Archives of Physiotherapy

Vercelli et al. Archives of Physiotherapy (2018) 8:1 DOI 10.1186/s40945-017-0042-8 RESEARCH ARTICLE Open Access Are they publishing? A descriptive cross- sectional profile and bibliometric analysis of the journal publication productivity of Italian physiotherapists Stefano Vercelli1* , Elisa Ravizzotti2 and Matteo Paci3 Abstract Background: In a clinical science-based profession such as physiotherapy, research is mandatory to update knowledge and to provide cost-effective, high quality treatments. This study aimed to provide point prevalence of Italian physiotherapists who are academics, holding a PhD degree, or being authors of scientific papers. The scientific journal productivity of physiotherapists was also thoroughly analyzed. Methods: A descriptive cross-sectional study was carried out on all Italian physiotherapists. Academics, postdoctoral research fellows, and PhD graduates were identified by searching the Italian Ministry of Education, University and Research (MIUR), Italian Society of Physiotherapy, and university websites. Then, authors of articles indexed in Scopus were searched. The following data were extracted: type of affiliation, authorship order, H-index, number of publications and citations, name of journals, year of publication, and journal’s Impact Factor. Results: The prevalence of academics, physiotherapists holding a PhD, or being author was 0.01%, 0.05%, and 0.56%, respectively. We identified 1083 papers co-authored by Italian physiotherapists, and their number has progressively increased over the years (p < 0.001). There was a significant difference between researchers and clinicians in both publication productivity (p < 0.01), citations (p < 0.01), and H-Index (p = 0.05). Articles were published in 359 different journals, receiving a total of 13,373 citations. Conclusions: Despite the low prevalence of faculty members and the reduced availability of PhD programs in Italy (forcing some students to study abroad), the quantity and quality of journal productivity is growing fast, and an increasing number of physiotherapists are involved in research activities. Keywords: Bibliometrics, Authorship, Scientific output indicators Background guidelines. To transfer knowledge to practice and obtain The credibility and professional development of physio- the expected impact, the results of research must be therapy is growing fast, thanks to the scientific know- documented in publications in order to have them ledge produced by their members and the clinical validated and made legitimate for use [2]. application of proven research findings [1]. In a health- care system that requires treatment to be of high quality In a previous study [3], the scientific productivity of and cost effective, research is mandatory to validate Italian physiotherapists was analyzed. Results indicated a current practice and guide the development of clinical point prevalence of authors in 2012 of about 0.34% - with authors being prevalently from the northern and * Correspondence: [email protected] central regions of Italy (Tuscany, Emilia Romagna and 1Laboratory of Ergonomics and Musculoskeletal Disorders Assessment, Piedmont) - and a steady increase in the number of pub- Division of Physical Medicine and Rehabilitation, Istituti Clinici Scientifici lished articles over years. The authors were working Maugeri SpA-SB, Via per Revislate 13, I-28010 Veruno, NO, Italy most frequently in non-scientific institutes (67%), but Full list of author information is available at the end of the article those working in research institutes had higher © The Author(s). 2017 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.

Vercelli et al. Archives of Physiotherapy (2018) 8:1 Page 2 of 6 bibliometric indicators. The academic professional To find physiotherapists holding a PhD, three strategies profile of authors was not fully explored, because at that were used. First, all PhD programs offered to physiothera- time there was only one physiotherapist working in a pists in the last 5 years were extracted from the MIUR full-time academic position. However, in the last few database. Each PhD program’s website - if available - was years substantial changes have taken place in physiother- then searched to find information on students and former apy education in Italy. The Master’s level degree (the students, and physiotherapists were identified. Second, the ‘Laurea Magistrale’) was introduced in late 2004, and list of PhDs and PhD students available on the Italian the first physiotherapists graduated in 2008. Since then, Society of Physiotherapy (SIF) website was downloaded a few programs have begun offering the Doctor of [7]. Third, each PhD or PhD student identified was con- Philosophy (PhD) degree to physiotherapists in our tacted to find any other colleagues or fellow students. country. Moreover, in 2012 the Italian Ministry of Education, University and Research (Ministero dell’Istru- Italian physiotherapists who published in peer-review zione, dell’Università e della Ricerca, MIUR) introduced journals indexed in the Scopus database were then a new process for the appointment of professors, which searched with a previously described strategy [3]. has been described elsewhere [4]. These substantive changes have prompted the need to determine more For each author the following data were extracted: precisely the profile of Italian physiotherapists involved type of work based on affiliation (academics - i.e. profes- in research activities, and to provide quantitative indica- sors and researchers - and those working in scientific tors of their productivity. institutes such as the Istituti di Ricovero e Cura a Carattere Scientifico, were termed ‘researchers,’ while those working The primary aim of this study was to provide point in hospitals, hospices, retirement homes, or professional prevalence of Italian physiotherapists who: a) have uni- studies were termed ‘clinicians’), total number of publica- versity roles; b) holding a PhD degree; and c) have writ- tions per author, number of citations received per author, ten at least one article published in peer-review journals authorship order (first, second, and last position were indexed in the Scopus database. The secondary objec- considered as ‘most relevant’ in multi-authored articles, po- tives were to describe the profile of academics and edu- tentially reflecting a greater contribution from those cational programs, and to analyze the scientific journal authors than of collaborators in other positions), and productivity of authors. H-index. The H-index (i.e. the number of articles published by the researcher which obtained a citation number greater Methods than or equal to the number of articles) was proposed in This was a descriptive cross-sectional study. 2005 to quantify the productivity and the impact of re- searchers, and it is used also by the MIUR to evaluate the The study population was represented by all Italian activity of researchers. physiotherapists with a bachelor’s degree (3 years), work- ing either in Italy or abroad. Foreign professionals work- After deleting duplicates generated by authors’ collab- ing in Italy were excluded. oration, all articles published in journals indexed in Scopus were then identified, excluding conference pa- The number of Italian professionals was recently esti- pers and book chapters. No limitation on publication mated by the MIUR and the Italian Association of period or language was set. For each publication, the fol- Physiotherapists (AIFI) in a survey conducted to estab- lowing information were extracted: name of journal, year lish the national educational needs [5]. In this study, the of publication, total number of citations received, and data provided by the ministerial survey was used to journal’s Impact Factor (IF). The IF is a measure reflect- calculate prevalence. ing the yearly average number of citations of recent arti- cles published in that journal. In this study, the Web of To identify those who have university roles - i.e. Science 2015 IF values were used. professors and postdoctoral research fellows - the MIUR website [6] was searched. The academic section area Counts, percentages, and ratios were used to summarize (Settore Scientifico Disciplinare, SSD) tag was used to the data. A Mann-Whitney U test was used to analyze dif- retrieve positions in the sector “Sciences of nursing, re- ferences between researchers and clinicians in H-Index, habilitation and neuropsychiatric techniques” (MED/48). number of publications and citations received per author. The MED/48 filter option was selected, and academics The correlation between year and number of published were classified as: full professors, associate professors, articles was estimated by the Spearman rank-order correl- and researchers. The curriculum of each person was ation coefficient (rho). For all tests, the level of statistical then carefully analyzed to identify those who were phys- significance was set at 0.05. iotherapists. The same strategy was used to find post- doctoral research fellows and the university educational Results programs (bachelor, graduate, and PhD) offered in Italy. In 2017, the Italian population of physiotherapists was estimated at about 65,000 professionals [5].

Vercelli et al. Archives of Physiotherapy (2018) 8:1 Page 3 of 6 To date, 7 physiotherapists - 1 full professor, 1 full distribution of H-Index, number of articles published, temporary professor, and 5 researchers - are listed in the and number of citations received by researchers and MED/48 section area of Italian university faculty mem- clinicians are shown in Table 1. There was a significant bers, wich resulted in a prevalence of 0.01%. The full difference in both publication productivity (p < 0.01), professor was appointed in 2005, and the associate pro- citations received (p < 0.01), and H-Index (p = 0.05) fessor in late 2016. Besides the academics, the MIUR between researchers and clinicians. website currently lists 4 postdoctoral research fellows among physiotherapists. In the sequences of authors, Italian physiotherapists were listed as authors in first, second, or last position (i.e. poten- A total of 85 bachelor (undergraduate) programs of tially reflecting higher contribution credit) 54% of times. physiotherapy are currently being taught in Italy, to which must be added 16 master (graduate) programs The number of articles published has increased stead- and 5 doctorate programs (XXXII cycle, year 2016) ily over the years (rho = 0.977, p < 0.001) (Fig. 2). accessible to physiotherapists. Thirty-one Italian physio- therapists holding a PhD have been identified (half of The 1083 papers identified received a total of 13,373 whom gained the title in the last 18 months), with a preva- citations. In 74% of the articles accounting for 71% of lence of 0.05%. At least 14 more physiotherapists are citations researches featured as authors; in 16% of the currently in training, either in Italy or abroad (Belgium, articles accounting for 21% of citations clinicians fea- Great Britain, Denmark, Netherlands, and Spain). tured, while in 10% of articles receiving 8% of the total citations the authors represented a collaboration be- Up to 2016, 363 Italian physiotherapists were identified as tween the two (Table 2). having been involved in scientific journal publication (Fig. 1), wich resulted in a prevalence of 0.56%. Of all authors, 242 Overall, 359 different journals were identified, only (67%) were researchers and 121 (33%) were clinicians. one-third of them being listed in the Web of Science Rehabilitation category; 252 journals had an IF, and 80% Overall, they appeared as authors 1820 times and, after of all articles were published in these journals (Table 3). deleting duplicates generated by collaborations, they produced a total of 1083 different articles published in Discussion journals indexed in the Scopus database. The Conducting research and publishing in scientific journals is regarded as a mandate for researchers, especially those Fig. 1 Flow chart of authors selection process

Vercelli et al. Archives of Physiotherapy (2018) 8:1 Page 4 of 6 Table 1 Distribution of articles published and citations received ministerial committee and may be called upon in the fu- by each author, and H-Index distribution for authors ture to fill an academic position [9]. All authors Researchers Clinicians Italian physiotherapists holding a Doctoral degree are N = 363 N = 242 N = 121 about 0.05%. This prevalence is still very low, but it is in line with emerging countries such as Brazil [10] and it is Articles published per author expected to grow very rapidly within the next few years, with many PhD students currently in training. Some of Cumulative 1820 1482 338 them have chosen to continue their studies abroad, be- cause of the few existing PhD programs in Italy. This Mean 5 6.1 2.8 highlights the intention of Italian physiotherapists to seek higher education and the urgent need to expand Interquartile Range 1–4 1–6 1–3 the number of research fellowships in this area. Min-Max 1–53 1–53 1–32 Despite the small number of academics and PhDs in physiotherapy in Italy, the national scientific output in the Citations per author rehabilitation area has achieved important bibliometric indicators worldwide: in the SCImago Country rank for Cumulative 21,063 16,551 4512 number of published articles Italy was 9th, with 301 pa- pers in 2016 and 4183 in the decade 1996–2016 [11]. Of Mean 58 68.4 37.3 note, not all of these articles were co-authored by physio- therapists, and not all of the authors were academics. Interquartile Range 2–48 3–54.5 2–36 In this study, more than 1000 articles were found to Min-Max 0–898 0–898 0–397 be co-authored by at least 363 physiotherapists, who most often had a scientific affiliation. In a previous re- H-Index view [3], 139 authors producing 517 articles were identi- fied between 1993 and 2012. In the last 5 years, the Cumulative 809 600 208 prevalence of authors passed from 0.34% to 0.56%, and both article production and number of authors has Mean 2.2 2.5 1.7 doubled. In the same period, the percentage of those af- filiated with universities or research institutes versus Interquartile Range 1–3 1–3 1–2 non-scientific institutes has reversed. Academic or scien- tific institute affiliations could facilitate initiatives and Min-Max 0–16 0–16 0–11 opportunities, establishing contacts between colleagues, producing new publications and sharing knowledge. % H-Index >2 17% 29% 12% Besides their own scientific activity, senior researchers working within universities. The primary purpose of this study was to calculate point prevalence of Italian physio- therapists who are academics, holding a PhD degree, or being authors of scientific papers. The very low prevalence (0.01%) of physiotherapists who are professors or academic researchers in Italy has already been described in the recent past [4, 8]. However, since 2012 the procedure of appointing new professors in Italy has changed and in April 2017 seven Italian physiotherapists were qualified by a national Fig. 2 Number of published articles per year

Vercelli et al. Archives of Physiotherapy (2018) 8:1 Page 5 of 6 Table 2 Distribution of citations per article among author of the academic milieu, the development of a profession affiliation types depends also on its members, working to refine and expand the body of knowledge that guides practice in All Only Only Researchers + that area. Although clinical physiotherapists generally devote little time to research activities, they are in a authors Researchers Clinicians Clinicians better place to carry out clinical trials, clinical auditing, case studies, and single-subject experimental designs [12]. Articles 1083 801 169 113 authored Criteria for authorship have been discussed at length, but a simple way to determine credit associated with the Citations per article sequence of authors’ names is still missing [13]. In multi-authored papers, the first author position is trad- Cumulative 13,373 9466 2847 1060 itionally assigned to the person who makes the greatest contribution, and the subsequent order of authors re- Mean 12.3 11.8 16.8 9.4 flects the descending importance of their contribution. However, evaluation committees and funding bodies Interquartile Range 1–14 0–13 1–23 1–12 often consider last authorship as a sign of successful group leadership and make this a criterion in hiring, Min-Max 0–194 0–194 0–182 0–52 granting, and promotion [13]. In this study, higher credit was arbitrarily assigned to the first, second, and last au- and professors also play an important role as mentors, thor’s position. Even in multi-authored, multidisciplinary serving as a catalyst for young researchers and promot- joint papers, Italian physiotherapists were ranked in one ing their development. However, it should be noted that of these positions more than half of the time, potentially they rarely are full-time researchers (as is the case in reflecting an important contribution to authorship. other countries) but are more often employed with a combined (or even prevalently) clinical role. Researchers Since the first publication co-authored by an Italian had a significantly higher H-index (p = 0.05), they were physiotherapist appeared in 1981 [14], the number of the profile most often involved in publishing (p < 0.01) articles published per year has steadily increased (Fig. 1), and they received more citations (p < 0.01) than clini- with more than 50% of all publications produced be- cians. While it was not surprising that researchers had a tween 2012 and 2016. However, it should be noted that greater scientific output (Table 1), it was surprising that a small group of highly productive physiotherapists the articles that received a higher average number of ci- (about 25) were responsible for more than 20% of all tations were those published by clinicians (Table 2). This publications. Articles were published in 359 different may be explained by different reasons. One may be that journals being indexed in Scopus, of which 252 had an researchers have published many more articles, and es- IF. The mean IF of these journals was 2.81 (ranging from pecially the recent ones not yet cited may have reduced 0.075 to 44.002, interquartile range [IQR] 1.839–3.057). the mean citations number. For example, one of the re- In 2015, the mean IF of the 113 journals indexed in the searchers published 40 articles that received 341 total Web of Science Rehabilitation category was 1.324 (ran- citations. Not considering the 15 articles that have been ging from 0.094 to 4.035, IQR 0.73–1.711). These data written in the last 2 years and which have received 13 demonstrate the high average quality of the scientific citations together, the average citations per article of that production of Italian physiotherapists, who have author would go from 8.5 to 13. On the other hand, this published a significant proportion of their studies in demonstrates also the good quality of the articles written top-level international journals. by clinicians. Over and above the scholarly expectations Journal productivity is a complex issue reaching be- Table 3 Counts of journals and articles published with their IF yond the existence of research degree programs [1], and distribution the results of this study demonstrate that good levels of publication productivity may be reached regardless of Number the insufficient number of academic figures, institutions, or leadership of the programs. In place of university, the Journals function of promoting culture and communication with professional networks, research emphasis, assertive- All 359 participative governance, mentoring and motivation, is currently carried out by scientific societies and associa- Journals with IF 252 tive scientific groups. However, we hope that in the near future the number of faculty members and PhD fellows Included in WOS Rehabilitation Category 113 will be further increased in Italy, and that journals Articles published All 1083 in Journals with IF 844 Mean Journal IF 2.81 Interquartile Range Journal IF 1.839–3.057 Min-Max Journal IF 0.075–44.002 Legend: WOS Web of Science, IF impact factor

Vercelli et al. Archives of Physiotherapy (2018) 8:1 Page 6 of 6 supported by scientific societies - such as the Archives of Ethics approval and consent to participate Physiotherapy - will get Scopus indexing. This could Not applicable. finally promote development of physiotherapy leadership and governance of university programs. Consent for publication Not applicable. The conclusions that we can draw from this investiga- tion may be limited by some factors. First, the approach Competing interests used to find authors was not exhaustive because no The authors declare that they have no competing interests. database provides search options by both profession and country. Second, only the journals indexed in the Scopus Publisher’s Note database were considered. Although this choice may have underestimated the results, we believe that it was Springer Nature remains neutral with regard to jurisdictional claims in justified since Scopus provides a strong coverage of the published maps and institutional affiliations. physiotherapy literature [15] and it is currently used by the MIUR to evaluate researcher’s journal production. Author details Then, internet sites were used to find academics and 1Laboratory of Ergonomics and Musculoskeletal Disorders Assessment, university programs instead of directly contacting the Division of Physical Medicine and Rehabilitation, Istituti Clinici Scientifici MIUR or universities. Therefore, the accuracy of infor- Maugeri SpA-SB, Via per Revislate 13, I-28010 Veruno, NO, Italy. 2Master of mation depends on the state of updating their websites. Science in Rehabilitation Sciences of the Health Professions, University of Moreover, publication productivity was measured by Genova, Varese, Italy. 3Unit of Functional Recovery, Azienda USL Toscana standardized bibliographic indicators based on number Centro, Area Prato, Italy. of articles and citations received, such as H-index and IF. We recognize that there are other forms of publica- Received: 9 September 2017 Accepted: 14 December 2017 tions and indexes, and that numerous criticisms have been made about the use of these particular parameters References [16]. However, this choice was done because of the avail- 1. Ritcher RR, Schlomer SL, Krieger MM, Siler WL. Journal publication ability of information and their widespread use to analyze the impact of publications on the scientific com- productivity in academic physical therapy program in the United States and munity [17] and to monitor scientific productivity of Puerto Rico from 1988 to 2002. Phys Ther. 2008;88:376–86. physiotherapists in other countries [1, 2, 10]. 2. Hamzat TK, Fatudimu MB. Autorship patterns in Nigerian physiotherapy and rehabilitation journals: are Nigerian physioterapists publishing? J Nig Soc Conclusions Physiother. 2009;17:19–22. This study presented an updated point prevalence of 3. Paci M, Plebani G. Scientific publication productivity of Italian Italian physiotherapists who are academics, holding a PhD physiotherapists. It J Physiother. 2013;3:170–3. degree, or being authors of scientific papers. The scientific 4. Gatti R, Paci M, Vercelli S, Baccini M. Has the Italian academia missed an journal productivity of physiotherapists was also thor- opportunity? Phys Ther. 2014;94(9):1358–60. oughly analyzed. Results indicate that faculty members 5. Available at: https://aifi.net/fabbisogno-miur-rivede-decreto-osservazioni-ora- among physiotherapists are still very low, and there are incontro/. Accessed 10 Nov 2017. insufficient PhD programs to meet the demand. However, 6. Available at: http://cercauniversita.cineca.it/php5/docenti/cerca.php. the quantity and quality of journal publication productiv- Accessed 1 Sep 2017. ity is growing fast, with an increasing number of physio- 7. Available at: http://www.sif-fisioterapia.it/?p=3191. Accessed 1 Sep 2017. therapists involved in research activities. 8. Gatti R, Paci M, Vercelli S, Baccini M. Teaching how to improve activities and participation of elderly subjects: the carelessness of the Italian academia Abbreviations shown by the national qualification for physiotherapists. Aging Clin Exp Res. IF: Impact factor; PhD: Doctor of Philosophy degree 2015;27(2):243–4. 9. Available at: http://www.sif-fisioterapia.it/?page_id=909. Accessed 1 Sep 2017. Acknowledgements 10. Sturmer G, Viero CCM, Silveira MN, Lukrafka JL, Plentz RDM. Profile and The author thanks Rosemary Allpress for the editorial help. scientific output analysis of physical therapy researchers with research productivity fellowship from the Brazilian National Council for Scientific and Funding Technological Development. Braz J Phys Ther. 2013;17(1):41–8. None. 11. Available at: http://www.scimagojr.com/countryrank.php?area=2700&year= 2016&category=2742. Accessed 1 Sep 2017. Availability of data and materials 12. Robertson VJA. Quantitative analysis of research in physical therapy. Phys Not applicable. Ther. 1995;75(4):313–27. 13. Tscharntke T, Hochberg ME, Rand TA, Resh VH, Krauss J. Author sequence and credit for contributions in multiauthored publications. PLoS Biol. 2007;5(1):e18. 14. Fattirolli F, Gori Savellini S, Martorana M, Bertini G, Margheri M, Baldini S, Bannister J, Cortini S, Fabbri D. Verification by the medico-psychological profile of the results of a rehabilitation program in patients with myocardial infarct. [article in Italian]. Boll Soc Ital Cardiol. 1981;26(10):1293–8. 15. Fell DW, Burnham JF, Buchanan MJ, Horchen HA, Scherr JA. Mapping the core journals of the physical therapy literature. J Med Libr Assoc. 2011;99(3):202–7. 16. Trapp J. Web of science, Scopus, and Google scholar citation rates: a case study of medical physics and biomedical engineering: what gets cited and what doesn't? Australas Phys Eng Sci Med 2016. [Epub ahead of print]. 17. Baccini A. Valutare la ricerca scientifica. Il Mulino: Uso e abuso degli indicatori bibliometrici. Bologna; 2010. Authors’ contributions All authors contributed to all the steps of conceiving, writing, and finalizing this manuscript.

Vancampfort et al. Archives of Physiotherapy (2018) 8:2 DOI 10.1186/s40945-018-0043-2 REVIEW Open Access Physiotherapy for people with mental health problems in Sub-Saharan African countries: a systematic review Davy Vancampfort1,2* , Brendon Stubbs3,4, Michel Probst1 and James Mugisha5,6 Abstract Background: There is a need for psychosocial interventions to address the escalating mental health burden in Sub-Saharan Africa (SSA). Physiotherapists could have a central role in reducing the burden and facilitating recovery within the multidisciplinary care of people with mental health problems. The aim of this systematic review was to explore the role of physiotherapists within the current mental health policies of SSA countries and to explore the current research evidence for physiotherapy to improve functional outcomes in people with mental health problems in SSA. Methods: The Mental Health Atlas and MiNDbank of the World Health Organization were screened for the role of physiotherapy in mental health plans. Next, we systematically searched PubMed from inception until August 1st, 2017 for relevant studies on physiotherapy interventions in people with mental health problems in SSA. The following search strategy was used: “physiotherapy” OR “physical therapy” OR “rehabilitation” AND “mental” OR “depression” OR “psychosis” OR “schizophrenia” OR “bipolar” AND the name of the country. Results: The current systematic review shows that in 22 screened plans only 2 made reference to the importance of considering physiotherapy within the multidisciplinary treatment. The current evidence (N studies = 3; n participants = 94) shows that aerobic exercise might reduce depression and improve psychological quality of life, self-esteem, body image and emotional stress in people with HIV having mental health problems. In people with depression moderate to high but not light intensity aerobic exercise results in significantly less depressive symptoms (N = 1, n = 30). Finally, there is evidence for reduction in post-traumatic stress symptoms (avoidance and arousal), anxiety and depression following body awareness related exercises (N = 1, n = 26). Conclusions: Our review demonstrated that physiotherapy is still largely neglected in the mental health care systems of SSA. This is probably due to poor knowledge of the benefits of physiotherapy within mental health care by policymakers, training institutes, and other mental health care professionals in SSA. Based on the current scientific evidence, this paper recommends the adoption of physiotherapy within mental health care services and investment in research and in training of professionals in SSA. Keywords: Physiotherapy, Physical therapy, Mental health services, Sub-Sahara Africa * Correspondence: [email protected] 1Department of Rehabilitation Sciences, KU Leuven – University of Leuven, Tervuursevest 101, 3001 Leuven, Belgium 2KU Leuven – University of Leuven, University Psychiatric Center KU Leuven, Leuvensesteenweg 517, 3070 Kortenberg, Belgium Full list of author information is available at the end of the article © The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.

Vancampfort et al. Archives of Physiotherapy (2018) 8:2 Page 2 of 8 Background cadres as compared to occupational therapists and phys- Mental and substance use disorders are the leading iotherapists [13]. cause of years lived with disability (YLD) in Sub- Saharan Africa (SSA), accounting for almost 20% of all There is however worldwide an increasing body of disability-associated burden [1]. The burden is esti- research demonstrating that physiotherapy can im- mated to increase even further with 130% and YLDs prove physical, mental and social health outcomes in will rise from between 20 to 45 million YLDs by 2050 people with mental health problems [17, 18]. For ex- [2] and will be equivalent to about two thirds the YLDs ample, it has been shown that aerobic and strength of the entire non-communicable diseases group, which exercises and yoga reduce mental health and cogni- is estimated at 67million YLDs [3]. The consequences tive symptoms, state anxiety, and psychological dis- of this rising burden of mental and substance use disor- tress while it improves health-related quality [17, ders will be devastating. The quality of life of those 18]. Progressive muscle relaxation reduces on its affected will be impacted and economic costs will be turn state anxiety and psychological distress [17, 18]. tremendous in these low-resourced settings [4]. More- Thus, implementation of physiotherapy within the over, secondary co-morbidities need to be considered care of people with mental health problems could which can add substantively to the increased disability reduce the mental, physical and social burden, while and burden [5, 6]. For example, severe mental illness facilitating functional recovery and consequently re- has been shown to be an independent risk factor for ducing disability. This will on its turn reduce the so- other important non-communicable disorders such as cietal costs. cardio-metabolic diseases, albeit inconsistently in SSA studies [7]. Consistent associations are however re- The aim of the current systematic review is twofold. ported between HIV/AIDS and chronic pain [8, 9] and First, we set out to explore the role of physiotherapy between HIV/AIDS and poor mental health [10, 11]. within the current mental health policies and plans of The burden of mental health problems, HIV/AIDS and SSA countries. Specifically, we wanted to explore which chronic pain will have larger consequences for women roles were reported and defined for physiotherapists. because if their productivity is affected, this directly af- Second, we explored the current research evidence for fects their family welfare and increases the scale of both physiotherapy in people with mental health problems in family and community poverty. Moreover, in SSA, Sub-Saharan Africa. many women are relying on labor-demanding jobs in the informal sector with no job security or compensa- Methods tion for lost income. Maintaining physical strength and an adequate activity level is therefore crucial for their Screening for the role of physiotherapists in mental livelihoods. health policies and plans in sub-Saharan Africa In a first stage, we screened the latest Mental Health So far, a mental health policy has been relatively Atlas [19]. If the country data were not available in the low on the priority list in most of SSA countries, al- latest edition, the penultimate edition was screened. though interest is increasing [12, 13]. However, in With data from 171 World Health Organization (WHO) most countries still less than 1% of the health budget Member States, the Mental Health Atlas provides a is spent on mental health [14]. As a result, mental comprehensive overview of mental health policies, plans health services are poorly resourced and treatment and services worldwide. Data abstracted were the pres- rates for people with mental disorders remain low, ence of a mental health policy and/or plan (yes or no). with treatment gaps over 90% [15]. To date, within low-to-middle income countries, community-based re- In a second stage, if a mental health policy and/or habilitation, psychoeducation and support for families plan was available the full-text documents were re- (delivered by non-specialists) are recommended for trieved via the MiNDbank of the World Health low resource settings, with assertive community care Organization [20]. Policies written in English, French, and cognitive therapy recommended as additions in Spanish or Portuguese were evaluated. If the mental higher resourced settings with stronger service- health policy and/or plan was not available, google delivery platforms [16]. scholar was screened using the search terms: “mental health” AND “plan” OR “policy” and the name of the Since the emphasis in health service delivery in SSA is country, or its equivalents in other languages. Mental based on the biomedical model (versus the biopsychoso- health policies and plans were screened for the role cial model) with an important focus on pharmacology in of physiotherapy. Search terms used, were: “physical the management of mental health problems, the poten- therap*” OR “physiotherap*” or its equivalents in tial role of physiotherapy seems to be neglected [13]. As other languages. a result, more doctors and nurses are recruited as key In a third stage, we summarized the reported role of physiotherapists in the policy plans.

Vancampfort et al. Archives of Physiotherapy (2018) 8:2 Page 3 of 8 Identification of physiotherapy studies conducted in Study design We included pre- and post-test studies people with mental health problems in SSA without a control group and randomized (RCTs) or non- Search strategy randomized clinical controlled trials (NRCCTs) in which We systematically search PubMed from inception until the experimental and control intervention were of simi- August 1st 2017 for relevant studies on physiotherapy in lar duration. Qualitative and quantitative data were people with mental health problems in SSA. The follow- included. ing search strategy was used: “physiotherapy” OR “phys- ical therapy” OR “rehabilitation” AND “mental” OR Exclusion criteria In case of overlap only the most “depression” OR “psychosis” OR “schizophrenia” OR “bi- recent study or the study with the largest dataset were polar” AND the name of the country. We did not place included. No additional exclusion criteria were applied. a language restriction upon these searches. In addition, the reference lists of all eligible articles were scanned to Study selection assess eligibility of additional studies. Two reviewers (DV and BS) screened titles and abstracts of all potentially eligible articles. Both authors applied Eligibility criteria eligibility criteria, and a list of relevant studies was developed through consensus. When necessary, the Participants As we were interested in prevention and protocol stated that the corresponding author of a study treatment of mental health problems, we did not exclude would be contacted up to two times in a 4-week period any people due to age or whether or not they were diag- to request data that would enable inclusion in the nosed with Statistical Manual (DSM) [21, 22], Inter- current analyses. national Classification of Disease (ICD) [23] or with validated diagnostic tools. Data extraction Two authors (DV, BS) extracted data using a predeter- Interventions The physiotherapy interventions could mined data extraction form. The data extracted for were comprise aerobic exercises, strength exercises, relaxation country, study setting, patient characteristics (diagnosis, training, basic body awareness exercises, or a combination age, % male). Duration (weeks) frequency (times per of these in accordance with the World Confederation for week), intensity (as defined by the authors), and type Physical Therapy position statement. A physiotherapy inter- (aerobic exercise, resistance training, relaxation training, vention could be used alone or in conjunction with other basic body awareness exercises or a mixed combination) interventions, with physiotherapy being considered the of the physiotherapy intervention. Finally, we did extract main or active element. Interventions that included physio- any relevant outcomes and motivational strategies used therapy in a multiple component weight management pro- to improve adherence and reduce dropout from the gram were excluded because the specific effects of the physiotherapy interventions. physiotherapy intervention could not be addressed. Other interventions could include any of the following: pharmaco- Methodological quality assessment of the RCTs therapy, psycho-education, and cognitive-behavioural or Two authors (DV and BS) assessed studies on the pres- motivational techniques related to changing physical activ- ence of high, low risk or unclear risk of bias according ity behaviour. Excluded were: (a) peer-led physical activity to the Cochrane Handbook definition [24]. Studies pre- interventions, and (b) mindfulness-based cognitive behav- senting adequate allocation concealment and complete ioural therapy. presentation of outcome data (intention-to-treat ana- lysis) and blinding outcome assessors are considered Control conditions The presence of control conditions studies with low risk of bias (high quality trials). was not an inclusion criterion. However, if studies were clinical or randomized controlled, usual-care or wait-list Results control conditions were included. Usual care, if available, was defined as care that people would normally receive Physiotherapy priorities in mental health policies and had they not been included in a research trial. Such care plans in sub-Saharan Africa would include medication, hospitalization, community In terms of policy, 69% (=33/48) of SSA countries report support, and outpatient care. having a mental health policy or plan. Ten policies were not found while one (Sudan) was written in Arabic, and Outcome measure The primary outcome measure was therefore not meeting our inclusion criteria and was not any mental or social health outcome measures. Second- screened. Two of 22 screened mental health policies or ary outcomes were physical health outcomes in mentally plans included a physiotherapy component. An overview ill populations. of the presence of a mental health policy or plan and the presence of a physiotherapy component is presented in

Vancampfort et al. Archives of Physiotherapy (2018) 8:2 Page 4 of 8 Table 1. The roles described in the mental health pol- depression were reported following body awareness re- icies of Namibia and Nigeria are prescribed in Table 2. lated exercises. Details of the physiotherapy intervention Briefly, while physiotherapists were only described as characteristics and outcomes are presented in Table 3. member of the multidisciplinary team in Nigeria, in Namibia a more extensive role was defined including fa- Discussion cilitating the integration of mental health in primary The current systematic review shows that in those SSA health care, identifying training needs and assisting in countries that have a mental health policy or plan (70%) public mental health campaigns. only two make reference to a role for physiotherapy. Therefore, although physiotherapy is becoming acknowl- Screening for physiotherapy studies conducted in people edged as an important adjunct in the management of with mental health problems in SSA mental health problems [30, 31], the potential is yet to Search results be embraced in SSA countries. The lack of priority given Out of 7859 search hits, 16 potentially eligible studies to physiotherapy as a complementary treatment is also were retrieved. After applying the eligibility criteria 5 mirrored in the limited number of studies exploring the [25–29] studies were included. An overview of the importance of physiotherapy interventions in the man- search results for each country is presented in Table 1. agement of mental health problems. Five studies, of Reasons for exclusion were: not limited to SSA countries which 4 RCTs, were found indicating that aerobic exer- (n = 1), peer-led interventions (n = 1), no relevant out- cise can improve mental health outcomes while there is comes (n = 2), sport-for-development by community qualitative evidence from one study for improvements in members with no input from physiotherapists (n = 2), quality of life following body awareness related exercises. mindfulness based cognitive behavioral therapy (n = 2), and cross-sectional studies on physical activity (n = 3). It is plausible that due to the strong biomedical focus on pharmacotherapy [13] health care professionals in Participants and study characteristics SSA are yet to be fully aware of the psychosocial effects Four RCTs [25–28] and one qualitative study [29] of physiotherapy. Hence, a need to re-orient the current involving in total 241 participants were included. Three health care systems including policy makers to embrace RCTs [25, 27, 28] focused on mental health outcomes in physiotherapy in the management of mental health people with HIV and one focused on mental health out- problems is needed. Budget holders should also invest comes in people with depression [26]. All RCTs explored more in mental health research. In three of the five stud- aerobic exercise. The qualitative study explored the ies mental health outcomes were explored in people with effect of dance movement and body awareness related HIV. This is not surprising and due to increased funding exercises in torture survivors and former child soldiers of HIV/AIDS care and research on the continent. Men- (Sierra Leone). The interventions ranged from 6 to tal health research and in particular research exploring 26 weeks, from 30 to 90 min per week, from 2 to 3 times the role physiotherapists could benefit current mental per week and from low to high intensity. Details of the health care systems. Future physiotherapy studies in dif- participants and study characteristics are presented in ferent contexts in SSA could explore: Table 3.  whether individually prescribed physiotherapy can Methodological quality of RCTs improve mood, reduce stress, promote well-being Two of the four RCTs were considered to be of high and address somatic co-morbidities associated with methodological quality with low risk of bias (see Table 3). mental health diagnoses in inpatients, outpatients The most important risks increasing bias in the other and community patients. RCTs were lack of intention-to-treat analyses or a lack of blinding of the assessors.  whether treatment of physical problems and physical pain of people with mental health diagnoses Physiotherapy outcomes by physiotherapists will facilitate social participation In the 3 RCTs [25, 27, 28] in people with HIV aerobic and recovery. exercise reduced depression and improved psychological quality of life, self-esteem, body image and emotional  whether physiotherapy interventions can minimise stress. In one RCT [26] in depressed adults only moder- the metabolic and motor side-effects of some psy- ate and high but not light intensity aerobic exercise chotropic medications. resulted in significantly less depressive symptoms. In the qualitative study [29] a reduction in post-traumatic  the benefits of fall prevention for elderly with mental stress symptoms (avoidance and arousal), anxiety and health problems or for those on antipsychotic medication.  whether physiotherapy could improve self- management and coping strategies in the context of mental and physical health issues.

Vancampfort et al. Archives of Physiotherapy (2018) 8:2 Page 5 of 8 Table 1 Overview of the presence of a mental health policy/ Table 1 Overview of the presence of a mental health policy/ plan, whether there was a role for physiotherapists and plan, whether there was a role for physiotherapists and physiotherapy research results in Sub-Saharan African countries physiotherapy research results in Sub-Saharan African countries (n = 48) (n = 48) (Continued) Country Official mental Role for PubMed Country Official mental Role for PubMed health policy physiotherapists search health policy physiotherapists search or plan results or plan results (potential (potential relevant / relevant / obtained) obtained) Angola (2011) Yes No 0/10 Nigeria (2014) Yes Yes 5/878 Benin (2014) No / 0/65 Rwanda (2014) Yes No 1/65 Botswana (2014) Yes No 0/38 São Tomé and Yes NA 0/0 Burkina Faso Yes No 0/25 Príncipe (2014) (2014) Senegal (2014) No / 0/50 Burundi (2014) Yes No 0/14 Seychelles No / 0/9 (2014) Cameroon No / 0/48 (2011) Sierra Leone Yes No 1/16 Cape Verde (2011) Yes No 0/6 (2014) Central African Yes No 0/4 Somalia (2014) No / 0/54 Republic (2014) South-Africa Yes No 6/1632 Chad (2011) Yes NA 0/37 (2014) Comoros Yes NA 0/5 South-Sudan No / 0/7 (2014) (2011) Congo (2014) No / 0/76 Sudan (2011) Yes Not checkeda 0/89 Côte d’Ivoire Yes No 0/21 Swaziland No / 0/13 (2014) (2014) Democratic Yes No 0/36 Togo (2014) Yes NA 0/55 Rep. of the Uganda (2014) Yes No 2/294 Congo (2011) United Yes NA 0/123 Djibouti (2014) No / 0/6 Republic of Equatorial No / 0/0 Tanzania Guinea (2014) (2011) Eritrea (2011) No / 0/19 Zambia (2014) Yes No 0/44 Ethiopia (2014) Yes No 0/277 Zimbabwe Yes NA 0/102 (2014) Gabon (2011) No / 0/1 Summary 69% (33/48) has an % (2/22) described a 16/7859 Gambia (2014) Yes No 0/17 official mental health role for Ghana (2014) Yes No 1/127 policy physiotherapists Guinea (2014) Yes NA 0/3151 UN Unknown, NA Not available aplan written in Arabic Guinea-Bissau No / 0/0 (2011)  whether body awareness exercise can boost a Kenya (2011) Yes No 0/207 person’s self-esteem. Lesotho (2014) No / 0/7 Within mental health care systems, physiotherapists should be seen as valuable members of a multidisciplin- Liberia (2014) Yes No 0/14 ary approach. Several strategies to initiate and stimulate physiotherapy within the mental health care systems of Madagascar Yes No 0/12 SSA are possible. For example, continued medical edu- cation, which is a common practice in SSA [13], should (2014) be used to inform supervising mental health care profes- sionals on the importance of physiotherapy. Policy Malawi (2014) Yes No 0/59 makers should be made aware that investment in physio- therapy could optimize mental and physical health Mauritania Yes NA 0/7 improvements while inclusion of physiotherapists in life- style interventions will improve adherence and reduce (2011) Mauritius No / 0/16 (2014) Mozambique Yes No 0/13 (2014) Namibia (2014) Yes Yes 0/16 Niger (2011) Yes NA 0/94

Vancampfort et al. Archives of Physiotherapy (2018) 8:2 Page 6 of 8 Table 2 The role of physiotherapists in mental health plans of therapeutic interactions (e.g. communication, attitude) Sub-Saharan African countries and in basic motivational skills (e.g., motivational inter- viewing). Research should explore the most optimal Country Role of physiotherapists form of physiotherapy delivery in different contexts in low-resourced settings. For example, physiotherapy tri- Namibia Physiotherapists should: als in different settings could explore whether assisting -facilitate the integration of the mental health policy into people with mental health problems in fulfilling three primary health care universal, psychological needs: (a) the need for auton- -identify specific training needs in mental health care omy (i.e., experiencing a sense of psychological freedom -provide input into the development of (the World Health when engaging in physiotherapy exercises), (b) the need Organization affiliated) Information, Education and for competence (i.e., ability to attain desired outcomes Communication materials, policies, guidelines and standards following physiotherapy), and (c) the need for related- for mental health ness (i.e., feeling socially connected) will increase the likelihood that they adopt or maintain for example an Nigeria Physiotherapists are involved in mental health care services, active lifestyle [34]. but a specific role was not defined. The current findings should be interpreted in light of dropout [32, 33] and consequently will be cost- some methodological limitations. First of all, as we only effective. Physiotherapy institutions should focus on the searched PubMed, several relevant articles might be importance of physiotherapy for mental health in order missed. However, to minimize this risk, we performed to improve the competencies of their graduates in this additional hand search by reviewing references listed in field. All physiotherapists should be able to recognize the included original publications. Second, due to the the signs and symptoms of the main disorders and demonstrate basic knowledge of the causes. Next to this, physiotherapists should be trained in basic Table 3 Mental and/or physical health outcomes in physiotherapy related studies in Sub-Saharan Africa First author Country Design Participants Physiotherapy intervention Mental and/or physical health MQ outcomes* Aweto Nigeria RCT 18 (32.1 ± 5.4 years) outpatients 6 weeks, 3*week, 30 min The Beck Depression Index score 2016 with HIV; BMI = 26.1 ± 1.4 vs 15 moderate intensity aerobic only reduced significantly in the exercise controls with HIV with care as exercise on a cycle ergometer group [10.3 ± 6.5 vs.3.5 ± 1.3;P < 0.001] usual (30.7 ± 5.8 years); 10♂/33 provided by a physiotherapist Balchin South- RCT 30♂ moderately depressed; 6 weeks, 3*week, 60 min high The HAM-D (15.9 ± 1.8 vs. 5.7 ± 5.8 and + 2016 Africa mean age = 25.4 years, mean vs moderate vs low intensity 16.4 ± 1.4 vs. 6.6 ± 5.0 vs. 17.1 ± 1.2 vs. BMI = 26.9 aerobic exercise; providers 11.8 ± 3.9, respectively) and MADRS 12.7 unknown ± 4.0 vs. 7.0 ± 6.7 and 14.4 ± 4.3 vs. 9.0 ± 6.7 vs. 18.8 ± 6.4 vs. 15.0 ± 5.2, respect ively) only reduced significantly in the high and moderate intensity aerobic exercise Maharaj South- RCT 26 (16♂) (19–58 years) weekly aerobic exercise on Significant improvements in all SF-36 + 2011 Africa outpatients on antiviral cycle ergometer (2*10 min) and domains (P < 0.05) for the experimental therapy vs 26 (18♂) treadmill (2*10 min) at 50–70% group compared with the control (22–51 years) of the age predicted maximum group, with the physical SF-36 summary heart rate for 3 months scores (P < 0.018) and mental SF-36 summary scores (P < 0.021) scores being significantly higher after exercise. Mutimura Rwanda RCT 50 (20♂) (37.5 ± 6.9 years) 26 weeks, 3*week, 90 min At 6 months, scores on psychological 2008 outpatients with HIV; 88% moderate intensity aerobic quality of life [1.3 ± 0.3 vs. 0.5 ± 0.1; employed; BMI = 24.4 ± 2.7; and resistance training; P < 0.0001], self-esteem [1.3 ± 0.8 vs. 0.1 20% smoking vs 50 (20♂) providers unknown ± 0.6); P < 0.001], body image [1.5 ± 0.6 controls with HIV with care vs. 0.0 ± 0.5; P < 0.001] and emotional as usual (37.8 ± 5.5 years) stress [1.6 ± 0.7 vs. 0.2 ± 0.5; P < 0.001], improved more in the exercise group Harris 2007 Sierra Qualitative Three studies: [1] 6♀ 9 to 16 weekly sessions of Reduction in self-reported post- NA Leone (16–17 years), [2] 8♂ (23–24 years), [3] 12♂ dance movement therapy with traumatic stress symptoms (avoidance (8 aged 18) child soldiers body awareness exercises within and arousal), anxiety and depression and torture survivors psychotherapy RCT Randomized controlled trial, HAM-D Hamilton depression score, MADRS Montgomery-Åsberg Depression Rating Scale, SF-36 Health Related Quality of Life Short Form – 36. MQ Methodological quality: risk of bias was assessed on random sequence generation, allocation concealment, blinding of participants, blinding of those delivering the intervention, blinding of outcome assessors, incomplete data outcome, selective reporting or others. Studies presenting adequate allocation concealment and complete presentation of outcome data (intention-to-treat analysis) and blinding outcome assessors are considered studies with low risk of bias (high quality trials, coded with “+”); NA = not applicable (no RCT)

Vancampfort et al. Archives of Physiotherapy (2018) 8:2 Page 7 of 8 limited literature available we were not able to perform a Received: 18 August 2017 Accepted: 18 January 2018 more rigorous meta-analysis. Conclusions References The current data shows that in SSA the importance 1. Whiteford HA, Degenhardt L, Rehm J, Baxter AJ, Ferrari AJ, Erskine HE, et al. of considering physiotherapy in mental health care is largely ignored in the current policies and within the Global burden of disease attributable to mental and substance use current research. Policy makers and existing work disorders: findings from the global burden of disease study 2010. Lancet. force should be informed about the potential of 2013;382(9904):1575–86. physiotherapy in the multidisciplinary treatment of 2. Institute for Health Metrics and Evaluation. The global burden of disease: people with mental health problems. There is a need generating evidence, guiding policy. Seattle: Institute for Health Metrics and to invest in physiotherapy in the mental health care Evaluation; 2013. settings, and research and training in institutes of 3. Charlson FJ, Diminic S, Lund C, Degenhardt L, Whiteford HA. Mental and SSA countries. There is already, however limited, evi- substance use disorders in sub-Saharan Africa: predictions of dence that the inclusion of physiotherapists in the epidemiological changes and mental health workforce requirements for the treatment will improve outcomes. next 40 years. PLoS One. 2014;9(10):e110208. 4. Bloom DE, Cafiero E, Jané-Llopis E, Abrahams-Gessel S, Bloom LR, Fathima S, Abbreviations et al. The global economic burden of noncommunicable diseases. Harvard: SSA: Sub-Saharan African; YLD: Years lived with disability Program on the Global Demography of Aging. Harvard School of Public Health; 2012. Acknowledgements 5. Vancampfort D, Stubbs B, Mitchell AJ, De Hert M, Wampers M, Ward PB, et None. al. Risk of metabolic syndrome and its components in people with schizophrenia and related psychotic disorders, bipolar disorder and major Funding depressive disorder: a systematic review and meta-analysis. World There was no funding. Psychiatry. 2015;14(3):339–47. 6. Vancampfort D, Correll CU, Galling B, Probst M, De Hert M, Ward PB, et al. Availability of data and materials Diabetes mellitus in people with schizophrenia, bipolar disorder and major Not applicable. depressive disorder: a systematic review and large scale meta-analysis. World Psychiatry. 2016;15(2):166–74. Authors’ contributions 7. Gelaye B, Williams MA, Lemma S, Berhane Y, Fann JR, Vander Stoep A, et al. Dr. D Vancampfort and Dr. J. Mugisha collected the data. Dr. D. vancampfort Major depressive disorder and cardiometabolic disease risk among sub- wrote a first draft which was reviewed and revised in several rounds by the Saharan African adults. Diabetes Metab Syndr. 2015;9(3):183–91. other co-authors. All authors approved the final version and all authors certify 8. Stubbs B, Mitchell AJ, De Hert M, Correll CU, Soundy A, Stroobants M, et al. that they have participated sufficiently in the work to believe in its overall The prevalence and moderators of clinical pain in people with validity and to take public responsibility for appropriate portions of its schizophrenia: a systematic review and large scale meta-analysis. Schizophr content. Res. 2014;160(1):1–8. 9. Stubbs B, Eggermont L, Mitchell AJ, De Hert M, Correll CU, Soundy A, et al. Ethics approval and consent to participate The prevalence of pain in bipolar disorder: a systematic review and large- Not applicable. scale meta-analysis. Acta Psychiatr Scand. 2015;131(2):75–88. 10. De Beaudrap P, Mac-Seing M, Pasquier E. Disability and HIV: a systematic Consent for publication review and a meta-analysis of the risk of HIV infection among adults with Not applicable. disabilities in sub-Saharan Africa. AIDS Care. 2014;26(12):1467–76. 11. Kinyanda E, Hoskins S, Nakku J, Nawaz S, Patel V. Prevalence and risk factors Competing interests of major depressive disorder in HIV/AIDS as seen in semi-urban Entebbe The authors declare that they have no competing interests. district. Uganda BMC Psychiatry. 2011;11(1):1. 12. Mugisha J. Towards understanding the dilemmas in prevention of mental Publisher’s Note illness in Uganda. New York: The Edwin Mellen Press; 2016. 13. Mugisha J, Ssebunnya J, Kigozi FN. Towards understanding governance Springer Nature remains neutral with regard to jurisdictional claims in issues in integration of mental health into primary health care in Uganda. published maps and institutional affiliations. Int J Ment Health Syst. 2016;10(1):1. 14. World Health Organization. Mental health atlas 2011. Geneva: World Health Author details Organization; 2011. 1Department of Rehabilitation Sciences, KU Leuven – University of Leuven, 15. Lund C, Alem A, Schneider M, Hanlon C, Ahrens J, Bandawe C, et al. Tervuursevest 101, 3001 Leuven, Belgium. 2KU Leuven – University of Leuven, Generating evidence to narrow the treatment gap for mental disorders in University Psychiatric Center KU Leuven, Leuvensesteenweg 517, 3070 sub-Saharan Africa: rationale, overview and methods of AFFIRM. Epidemiol Kortenberg, Belgium. 3Physiotherapy Department, South London and Psychiatr Sci. 2015;24(03):233–40. Maudsley NHS Foundation Trust, London, UK. 4Health Service and Population 16. de Jesus MJ, Razzouk D, Thara R, Eaton J, Thornicroft G. Packages of care for Research Department, Institute of Psychiatry, Psychology and Neuroscience, schizophrenia in low-and middle-income countries. PLoS Med. 2009;6(10): King’s College London, De Crespigny Park, London, UK. 5Butabika National e1000165. Referral and Mental Health Hospital, Kampala, Uganda. 6Kyambogo 17. Vancampfort D, Probst M, Skjaerven LH, Catalán-Matamoros D, Lundvik- University, Kampala, Uganda. Gyllensten A, Gómez-Conesa A, et al. Systematic review of the benefits of physical therapy within a multidisciplinary care approach for people with schizophrenia. Phys Ther. 2012;92(1):11–23. 18. Vera-Garcia E, Mayoral-Cleries F, Vancampfort D, Stubbs B, Cuesta-Vargas AI. A systematic review of the benefits of physical therapy within a multidisciplinary care approach for people with schizophrenia: an update. Psychiatry Res. 2015;229(3):828–39. 19. World Health Organization. Mental health atlas 2014: World Health Organization; 2015. 20. World Health Organisation. WHO MiNDbank: more inclusiveness needed in disability and dvelopment. Geneva: World Health Organization; 2016.

Vancampfort et al. Archives of Physiotherapy (2018) 8:2 Page 8 of 8 21. American Psychiatric Association. Diagnostic and statistical manual of mental disorders (DSM-5). Washington: American Psychiatric Association; 2013. 22. Association AP. Diagnostic and statistical manual of mental disorders (DSM-IV). Washington: American Psychiatric Association; 1994. 23. World Health Organization. The ICD-10 classification of mental and BehaviouralDisorders- diagnostic criteria for research. Geneva: World Health Organization; 1993. 24. Higgins J. Cochrane handbook for systematic reviews of interventions. Version 5.1. 0. The Cochrane collaboration. 2011. 25. Aweto HA, Aiyegbusi AI, Ugonabo AJ, Adeyemo TA. Effects of aerobic exercise on the pulmonary functions, respiratory symptoms and psychological status of people living with HIV. J Res Health Sci. 2016;16(1): 17–21. 26. Balchin R, Linde J, Blackhurst D, Rauch HL, Schönbächler G. Sweating away depression? The impact of intensive exercise on depression. J Affect Disord. 2016;200:218–21. 27. Maharaj SS, Chetty V. Rehabilitation program for the quality of life for individuals on highly active antiretroviral therapy in KwaZulu-Natal, South Africa: a short report. Int J Rehabil Res. 2011;34(4):360–5. 28. Mutimura E, Stewart A, Crowther NJ, Yarasheski KE, Cade WT. The effects of exercise training on quality of life in HAART-treated HIV-positive Rwandan subjects with body fat redistribution. Qual Life Res. 2008;17(3):377. 29. Harris DA. Dance/movement therapy approaches to fostering resilience and recovery among African adolescent torture survivors. Torture. 2007;17(2): 134–55. 30. Stubbs B, Probst M, Soundy A, Parker A, De Herdt A, De Hert M, et al. Physiotherapists can help implement physical activity programmes in clinical practice. Br J Psychiatry. 2014;204(2):164. 31. Stubbs B, Soundy A, Probst M, Parker A, Skjaerven LH, Lundvik GA, et al. Addressing the disparity in physical health provision for people with schizophrenia: an important role for physiotherapists. Physiotherapy. 2014; 100(3):185–6. 32. Stubbs B, Vancampfort D, Rosenbaum S, Ward PB, Richards J, Soundy A, et al. Dropout from exercise randomized controlled trials among people with depression: a meta-analysis and meta regression. J Affect Disord. 2016;190: 457–66. 33. Vancampfort D, Rosenbaum S, Schuch FB, Ward PB, Probst M, Stubbs B. Prevalence and predictors of treatment dropout from physical activity interventions in schizophrenia: a meta-analysis. Gen Hosp Psychiatry. 2016; 39:15–23. 34. Deci EL, Ryan RM. The\" what\" and\" why\" of goal pursuits: human needs and the self-determination of behavior. Psychol Inq. 2000;11(4):227–68. Submit your next manuscript to BioMed Central and we will help you at every step: • We accept pre-submission inquiries • Our selector tool helps you to find the most relevant journal • We provide round the clock customer support • Convenient online submission • Thorough peer review • Inclusion in PubMed and all major indexing services • Maximum visibility for your research Submit your manuscript at www.biomedcentral.com/submit

Miciak et al. Archives of Physiotherapy (2018) 8:3 https://doi.org/10.1186/s40945-018-0044-1 RESEARCH ARTICLE Open Access The necessary conditions of engagement for the therapeutic relationship in physiotherapy: an interpretive description study Maxi Miciak1* , Maria Mayan2, Cary Brown3, Anthony S. Joyce4 and Douglas P. Gross5 Abstract Background: The therapeutic relationship between patient and physiotherapist is a central component of patient-centred care and has been positively associated with better physiotherapy clinical outcomes. Despite its influence, we do not know what conditions enable a physiotherapist and patient to establish and maintain a therapeutic relationship. This knowledge has implications for how clinicians approach their interactions with patients and for the development of an assessment tool that accurately reflects the nature of the therapeutic relationship. Therefore, this study’s aim was to identify and provide in-depth descriptions of the necessary conditions of engagement of the therapeutic relationship between physiotherapists and patients. Methods: Interpretive description was the qualitative methodological orientation used to identify and describe the conditions that reflect and are practically relevant to clinical practice. Eleven physiotherapists with a minimum 5 years of clinical experience and seven adult patients with musculoskeletal disorders were purposively sampled from private practice clinics in Edmonton, Canada. The in-person, semi-structured interviews were completed in a location of the participant’s choice and were audio recorded and transcribed. Qualitative content analysis was used to analyze the textual data and constant comparison techniques were integrated to refine the categories and sub-categories. Rigour strategies used throughout the study were peer debrief, interview notes, reflexive journaling, memoing, member reflections, audit trail, and external audit. Results: Four conditions were identified as necessary for establishing a therapeutic relationship: present, receptive, genuine, and committed. These conditions represent the intentions and attitudes of physiotherapists and patients engaging in the clinical interaction. Although distinct, the conditions appear related as being present and receptive create a foundation for being genuine and committed. Conclusions: These conditions of engagement are needed for physiotherapist and patient to “be” in a therapeutic relationship. Although communication skills are important for advancing therapists’ relational abilities, awareness and integration of intentions and attitudes are essential for shaping behaviors that develop the therapeutic relationship. These findings also suggest there are characteristics of the therapeutic relationship specific to physiotherapy. Therefore, theories from other contexts (e.g., psychotherapy) should be used judiciously to guide physiotherapy practice and research. Keywords: Therapeutic alliance, Working alliance, Psychotherapy, Patient-therapist interaction, Patient-therapist relationship, Patient-centred care * Correspondence: [email protected]; [email protected] 1Alberta Innovates, 1500, 10104 – 103 Avenue NW, Edmonton, AB T5J 0H8, Canada Full list of author information is available at the end of the article © The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.

Miciak et al. Archives of Physiotherapy (2018) 8:3 Page 2 of 12 Background likely to form consistent relationships with their pa- The therapeutic relationship between patient and pro- tients (i.e., same therapist sees the patient) over the vider is considered a central component of patient- course of a treatment period than, for example, centred care [1, 2] and patient engagement [3, 4]. In nurses working in a hospital where shift changes re- physiotherapy, the therapeutic relationship is integrated quire a patient work with more than one nurse. into various practice standards [5], indicating its import- These factors could shape how physiotherapists ap- ance in shaping competent care. Research demonstrating proach interactions with patients and create an en- a positive association between better therapeutic rela- vironment that provides the opportunity to develop tionships and patient satisfaction [6], adherence with the therapeutic relationship as a central component treatment [7], and clinical outcomes [8–10] supports of the clinical interaction, as well as direct how the physiotherapists’ beliefs that the therapeutic relationship therapeutic relationship should be assessed. influences clinical outcomes [11]. The concept of engagement is an influential factor in Study of the therapeutic relationship in physiotherapy outcomes and has been linked to the therapeutic rela- is in its infancy, especially when compared to theoretical tionship. In their content analysis of patient engagement, development and empirical investigation in the psycho- Higgins et al. [3] determined that the therapeutic alli- therapy context. Despite its potential to impact clinical ance (a term used broadly as synonymous with the outcomes, we know very little about what constitutes a therapeutic relationship) was an attribute of patient en- therapeutic relationship in physiotherapy. Due to this gagement because, as a supportive partnership, it en- gap in physiotherapy literature, combined with the ad- courages patients to engage in rehabilitation. But who is vanced knowledge development in psychotherapy relative responsible for engaging the therapeutic relationship, to other healthcare disciplines (e.g., medicine) and the po- physiotherapist or patient? For instance, Higgins et al. tential benefits of adopting a psychologically-informed [3] define engagement as “.. . the desire and capability to perspective in rehabilitation [12], physiotherapy research actively choose to participate in care in a way uniquely and practice have been influenced by psychotherapy the- appropriate to the individual in cooperation with a ory [9, 10, 13]. For instance, physiotherapy researchers healthcare provider or institution for the purposes of have used Bordin’s theory of the working alliance [14], maximizing outcomes or experiences of care” (p. 33). while educators reference Freudian [15, 16] and Rogerian This implies a substantial degree of patient investment principles [15]. Of these theories, Rogers’ [17] “necessary along side the provider. and sufficient conditions” of genuineness (freedom to be one’s self), empathic understanding (understanding of the Given the importance placed on patient engagement patient’s feelings and meanings combined with congruent in rehabilitation, understanding the therapeutic relation- interactional behaviours) and unconditional positive re- ship in physiotherapy from patient and physiotherapist gard (accepting attitude) have contributed, implicitly or perspectives is needed. Although this view is supported explicitly, to the understanding of the therapeutic relation- in research of physiotherapy services [23], historically, ship in psychotherapy [18]. These guiding principles are patient involvement in research of the therapeutic rela- broad and arguably extend to human relationships in tionship has focused more on therapist perspectives. general [17, 19] and in a way that can be understood Moreover, patients’ experiences of the therapeutic rela- by practitioners and patients alike. Meta-analyses have tionship may have greater weight than therapists’ consid- demonstrated that empathy [20] and positive regard ering their ratings of therapeutic relationship quality can [21] are moderately associated with clinical outcomes be more predictive of successful psychotherapy interven- in psychotherapy. tions [24]. Therefore, patient contributions are essential for developing foundational knowledge of the thera- While Rogers’ conditions are broad and could apply to peutic relationship in physiotherapy. physiotherapy as well as other healthcare disciplines [17], there might also be aspects specific to physiother- Assuming meaningful engagement relies on a positive apy [22]. For instance, physiotherapists often use touch supportive relationship between patient and provider, during assessment and treatment, which is likely not the we posed the question: what conditions are necessary case in psychotherapy-oriented disciplines, such as for both physiotherapist and patient to engage in a psychology. It is also relevant to note that delivery of therapeutic relationship? Given the nature of the ques- physiotherapy services differs practically from other tion and the limited understanding of the therapeutic healthcare professions. For instance, physiotherapy treat- relationship in physiotherapy, we undertook a qualita- ment sessions can be longer in duration and occur on a tive investigation, using physiotherapist and patient more frequent basis during a particular treatment period perspectives, to identify and provide in-depth descrip- (e.g., number of sessions per week) compared to phys- tions of the conditions of engagement necessary for a ician visits. In addition, physiotherapists may be more therapeutic relationship between physiotherapist and patient.

Miciak et al. Archives of Physiotherapy (2018) 8:3 Page 3 of 12 Methods how much emphasis physiotherapists placed on the therapeutic relationship to build a caseload. Research team and reflexivity The research team consisted of 4 clinicians (2 physio- Participants therapists, 1 occupational therapist, and 1 psychologist), Physiotherapists were eligible if they had a minimum of and a qualitative methodologist from human ecology. 5 years of clinical experience and were currently working Two of the 5 researchers had significant experience in private practice. Adult (18–64 years of age) patients using qualitative methods in health research and a third, with musculoskeletal complaints were eligible if they re- the lead author, was completing this project as a compo- ceived at minimum 3 treatment sessions and were nent of her doctoral thesis and led all aspects of the within 12 weeks of their last session. Patients were ineli- study. In doing so, the lead author was informed by pre- gible if they had co-morbid conditions limiting their vious and extensive training in qualitative methods as cognitive capacity or ability to communicate, neuro- well as meta-theoretical perspectives from critical real- logical or systemic inflammatory conditions, or if they ism [25, 26] and psychotherapeutic contextual theory had received wage replacement or pain and/or suffering [27]. The lead author also applied experience gained as a compensation. contributor on other qualitative research studies. It is also relevant to note that the lead author had post- Sampling strategy and recruitment graduate training in psychotherapy, which informed Physiotherapist sampling strategy and recruitment prior clinical practice as a physiotherapist as well as her Purposive sampling was used to recruit 11 physiothera- interest in therapeutic relationship as a research topic. pists (6 female). Two authors (including the lead author) The therapeutic relationship was a central component of who are physiotherapists used their knowledge of the the clinical psychologist’s research program. private practice community to identify physiotherapists who could provide in-depth accounts of their thera- Design peutic relationship experiences. Administrative staff in Interpretive description was the qualitative methodo- the Department of Physical Therapy, University of Al- logical orientation [28, 29] used to address the research berta sent an email invitation to therapists, directing question [28]. Grounded in naturalistic inquiry [30], in- them to contact the first author with questions or if in- terpretive description is a framework that guides re- terested in participating. Upon contact, the lead author searchers to maintain a path toward pragmatic versus reviewed the study information sheet with all potential theoretical findings when addressing clinical or applied participants. Three therapists did not respond to the problems. Interpretive description does not prescribe the email and 1 declined to participate after speaking with use of a specific theoretical framework, as do traditional the lead author. Purposive sampling enabled sampling methods (e.g., grounded theory, phenomenology). When across factors such as treatment specializations (e.g., designing a study, Thorne suggests researchers consider manual therapists) and areas of interest (e.g., chronic various factors that could influence practice, including pain). Therapists’ ages ranged between 36 and 60 years the disciplinary mandate (e.g., physiotherapy’s social (mean age 47.8 years); demographic data were missing mandate to help others), current practice theories or for 2 therapists. All physiotherapists had been practicing models (e.g., patient-centred care), and the research in private practice for at least 10 years. The majority question. This practice-oriented scope is meant to en- (10/11) used at least one advanced restricted activity sure that “... at least one foot be firmly placed on the (i.e., activity requiring authorization from the regulatory solid ground that is the ‘real world’...”(p 201) of clinical body), such as acupuncture or spinal manipulation [32]. practice. For this reason, an inductive approach was Post-graduate training was reported in women’s health, taken, eliminating the use of a theoretical framework or vestibular rehabilitation, temporomandibular joint re- themes at the outset of the study, including psychother- habilitation, and sports physiotherapy. apy theories or approaches. Patient sampling strategy and recruitment Setting Purposive and convenience sampling were used to re- The setting was private practice physiotherapy clinics in cruit 7 patient participants (4 male). Ages ranged be- Edmonton, Canada. Reasons for situating the study in tween 18 and 62 years (mean age of 42.3 years). these clinics included: the notable percentage of physio- Administrative staff in 3 clinics purposively identified therapists working in these settings (48.2% in 2016) [31]; patients they believed would be able to provide candid their community location, which provided direct and accounts of the relationships with their therapists. Staff possibly greater access to physiotherapy services; and the provided patients with study information sheets and di- potential that a for-profit business model could influence rected them to contact the lead author with questions or

Miciak et al. Archives of Physiotherapy (2018) 8:3 Page 4 of 12 if interested in participating. Study information was also and critique of, for example, the interview setting as well distributed to a large athletic club via a coach’s email, as the interaction between researcher and participant, with instructions to contact the lead author. Upon con- including how the researcher’s perspectives on the thera- tact, the study information sheet was reviewed with all peutic relationship might have influenced the interview. potential participants. One patient was deemed ineligible Two mock interviews were completed, which informed for the study after speaking with the lead author. Most refinement of the interview guides prior to initiating par- patients (6/7) had previously accessed physiotherapy ticipant interviews. Concurrent data generation and ana- services and most (5/7) had experienced their physical lysis allowed for interview guide revisions to reflect the issues for greater than 3 months prior to seeking evolving analysis. The lead author completed all inter- treatment. views and data analysis. Data were generated until a point of saturation [36] was achieved representing a Data generation and analysis meaningful reflection of clinical reality. Data generation and analysis were inductive and itera- tive. After receiving informed consent, semi-structured Data analysis occurred in 2 concurrent phases: (1) a one-on-one interviews were completed in a public loca- systematic process of data (audio and transcript) review, tion of the participant’s choice, audio-recorded, and pro- reflexive journaling, and memoing prior to coding; and fessionally transcribed. One interview lasting between 40 (2) formal coding guided by qualitative content analysis and 90 min was completed with each participant, al- [37] and constant comparison principles [38]. To sup- though participants were informed they may be con- port an inductive process that would generate findings tacted to clarify their statements. An interview guide congruent with the physiotherapy context, psychother- [33] of open-ended questions was used to facilitate de- apy theory (e.g., Rogerian theory) was not used to guide scriptions of participants’ experiences of the therapeutic the analysis. Content analysis began with initial coding relationship. Although physiotherapists and patients had [38] or the assignment of a specific word or phrase to separate interview guides, they were similar in that both summarize a key attribute of a portion of text [39]. As began with broad questions to provoke responses on the patterns of codes were recognized [40, 41], they were clinical interaction in general (e.g., What do you call grouped into categories and sub-categories [42]. At this yourself – a patient or a client?) then became specific to point, constant comparison strategies were integrated to aspects of the interaction that physiotherapists and pa- refine the analysis and assist in the process of thinking tients believed could influence or were a part of the about the categories’ properties (i.e., characteristics of therapeutic relationship. However, questions in the inter- the category) and conditions (i.e., circumstances that view guides differed since physiotherapists form thera- foster the category) [36]. Negative cases [43] within par- peutic relationships with many patients whereas patients ticipant accounts contributed to clarifying aspects of the will not have this breadth of experience. For example, conditions of engagement. physiotherapists were asked about their views on ‘fixing patients’ in order to encourage responses regarding their The lead author completed all interviews and analysis treatment philosophies in general whereas patients could in partial fulfillment of her doctoral thesis. It is worth be asked to compare their therapeutic relationship with noting that the lead author had not met the patient par- their physiotherapist to the one with their physician. We ticipants prior to the study. However, given the lead au- have described the rationale for both patient and physio- thor had previously worked in private practice therapist interview guides elsewhere [34]. Probing ques- physiotherapy, she knew some of the physiotherapist tions (e.g., How did that make you feel? or What participants on a professional basis, to varying degrees, happened then?) or contact statements to check for clar- prior to the study commencing. Various rigour strategies ity (e.g., It sounds like your physiotherapist was con- that involved researcher, participants, and external re- cerned about your well-being?) were used to build on views were used throughout the study to address trans- participant responses in-the-moment to encourage thor- parency and trustworthiness of the research process and ough description and to disrupt the researcher’s pre- findings. Personal researcher strategies involved journal- conceived notions. Various rigour strategies, described ing to: maintain an audit trail [44, 45]; reflexively engage below, were used to critique the data generation process [46] throughout the research process; and memo ques- in order to continually inform interview quality. For ex- tions and ideas during the analysis [47]. Two patient ample, interview notes were a component of the lead au- participants engaged in member reflections [45] about thor’s intersubjective reflection on the “... situated, the ongoing analysis and 2 researchers and healthcare emergent, and negotiated nature of the research encoun- providers were involved in peer debrief [44, 47]. An ex- ter” ( [35] p8). Interview notes were completed after ternal audit [44] was completed at project completion, each interview to capture the researcher’s impressions confirming that the research process was thorough and the quality and nature of the findings were congruent with the process. NVivo 10 for Windows software (QSR

Miciak et al. Archives of Physiotherapy (2018) 8:3 Page 5 of 12 International Pty Ltd.) was used to manage the data and down” (PT-E). Patients also spoke of their need to be analysis. present during the interaction. Notably, they spoke of being in-the-moment to understand their bodies and Results “feel the treatment” (Patient-E) because “if I can’t tell Four foundational conditions fostering engagement be- her [PT] how it’s feeling or how it’s reacting, I can’t help tween physiotherapist (PT) and patient within a thera- her” (Patient-A). peutic relationship were identified and labeled: (a) Present, (b) Receptive, (b) Genuine, and (d) Committed. Receptive To be receptive, physiotherapists and patients must Present enter interactions with: a) an open attitude to negotiate Being present reflects physiotherapists’ and patients’ in- appropriate treatment plans; and b) a focused receptivity tentions and abilities to be in-the-moment or embodied to identify salient issues and needs. in time and space. Physiotherapists make conscious choices about the amount of time they spend in direct Open attitude proximity with patients in a potentially chaotic setting Having an open attitude requires physiotherapists and laden with competing responsibilities. Therapists de- patients to manage personal agendas and be willing to scribed instances when remaining with the patient was be- be “open to all these things [treatments]” (Patient-A). lieved to be of utmost importance, such as when a patient Even though therapists have specific knowledge and needed “more one-on-one time” (PT-J) for guidance with skills that inform treatment plans, they also need “... to exercises or when experiencing emotional distress: be open and listening and not go into this [interaction] with a pre-determined agenda”(PT-B). This includes a PT-B:... they start crying. .. the biggest thing. .. is don’t willingness to listen to the patient’s story because it is “... pull away. Don’t walk out of the room. Don’t leave them. important to me as the patient that you hear and under- stand what I need you to help me [with]” (Patient-E). While scheduling longer sessions (e.g., 30 min) was an Allowing patients to tell their stories can be important option, physiotherapists also described many impromptu for developing a safe and receptive atmosphere: situations where a decision was made to remain with a patient, despite the allotted timeframe: PT-I: The big thing is that patients that are struggling and... really have big problems, they need to tell their PT-I: I think that if I’m with somebody who’s gone story. You need to listen and shut your mouth. through 20 years of struggle with this, I think I have to take more time at the beginning. The same is true for patients. Just as therapists need to “... listen to all their [patients’] fears, all their issues...” Patients noticed their therapists’ efforts to “spend (PT-G), to create a working relationship, patients also more time with me than they should” (Patient-B). need to listen and be open to physiotherapists’ Patient-E appreciated that “time was of no consequence” suggestions: because it gave the impression that the therapist was willing to do “whatever it takes” to address the issue. Pa- PT-G: You try to explain what you are doing and they tients also noticed when therapists were not present and keep interrupting you. They keep challenging everything the negative impact this had on their experiences, such you say... They don’t listen to anything you say. That I as when they perceived therapists were rushing. More- find really difficult. over, patients were able to distinguish between a ‘busy’ therapist and a ‘rushed’ therapist, where a busy therapist Focused receptivity could be present despite the hectic environment: In addition to an open attitude, physiotherapists must also be attentive to the situation at hand. This is Patient-D: They were busy as can be, just on a cycle achieved by actively considering patients’ verbal and going from one to the next to the next and coming back. non-verbal cues. For example, focused receptivity helps They always took the time to make you feel like you therapists gain insight into patients’ physical and psy- were a decent person. chological states: In addition, physiotherapists and patients described PT-B: They are guarded, they are tightening. .. you can the importance of creating a “bubble” (PT-K) that allows just see that they are upset. full engagement. Although therapists could be distracted by multiple responsibilities, a busy caseload, and per- PT-A: If they are not talking to you. .. or if their tone sonal factors (e.g., family stressors), they took personal has raised or heightened then you know something is responsibility to “turn those issues off” (PT-G) when going on... with patients. Therapists also described using non-verbal cues and manipulating material space, such as adjusting In addition to focusing on behaviours, therapists also seating arrangements and using private rooms versus spoke of how being receptive to patients’ comments, curtained cubicles to help patient and therapist “narrow often noted either mentally or in the chart, was essential

Miciak et al. Archives of Physiotherapy (2018) 8:3 Page 6 of 12 for identifying how to connect with patients about their PT-B:... being realistic about what’s going to happen. .. lives. This enabled them to “... gauge where that person’s I’m really honest with people about that and I explain to at and what their interests are...” (PT-E). This receptivity them and especially with those more complex, that they fosters deeper engagement during the immediate inter- are 80% of what’s going to make a difference. action and provides opportunity for the same in the future. Patients must also be transparent about information related to their conditions, or as Patient-C claims, it “... Genuine is important for the patient to tell the whole truth...” To be genuine is to be real or convey sincerity in the Physiotherapists agreed they needed to trust that “... they present. Being genuine in a therapeutic relationship has [patients] are telling you the truth... all the factors that three aspects: a) being yourself; b) being honest; and c) are contributing.” (PT-E). investing in the personal. In addition to being transparent, the physiotherapist Being yourself must also be direct in the tone and manner of communi- To convey genuineness, individuals must remain con- cation. Specifically, therapists must be clear and forth- gruent with their personal qualities and values, while right. Although being direct might be interpreted as maintaining an accepting attitude. To do this, physio- stern, especially in challenging situations, the tone can therapists and patients must feel comfortable enough to also convey concern or compassion. Ultimately, the ther- sincerely present themselves, not putting on a facade: apist’s intention is to be clear, leaving little doubt about the message: PT-I: I’m pretty open with people. I can talk to any- body... I don’t change who I am in any role in my life. .. PT-H: She did have an injury but I had to explain to I am who I am. I think patients probably feel comfort- her that, “The injury that you have cannot cause all of able asking me that because that’s kind of how we inter- the problems that you are having. Let’s try to figure out act as people. what else is causing it.” Patients notice when physiotherapists are being them- Investing in the personal selves or have “warm”, “personable”, or “approachable” A primary focus of physiotherapy is to restore or main- personalities. In doing so, therapists create an environ- tain physical mobility and function. However, many pa- ment where patients can also express themselves. Thera- tients and physiotherapists revealed that a personal pists curb judgment of patients and are open to “where aspect was important to the overall quality of the thera- that individual is” (PT-E) by acknowledging their unique peutic relationship. Being invested in the personal was personalities, life stories, and social and cultural realities. revealed through an interest in the person and a willing- In addition, freedom for patients to be themselves ex- ness to disclose about self. tends to their bodies and injuries. Physiotherapists can mitigate patients’ feelings of vulnerability that give rise Taking an interest in the person pertains to therapists’ to negative perceptions of their bodies and injuries: or patients’ desires to broaden the scope of caring to an interest in the other’s life beyond the reason for referral: Patient-D:... [he] was very good at making me feel like you weren’t abnormal... I don’t want to be singled out as PT-C:... folks that ask me how I’m doing, folks that ask out of shape or old or... I didn’t quite know what to ex- me how things are going, we end up talking about things pect when the physiotherapist came in... I expected a fair unrelated to their condition or the weather... We have bit of judgmenty-type things the way that doctors would an interest in each other. sometimes. Even though therapists often need to know about pa- Being honest tients’ lives for therapeutic reasons, those invested in the While honesty is likely a necessary condition for any personal are willing to get to know the patient as a per- healthy relationship, there are two main qualities that son, demonstrating an authentic interest in people’s describe being honest in the physiotherapy context: lives. This investment can put the patient at ease: transparency and directness. Being transparent involves therapists and patients providing the necessary informa- PT-I: Even when my questioning starts, you know I al- tion to help the patient progress in a safe and meaning- ways ask them about them first. So, I always make it ful way. This can include impressions of the physical clear that that’s really important to me... I ask them to problem and rehabilitation process; personal limitations tell me a little bit about yourself outside of what’s in skill and knowledge; patient participation and out- brought you here... What sorts of things do you enjoy come expectations; and the therapist’s role and doing? Even the way I ask those questions is very differ- responsibilities: ent. ... I can get to a person’s level of comfort and they can relax a little bit if I ask them questions that are not directed to their sore knee or sore shoulder. . . Even though roles and professional boundaries might make it difficult for patients to express an interest in

Miciak et al. Archives of Physiotherapy (2018) 8:3 Page 7 of 12 their physiotherapists’ lives, they could be “genuinely in- to stay outside of those kinds of conversations”, nor was terested” in getting to know their therapists, asking “... he interested in discussing anything outside of the clin- almost as many questions as you ask them” (PT-J). Fur- ical problem: thermore, some patients found value in knowing their physiotherapists on a human level: PT-D: I really don’t talk much on the personal side. I really don’t think any of my patients even know how Patient-B: It makes a huge difference knowing that many kids I have or what I do in my spare time. I don’t they can relate to you, first of all and they have a real life. think any one of them knows that... that’s purely on the They are not just a physio... these people go home and have personal side. kids and have a family. It’s nice. You are both real people so you should probably treat each other like people. Therefore, it appears there are different ways to be in a therapeutic relationship: Another aspect of investing in the personal is demon- strating a willingness to disclose. Being willing to disclose PT-K: My partner is exactly the opposite of me. .. my means offering something more personal and not neces- professional boundaries and his professional boundaries sarily related to the primary intent of the interaction. are on either side of the continuum of professional Therefore, disclosures can be social or therapeutic. Most boundaries. therapists recalled they had different perceptions of what constituted an appropriate disclosure: Committed To be engaged, physiotherapists and patients must be PT-F: ... you can talk about personal interests and not committed to their roles within the therapeutic relation- get personal so hobbies and what you might do in your ship. A patient’s well-being matters, or, as PT-A claimed, non-professional life that doesn’t have to do with any- “their well-being is your well-being...” This speaks to an thing intimate... sports are good, music is good, leisure ethic of care that encompasses physiotherapists’ profes- activities. . . sional duty and the desire to be of service to others to restore patients’ well-being. Some physiotherapists and Patients’ investments in the personal also included dis- patients stated that therapists do not “fix” patients, but closing more personal aspects of their physical and emo- that both have roles they must commit to: tional challenges, including issues pertaining to sexuality or mood. Although one therapist commented that there Patient-B: You have to take care of yourself in order are some patients who “... are comfortable disclosing for them [physiotherapists] to be able to take care of you that information to you” (PT-A), this same therapist also too. If you are just going to go and expect them to do it claimed that patient disclosures sometimes required a all for you, it’s not going to happen. You’re not going to “leap of faith” in the therapist. Patients agreed, com- get better, I find. menting that disclosure of their physical issues and per- sonal lives was easier as “... you get more comfortable so These points considered, there are two aspects that you’re more willing to tell them what you are feeling” characterize being committed: (a) committed to under- (Patient-C). standing and (b) committed to action. There is a spectrum of how much physiotherapists Committed to understanding and patients are willing to invest in the personal (see Both physiotherapists and patients must be motivated to Fig. 1). For example, PT-D was very clear he was not in- understand the patient’s situation. When the physiother- terested in his patients’ personal lives, making “... a point apist is committed to understanding the patient, there is Fig. 1 Spectrum of Personal Engagement. The figure illustrates personal engagement as a spectrum involving a relationship between the nature of engagement and the degree of personal engagement. The degree of personal engagement is dependent on the intentions and behaviours of physiotherapist and patient

Miciak et al. Archives of Physiotherapy (2018) 8:3 Page 8 of 12 a “... need to understand more about what you [patient] “Oh hey (name of patient). What do we need to work on are describing...”(PT-B). Therapists were not satisfied today?” He already knows how much I exercise and with a generic overview of the patient’s situation: everything. PT-D:. .. if you give out the impression that you know The Conditions of Engagement Form a Safe Therapeutic what’s happening in this person’s back without showing Container them the interest or without making an effort in under- The conditions of engagement work in concert to form standing it, you won’t be able to help them. a safe therapeutic container for the therapeutic relation- ship to manifest (Fig. 2). The foundational components The physiotherapist is not only dedicated to under- of the container – the bottom and the walls – are repre- standing the patient’s physical situation, but also “a pic- sented by the cornerstone conditions being present and ture of the unspoken” (PT-C) or the psychosocial factors being receptive. Being present is the foundation that al- that could be influential: lows the other conditions to unfold, while being receptive provides the structure that enables pertinent information PT-H: If a person has what we would call a chip on to be gathered. There is more of a personal aspect to be- their shoulder let’s say, you try to find out what the chip ing genuine and being committed; the degree to which ei- is. I see it as part of my job to get over that chip... If I ther condition is established is reliant upon individuals’ can find out what brought it on... Empathize. Sort of uniqueness and circumstances. Essentially, the condi- understand. tions of engagement set the tone for “being” with other and self, representing the dynamic intent to en- Even though the physiotherapist is expected to try to gage that both physiotherapist and patient bring to understand, it was also clear that patients needed to in- the relationship. vest in understanding their situations: Discussion Patient-E: I felt I needed to understand as much of my We found there are necessary conditions of engagement own physiology and biology in order to help what it is that facilitate the therapeutic relationship in physiother- that she was trying to do for me, so I could help myself. apy. In addition to providing needed clarity specific to physiotherapy, these conditions offer insight into the Committed to action relevance of psychotherapeutic principles in physiother- Being committed to action involves making “all efforts” apy and how they can best serve practice and research. (PT-D) to honour the best interests of the patient. Phys- iotherapists “... do their best to do the best that they Fig. 2 The Safe Therapeutic Container Formed by the Conditions can...” (Patient-C), and will go beyond due diligence to of Engagement. The foundation and the walls of the therapeutic help patients achieve goals. Therapists committed to ac- container represent the two cornerstone conditions, “present” tion recognize there are many facets of care to be con- and “receptive”, respectively. “Committed” and “genuine” are sidered, and that they may need to “go that extra little more variable and are therefore represented by the mobile mile” (PT-A) in complicated situations. nature of the lids of the container Patients must also be committed to act in their own best interests. Physiotherapists spoke about the necessity of patient “buy in” or as PT-G stated, “... they also have to agree with what you are saying and be motivated to take part in the treatment themselves because it's not just passive.” Patients seemed to understand that their motivation to participate was essential: Patient-G:... they [patients] are expecting the physio- therapist to “fix them” and they don’t need to fix them- selves... I understand what physio means and how I need to aid myself as well. Patients highlighted that continuity, described as the patient seeing the same therapist versus being shuttled between therapists, is an important part of being com- mitted. Having “your therapist” (Patient-B) facilitates progression of the session, reduces the need for the pa- tient to familiarize a new therapist, and allows the physiotherapist to get to know the patient’s body, activity levels, and treatment history: Patient-G: “What’s your past injuries? How many in- juries have you had? What's your sport history?” All that stuff. When I saw (name of physiotherapist), it was like,

Miciak et al. Archives of Physiotherapy (2018) 8:3 Page 9 of 12 The Relevance of Psychotherapeutic Principles in must contribute. We agree that the practitioner is re- sponsible for establishing the conditions that provide a Physiotherapy safe space for the patient to engage. Indeed, the physio- therapist’s capacity to do so could be the deciding factor Physiotherapy has previously borrowed theory from psy- in some patients’ willingness to engage. However, our chotherapy to inform research and practice. In fact, participants were clear that engagement involves the de- physiotherapy can be seen to clearly align with the ele- liberate participation of both patient and therapist for ments of a psychotherapeutic process that involves an the conditions to flourish. This is consistent with Bright individual seeking healing (e.g., patient) and a healing et al.’s [4] concept analysis of engagement in rehabilita- agent (e.g., physiotherapist) investing in a relationship in tion, which concluded that both clinicians and patients order to relieve disability and suffering while addressing have roles in patient engagement. Lequerica et al. [54] the individual’s beliefs and attitudes [27, 48]. This is a also found that therapists’ ability to facilitate patient en- compelling perspective as researchers and clinicians turn gagement was supported by “... taking time to simply their attention to psychological aspects rehabilitation, talk to the patient about their life...” (p. 757), indicating such as patient expectations, beliefs, and emotions that engagement is two-way and that both therapist and alongside addressing physical impairment and function patient engagement can be essential in developing the [12, 48]. Given our results appear to have similarities conditions. with Rogers’ necessary conditions of patient-centred care [49], and that these principles inform motivational inter- The Impact of Conditions of Engagement on viewing [50], an intervention increasingly being used in healthcare settings with positive results [50–53], we feel Physiotherapy Research and Practice there is an opportunity to consider their relevance in Psychotherapeutic theories such as Rogers’ conditions physiotherapy. This is also important considering the [17] and Bordin’s working alliance [14] are claimed to be pragmatic differences between physiotherapy and psy- universal. Moreover, physiotherapy researchers tend to chotherapy, including the conditions (e.g., mental illness assume that these theories directly transfer to physio- versus physical conditions) [12] and the subsequent therapy. By clarifying physiotherapy-specific conditions treatment goals. of engagement, our findings clearly have the potential to impact physiotherapy research and practice. Aspects of Rogers’ genuineness, unconditional positive regard, and empathic understanding weave into our con- Regarding research, we need to consider whether ditions of engagement. For instance, Rogers describes measurement scales developed through a psychothera- genuineness as being the expression of an integrated self peutic lens are valid within the physiotherapy context. through self-awareness and transparency [18], which is This view is congruent with Besley et al.’s conceptual congruent with our description of genuine as physiother- [22] and evaluative [55] findings. In particular, the evalu- apists honouring their personal psychosocial situations, ative findings clarify that while the measurement proper- disclosing personal information, and being direct with ties for the Working Alliance Inventory [56] and patients. Helping Alliance Questionnaire [57] were “adequate” [55], there were also aspects missing. The authors called While our findings complement Rogers’ conditions, for a better conceptual understanding within the physio- clear nuances are also present. Specifically, we added be- therapy context in order to develop more rigorous meas- ing present and receptive. Rogers [17] limits his discus- urement tools. sion of being present to the basic level of patient and therapist being “in contact” (p. 90), and to some degree Regarding practice, the conditions of engagement influencing the experience of the other. However, we de- speak to the essence of what is required to have a meaning- fine being present as a foundational condition and ful therapeutic relationship. Much literature has focused on clearly describe the focused manner and intentional use the importance of communication in developing the thera- of time and space in creating a safe therapeutic environ- peutic relationship [58]. However, relationships are more ment. In addition, we explicitly identify receptive as a than a compilation of skills and behaviours that can be condition. Although some might interpret being recep- dutifully checked off when completed. Relationships are dy- tive as an aspect of Rogers’ empathic understanding, we namic, requiring intent to ensure behaviours and skills are understand it to have distinguishing characteristics, congruent with the situation. Additionally, it is important namely that a therapist can be receptive but not be to note that a personal aspect can be important for physio- empathic. therapy therapeutic relationship. Even though participants described a spectrum of perspectives and practices regard- One difference between Rogers’ and our conditions re- ing the nature and boundaries of the personal, the majority lates to who is responsible for developing the conditions agreed that a personal aspect, understood as patients’ and of engagement. Rogers describes the psychotherapist as therapists’ authentic interest in the other’s life outside of cultivating the conditions of engagement whereas our findings indicate both physiotherapists and patients

Miciak et al. Archives of Physiotherapy (2018) 8:3 Page 10 of 12 the rehabilitation context or the disclosure of information study. Third, these findings would likely be most ap- perceived as private, was important. This study illustrates plicable for therapists in private practice physiotherapy. that patients and therapists may want to know one another Future research in other settings (e.g., hospitals, rehabilita- as people while respecting professional boundaries. More- tion centres) and systems (e.g., workers compensation) over, the conditions provide the foundation for a patient- will contribute to understanding the conditions that influ- centred approach to be operationalized in clinical practice. ence patients’ and physiotherapists’ abilities to engage in Being receptive, committed, and genuine create the safe the therapeutic relationship. However, because the condi- therapeutic space necessary for a patient-centred exchange tions of engagement are conceptual in nature, they could that highlights collaboration in order to establish meaning- be useful across a wide range of physiotherapy contexts ful patient-driven goals [1]. and health care professions as a foundational starting point regardless of practice area. As alluded to above, it is worth noting that the condi- tions supporting the therapeutic relationship do not ‘just Conclusions happen’ by completing a list of behaviours. The condi- Participants in this study have made it clear that thera- tions in this context exist, at least in part, as a function peutic relationships do not ‘just happen’. Through par- of physiotherapist and patient states or the quality of ticipants’ candid accounts we have highlighted that consciousness experienced by an individual in any given conditions specific to the physiotherapy encounter cre- moment, whereas a condition of engagement can be de- ate a safe environment and facilitate mutual engagement scribed as the sentiment or circumstances between two of therapist and patient. Cultivating these conditions, in individuals. An individual’s state is informed by a com- conjunction with applying communication skills (e.g., ac- plex merging of momentary thoughts, feelings, and sen- tive listening), will result in situation-appropriate re- sations in addition to more enduring attitudes, values, sponses. Findings suggest that theories developed in and beliefs, which will inform that individual’s intentions other disciplinary contexts (e.g., psychotherapy) should and ability to behave in ways that carry out those inten- be used judiciously when developing theory that guides tions. Moreover, behaviours that are genuine and con- physiotherapy practice and research regarding the thera- gruent with a situation arise from appropriate states. peutic relationship. Therefore, if physiotherapists are aware of and able to critique their thoughts, emotions, attitudes and assump- Abbreviation tions [59] and adjust as needed, conditions can be devel- PT: physiotherapist oped, maintained, or deepened. This reflection can occur outside of the clinical interaction, which is more Acknowledgements likely with novice physiotherapists or within the clinical The lead author would like to thank Christine Daum, PhD and Geoff Bostick, interaction, otherwise known as ‘reflection-in-action’ PhD for participating in peer debrief sessions and Mark Hall, PhD for [59]. Reflective practice targeting therapeutic relation- completing the external audit. ships is critical for encouraging physiotherapists’ abilities to cultivate the conditions of engagement. Without it, Funding physiotherapists risk self-limiting their ability to influ- This research was supported by a Thesis Research Operating Grant from the ence what is considered a key contextual factor [48] Department of Physical Therapy, University of Alberta. The funder did not impacting clinical outcomes [8–10]. have any role in designing the study or collecting, analysing, or interpreting the data. Limitations of the study There are three main limitations in this study. First, both Availability of data and materials patient and physiotherapist accounts often centred on The datasets generated and/or analysed during the current study are not therapists’ contributions to the therapeutic relationship publicly available due to ethical considerations. Participants were not and conditions of engagement. Although this might be informed that their data (i.e., verbatim interview transcriptions) would be expected given the therapist’s role and position of power made public, and therefore, did not provide informed consent. within the clinical interaction, a second interview with patients would have provided opportunity to probe them Authors’ contributions about their role in establishing the conditions of engage- MM (lead author) made substantive contributions to the conceptual and ment. Second, the exclusion of some patients limited the design development, collected and analysed all data, and was the lead nature of the data and hence, the possible breadth of the writer. MM (second author), CB, ASJ, and DPG all contributed to conceptual findings. Other populations likely have additional con- and design development and writing. All authors read and approved the siderations (e.g., long term therapy and family involve- final manuscript. ment for patients with neurological conditions) that require focused investigation beyond the scope of this Author’s information Maxi Miciak worked as a physiotherapist for 13 years in private practice and interdisciplinary rehabilitation settings, with a special interest in providing patient-centred care for people managing chronic conditions. Maxi is currently a post-doctoral fellow in research impact assessment. Her research focuses on operationalizing patient-centred care by developing, implementing, and evaluating care models that impact the patient-practitioner therapeutic relationship, including how health services and policies support this relationship.

Miciak et al. Archives of Physiotherapy (2018) 8:3 Page 11 of 12 Ethics approval and consent to participate 12. Main CJ, George SZ. Psychologically informed practice for management of The University of Alberta Health Research Ethics Board approved the study. low back pain: future directions in practice and research. Phys Ther. 2011;91: All participants provided informed consented prior to participating. 820–4. https://doi.org/10.2522/ptj.20110060. Consent for publication 13. Cheing GL, Lai AK, Vong SK, Factorial CFH. Structure of the pain Not applicable. rehabilitation expectations scale: a preliminary study. Int J Rehabil Res. 2010; 33:88–94. https://doi.org/10.1097/MRR.0b013e32832e9884. Competing interests MM (lead author) was supported during her doctoral studies by the 14. Bordin ES. The generalizability of the psychoanalytic concept of the working University of Alberta, the Canadian Institutes of Health Research, and the alliance. Psychotherapy (Chic). 1979;16:252–60. https://doi.org/10.1037/ College of Physical Therapists of Alberta. She is currently the Cy Frank Fellow h0085885. at Alberta Innovates, Edmonton, Canada. MM (lead author) has received honorariums as an invited speaker for presentations related to the content of 15. Davis CM. Patient practitioner interaction: an experiential manual for this manuscript. MM (second author), CB, ASJ, and DPG declare they have no developing the art of health care. Thorofare: SLACK; 2006. competing interests. 16. Purtilo RB, Haddad AM. Health professional and patient interaction. 7th ed. Publisher’s Note St. Louis: Saunders Elsevier; 2007. Springer Nature remains neutral with regard to jurisdictional claims in 17. Rogers CR. The necessary and sufficient conditions of therapeutic personality published maps and institutional affiliations. change. J Consult Psychol. 1957;21:95–103. https://doi.org/10.1037/h0045357. Author details 18. Truscott D. Becoming an effective psychotherapist: adopting a theory of 1Alberta Innovates, 1500, 10104 – 103 Avenue NW, Edmonton, AB T5J 0H8, psychotherapy that’s right for you and your client. Washington: American Canada. 2Faculty of Extension, University of Alberta, 10230 – Jasper Ave, Psychological Association; 2010. Edmonton, AB T5J 4P6, Canada. 3Department of Occupational Therapy, University of Alberta, 8205 114 Street, 2-64 Corbett Hall, Edmonton, AB T6G 19. Cornelius-White JHD, Motschnig-Pitrik R, Lux M. Interdisciplinary 2G4, Canada. 4Department of Psychiatry, University of Alberta, 1E1 Walter research and theory. In: Cornelius-White JHD, Motschnig-Pitrik R, Lux M, Mackenzie Health Sciences Centre, 8440 112 St NW, Edmonton, AB T6G 2B7, editors. Interdisciplinary handbook of the person-centered approach: Canada. 5Department of Physical Therapy, University of Alberta, 8205 114 research and theory. New York: Springer; 2013. p. 23–34. https://doi.org/ Street, 2-50 Corbett Hall, Edmonton, AB T6G 2G4, Canada. 10.1007/978-1-4614-7141-7. Received: 15 September 2017 Accepted: 26 January 2018 20. Elliott R, Bohart AC, Watson JC, Greenberg LS. Empathy. Psychother (Chic). 2011;48:43–9. https://doi.org/10.1037/a0022187. References 1. Mead N, Bower P. Patient-centredness: a conceptual framework and review 21. Farber BA, Doolin EM. Positive regard. Psychotherapy (Chic). 2011;48:58–64. https://doi.org/10.1037/a0022141. of the empirical literature. Soc Sci Med. 2000;51:1087–110. https://doi.org/ 10.1016/S0277-9536(00)00098-8. 22. Besley J, Kayes NM, KM MP. Assessing therapeutic relationships in 2. Macleod R, Mcpherson KM. Care and compassion: part of person-centred physiotherapy: literature review. N Z J Physiother. 2011;39:81–91. https:// rehabilitation, inappropriate response or a forgotten art? Disabil Rehabil. www.thefreelibrary.com/Assessing+therapeutic+relationships+in 2007;29:1589–95. https://doi.org/10.1080/09638280701618729. +physiotherapy%3a+literature...-a0288537990. Accessed 06 Jan 2018 3. Higgins T, Larson E, Schnall R. Unraveling the meaning of patient engagement: a concept analysis. Patient Educ Couns. 2017;100:30–6. https:// 23. Hills R, Kitchen S. Satisfaction with outpatient physiotherapy: focus groups doi.org/10.1016/j.pec.2016.09.002. to explore the views of patients with acute and chronic musculoskeletal 4. Bright FAS, Kayes NM, Worrall L, McPherson KM. A conceptual review of conditions. Physiother Theor Pract. 2007;23:1–20. https://doi.org/10.1080/ engagement in healthcare and rehabilitation. Disabil Rehabil. 2015;37:643– 09593980601023705. 54. https://doi.org/10.3109/09638288.2014.933899. 5. Physiotherapy Alberta - College + Association. Standards of Practice for 24. Horvath AO, Bedi RP. The alliance. In: Norcross JC, editor. Psychotherapy Physiotherapists in Alberta. 2017. Physiotherapy Alberta - College + relationships that work: therapist contributions and responsiveness to Association.. https://www.physiotherapyalberta.ca/files/standards_of_ patients. New York: Oxford University Press; 2002. p. 37–70. practice.pdf. Accessed 01 Sept 2017. 6. Hush JM, Cameron K, Mackey M. Patient satisfaction with musculoskeletal 25. Clark AM, Lissel SL, Davis C. Complex critical realism: tenets and application physical therapy care: a systematic review. Phys Ther. 2011;91:25–36. https:// in nursing research. ANS Adv Nurs Sci. 2008;31:E67–79. https://doi.org/10. doi.org/10.2522/ptj.20100061. 1097/01.ANS.0000341421.34457.2a. 7. Schönberger M, Humle F, Zeeman P, Teasdale TW. Working alliance and patient compliance in brain injury rehabilitation and their relation to 26. Pawson R, Tilley N. Realistic evaluation. London: Sage; 1997. psychosocial outcome. Neuropsychol Rehabil. 2006;16:298–314. https://doi. 27. Frank JD, Frank J. Persuasion and healing: a comparative study of org/10.1080/09602010500176476. 8. Fuentes J, Armijo-Olivo S, Funabashi M, Miciak M, Dick B, Warren S, et al. psychotherapy. Baltimore: Johns Hopkins University Press; 1991. Enhanced therapeutic alliance modulates pain intensity and muscle pain 28. Thorne S. Interpretive description. Walnut Creek: Left Coast Press; 2008. sensitivity in patients with chronic low back pain: an experimental 29. Thorne S, Oliffe J, Kimsing C, Hislop TG, Stajduhar K, Harris SR, et al. Helpful controlled study. Phys Ther. 2014;94:477–89. https://doi.org/10.2522/ptj. 20130118. communications during the diagnostic period: an interpretive description of 9. Hall AM, Ferreira PH, Maher CG, Latimer J, Ferreira ML. The influence of the patient preferences. Eur J Canc Care (Engl). 2010;19:746–54. https://doi.org/ therapist-patient relationship on treatment outcome in physical 10.1111/j.1365-2354.2009.01125.x. rehabilitation: a systematic review. Phys Ther. 2010;90:1099–110. https://doi. 30. Lincoln YS, Guba EG. Naturalistic inquiry. Beverly Hills: Sage Publications; 1985. org/10.2522/ptj.20090245. 31. Physiotherapy Alberta - College + Association 2016 Annual Report. 10. Ferreira PH, Ferreira ML, Maher CG, Refshauge KM, Latimer J, Adams RD. The Physiotherapy Alberta - College + Association. 2017. https://www. therapeutic alliance between clinicians and patients predicts outcome in chronic physiotherapyalberta.ca/files/annual_report_2016.pdf. Accessed 06 Jan 2018. low back pain. Phys Ther. 2013;93:470–8. https://doi.org/10.2522/ptj.20120137. 32. Physiotherapy Alberta College + Association. Advanced activities. Physiotherapy 11. Stenmar L, Nordholm LA. Swedish physical therapists’ beliefs on what makes Alberta - College + Association. http://www.physiotherapyalberta.ca/ therapy work. Phys Ther. 1994;74:1034–9. https://doi.org/10.1093/ptj/74.11.1034. physiotherapists/what_you_need_to_know_to_practice_in_alberta/restricted_ activities/advanced_activities. Accessed 08 Sept 2017. 33. Patton MQ. Qualitative research & evaluation methods. Thousand Oaks: Sage; 2002. 34. Miciak M, Mayan M, Brown C, Joyce AS, Gross DP. A framework for establishing connections in physiotherapy practice. Physiother Theory Pract. In press. 35. Finlay L. The reflexivity journey: mapping multiple routes. In: Finlay L, Gough B, editors. Reflexivity: a practical guide for researchers in health and social sciences. Oxford: Blackwell Science; 2003. p. 3–20. 36. Glaser BG, Strauss AL. The discovery of grounded theory: strategies for qualitative research. New York: Aldine de Gruyter; 1967. 37. Hsieh HF, Shannon SE. Three approaches to qualitative content analysis. Qual Health Res. 2005;15:1277–88. https://doi.org/10.1177/ 1049732305276687. 38. Charmaz K. Constructing grounded theory. 2nd ed. Thousand Oaks: Sage; 2014.

Miciak et al. Archives of Physiotherapy (2018) 8:3 Page 12 of 12 39. Saldana J. An introduction to codes and coding: the coding manual for Submit your next manuscript to BioMed Central qualitative researchers. Thousand Oaks: Sage; 2009. and we will help you at every step: 40. Braun V, Clarke V. Using thematic analysis in psychology. Qual Res Psychol. • We accept pre-submission inquiries 2006;3:77–101. https://doi.org/10.1191/1478088706qp063oa. • Our selector tool helps you to find the most relevant journal • We provide round the clock customer support 41. Joffe H, Yardley L. Content and thematic analysis. In: Marks DF, Yardley L, • Convenient online submission editors. Research methods for clinical and health psychology. London: Sage; • Thorough peer review 2004. p. 58–69. http://srmo.sagepub.com/view/research-methods-for-clinical- • Inclusion in PubMed and all major indexing services and-health-psychology/n4.xml. • Maximum visibility for your research 42. Richards L, Morse JM. Readme first for a user’s guide to qualitative methods. Submit your manuscript at Thousand Oaks: Sage; 2007. www.biomedcentral.com/submit 43. McPherson G, Thorne S. Exploiting exceptions to enhance interpretive qualitative health research: insights from a study of cancer communication. Int J Qual Methods. 2006;5:1–11. https://doi.org/10.1177/ 160940690600500210. 44. Creswell JW. Research design: qualitative, quantitative, and mixed methods approaches. Thousand Oaks: Sage Publications; 2009. 45. Tracy SJ. Qualitative quality: eight “big-tent” criteria for excellent qualitative research. Qual Inq. 2010;16:837–51. https://doi.org/10.1177/ 1077800410383121. 46. Finlay L. “Outing” the researcher: the provenance, process, and practice of reflexivity. Qual Health Res. 2002;12:531–45. https://doi.org/10.1177/ 104973202129120052. 47. Mayan M. Essentials of qualitative inquiry. Walnut Creek: Left Coast Press; 2009. 48. Miciak M, Gross DP, Joyce A. A review of the psychotherapeutic ‘common factors’ model and its application in physical therapy: the need to consider general effects in physical therapy practice. Scand J Caring Sci. 2012;26:394–403. https:// doi.org/10.1111/j.1471-6712.2011.00923.x. 49. Rogers CR. Foundations of the person-centered approach. Education. 1979; 100:98–107. 50. Shannon R, Hillsdon M. Motivational interviewing in musculoskeletal care. Musculoskeletal Care. 2007;5:206–15. doi: org/https://doi.org/10. 1002/msc.119. 51. Lundahl B, Moleni T, Burke BL, Butters R, Tollefson D, Butler C, Rollnick S. Motivational interviewing in medical care settings: a systematic review and meta-analysis of randomized controlled trials. Patient Educ Couns. 2013;93: 157–168. doi: rg/https://doi.org/10.1016/j.pec.2013.07.012. 52. Park J, Esmail S, Rayani F, Norris CM, Gross DP. Motivational interviewing for workers with disabling musculoskeletal disorders: results of a cluster randomized control trial. J Occup Rehabil. 2017; https://doi.org/10.1007/ s10926-017-9712-3. 53. Gross DP, Park J, Rayani F, Norris CM, Esmail S. Motivational interviewing improves sustainable return to work in injured workers after rehabilitation: a cluster randomized controlled trial. Arch Phys Med Rehabil. 2017;98:2355– 2363.doi: org/https://doi.org/10.1016/j.apmr.2017.06.003. 54. Lequerica AH, Donnell CS, Tate DG. Patient engagement in rehabilitation therapy: physical and occupational therapist impressions. Disabil Rehabil. 2009;31:753–60. https://doi.org/10.1080/09638280802309095. 55. Besley J, Kayes NM, KM MP. Assessing the measurement properties of two commonly used measures of therapeutic relationship in physiotherapy. N Z J Physiother. 2011;39:75–80. http://www.thefreelibrary.com/Assessing+the +measurement+properties+of+two+commonly+used+measures+of...- a0288537989. Accessed 06 Jan 2018 56. Horvath AO, Greenberg LS. Development and validation of the working alliance inventory. J Couns Psychol. 1989;36:223–33. https://doi.org/10.1080/ 10503300500352500. 57. Luborsky L, Barber JP, Siqueland L, Johnson S, Najavits LM, Frank A, et al. The revised helping alliance questionnaire (HAq-II): psychometric properties. J Psychother Pract Res. 1996;5:260–71. https://www.med.upenn.edu/cpr/ assets/user-content/documents/Luborskyetal1996_HAQ-II.pdf. Accessed 06 Jan 2018 58. Pinto RZ, Ferreira ML, Oliveira VC, Franco MR, Adams R, Maher CG, et al. Patient-centred communication is associated with positive therapeutic alliance: a systematic review. J Physiother. 2012;58:77–87. https://doi.org/10. 1016/s1836-9553(12)70087-5. 59. Paterson C, Chapman J. Enhancing skills of critical reflection to evidence learning in professional practice. Phys Ther Sport. 2013;14:133–8. https://doi. org/10.1016/j.ptsp.2013.03.004.

Corbetta et al. Archives of Physiotherapy (2018) 8:4 https://doi.org/10.1186/s40945-018-0047-y CASE REPORT Open Access Mirror therapy for an adult with central post-stroke pain: a case report Davide Corbetta1*, Elisabetta Sarasso1,2, Federica Agosta2, Massimo Filippi3 and Roberto Gatti4 Abstract Background : Treatment of central post-stroke pain (CPSP) after a thalamic-capsular stroke is generally based on pharmacological approach as it is low responsive to physiotherapy. In this case report, the use of mirror therapy (MT) for the reduction of CPSP in a subject after a stroke involving thalamus is presented. Case presentation: Five years after a right lenticular-capsular thalamic stroke, despite a good recovery of voluntary movement that guaranteed independence in daily life activities, a 50-year-old woman presented with mild weakness and spasticity, an important sensory loss and a burning pain in the left upper limb. MT for reducing arm pain was administered in 45-min sessions, five days a week, for two consecutive weeks. MT consisted in performing symmetrical movements of both forearms and hands while watching the image of the sound limb reflected by a parasagittal mirror superimposed to the affected limb. Pain severity was assessed using visual analogue scale (VAS) before and after the intervention and at one-year follow-up. After the two weeks of MT, the patient demonstrated 4.5 points reduction in VAS pain score of the hand at rest and 3.9 points during a maximal squeeze left hand contraction. At one-year follow-up, pain reduction was maintained and also extended to the shoulder. Conclusion: This case report shows the successful application of a motor training with a sensory confounding condition (MT) in reducing CPSP in a patient with a chronic thalamic stroke. Keywords: Stroke, Physical therapy modalities, Pain perception, Case reports Background pain (CPSP) syndrome [7, 8]. The estimated incidence of Stroke often causes impairment in movement control CPSP comes up to 1 every 6 patients presenting a vascu- but can also affect perception [1, 2]. Alterations of lar lesion in the thalamus [8, 9], but its prevalence is stimulus integration are common after a stroke, with difficult to estimate because of the co-occurrence of variable reported prevalence ranging from 11 to 85% [3], other painful conditions, such as spasticity or shoulder and sometimes these alterations of perception result in pain [4]. The pathophysiological mechanisms underlying pain. Pain relates with the site of lesion and it is com- the development of CPSP are thought to be related to pletely distinct from other painful conditions such as the hyperexcitability or to the spontaneous discharge of shoulder pain or spasticity [4]. It typically emerges from damaged neurons located in the thalamus or in the cor- hemispheric lesions that involve the spinothalamic and tex [10]. The CPSP syndrome is one of the less respon- thalamocortical pathways, leading patients to complain sive conditions to physiotherapy treatment and it usually of sharping, stabbing, or burning through an experience requires a pharmacological approach through the use of of hyperpathia and allodynia [5, 6]. This association Amitriptyline, Gabapentin and Pregabalin [2]. between sensory abnormalities and constant or intermit- tent central neuropathic pain, arising from damage of Mirror therapy (MT), defined as the use of a mirror the sensory tracts, is known as the central post-stroke reflection of unaffected limb movements superimposed on the affected extremity, is often used to treat motor * Correspondence: [email protected] and perception problems [11, 12]. This technique was 1Laboratory of Analysis and Rehabilitation of Motor Function, San Raffaele described for the first time in 1995 in studies reporting Scientific Institute, Vita-Salute San Raffaele University, Via Olgettina 60, 20132 the reduction of phantom limb pain in arm amputees Milan, Italy [13]; more recently, its use was described also for recov- Full list of author information is available at the end of the article ery of motor function after stroke [14, 15], for the © The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.

Corbetta et al. Archives of Physiotherapy (2018) 8:4 Page 2 of 6 treatment of complex regional pain syndrome type I [12] was autonomous in everyday life activities (Functional and other painful conditions (e.g., brachial plexus avul- Independence Measure = 120) [18], used a cane for walk- sion and after surgery) [16, 17]. ing outdoor, and needed augmented time for self-care. This case report describes the beneficial effect of MT for The patient suffered from persistent allodynia and the reduction of pain of the upper limb in a subject pre- dysesthesia at her left upper extremity and at the left senting CPSP in the left body side combined to sensory side of her face. The soft touch of an open hand was loss and mild movement disorders after a right haemor- perceived as burning. Proprioception of the left upper rhagic lenticular-capsular, thalamic stroke occurred five and lower limbs was also impaired, as in the absence of years before. To the best of our knowledge, the effect of vision she was unable to locate her left arm and leg. MT for the treatment of CPSP has never been observed Three years before, she was prescribed Pregabalin to re- despite it has been defined deserving to be explored [11]. lieve pain (the patient does not remember the dosage), however she quit its use after few months because of its Case presentation inefficacy. At the moment of the treatment she was not taking medications. At clinical examination, she pre- Case description sented low weakness at her left side (grade 4 out of 5 at The patient was a 50-year old, right-handed woman who Medical Research Council grading for strength of shoul- experienced a haemorrhagic stroke in 2010. The com- der abductors, elbow flexors, wrist extensors, hip exten- puted tomography scan performed immediately after the sors, hip abductors, knee extensors, ankle dorsiflexors acute event revealed a right lenticular-capsular, lateral and plantarflexors) [19] and slight spasticity (modified thalamic and intraparenchymal hematoma, with a mid- Ashworth Scale grade 1) during the elbow and wrist line shift toward the left side. Two days after the acute extension; the knee flexion and the foot dorsal flexion event a magnetic resonance imaging scan of the brain [20]. She was able to perform isolated movements of the confirmed the presence of this lesion (Fig. 1). After six foot correctly and to move each finger of the left hand weeks as inpatient for a rehabilitation program in our singularly. Institute, the patient entered a daily physiotherapy pro- gram for two months as outpatient, which led to a good Intervention recovery in motor control and strength. The postural balance improved after few sessions of spe- cific training on the treadmill and on instable surfaces, Five years after the acute event, she returned to the the patient achieved the ability to maintain the standing Outpatient Service complaining of a CPSP syndrome, a position on the left leg for some seconds. Afterwards, persisting difficulty in the functional use of the left side during one session targeted at improving coordination of her body, especially of the lower limb during walking and stair climbing and a reduced postural balance. She Fig. 1 Magnetic resonance images showing the acute right lenticular-capsular-thalamic stroke lesion assessed with T1-weighted, Fluid-attenuated inversion recovery (FLAIR), and T2-weighted sequences (a); one-year follow-up MRI showing the lesion evolution assessed with FLAIR sequence. The red arrows show the posterior thalamic involvement (b)

Corbetta et al. Archives of Physiotherapy (2018) 8:4 Page 3 of 6 of the lower extremity, the patient performed exercises Results with a visual feedback provided by a mirror. After these Before starting MT, the patient was stable as for func- exercises, the patient presented a positive good sensation tional use of the upper extremity and pain (Table 1). at the leg, not related to movement, reporting that the After MT training, the patient showed a reduction of 4.5 leg was “more sensitive”. In the light of this unexpected points at rest and of 3.9 points during the maximal vol- finding, she was proposed to start MT in order to reduce untary contraction in the VAS score of the trained hand. pain at the upper extremity. The patient completed two The patient reported a reduction of pain intensity while consecutive weeks of MT training for five days a week. burning sensation was still present, however the reduc- In each session, she was asked to perform symmetrical tion of pain was described as a “significant relief”. A bilateral movements with the upper extremities while slight reduction of VAS score for shoulder pain occurred watching the image of the sound limb reflected by a also, 1.2 points at rest and 2.3 points during the maximal parasagittal mirror superimposed to the image of the voluntary contraction. Hand strength and dexterity did affected arm. Each session lasted 45 min. The requested not show relevant variations after treatment, a difference movements were: forearm prono-supination, wrist ex- of 3.7 N and 2.5 s from pre-treatment to post-treatment tension and opening and closing the hand (Fig. 2). These respectively reflect the difference observed during the movements were always proposed in a random order. two baseline assessments (Table 1). Each movement was performed for 10 min at spontan- eous speed (about one movement every second). Five The subject remained positively impressed by pain minutes were spent for resting and for self-mobilization reduction after MT and autonomously decided to get a of the left arm and hand without the mirror. During the “mirror box” in order to continue the training at home. exercise, the patient was supervised by a physiotherapist. No further instructions, corrections or encouragements One year later, relative to pre-treatment, the reduction were given. of hand pain persisted both at rest and maximal voluntary contraction (Table 1). In addition, shoulder pain was fur- Outcomes ther reduced (Table 1). It is worth noting that during the Pain severity was assessed by the visual analogue scale year, the patient performed spontaneously and steadily at (VAS 0-10 cm): the patient was asked to draw a vertical home a similar MT training for the shoulder with flexion, line on a horizontal 10 cm line, where she felt the pain abduction and external rotation with elbow flexed at 90°. intensity would be better represented, in a range from the left end of the line indicating “0 = no pain” to the Discussion and conclusions right one indicating “10 = worst pain imaginable”. Hand This case report shows the application of MT on the and finger strength was assessed by the dynamometers upper extremity for the reduction of CPSP in a patient Jamar and Pinch Gauge, and finger dexterity by the 9- with a thalamic stroke occurred five years before. Find- Hole Peg Test [21]. The patient was evaluated at baseline, ings from literature support the use of MT at least as about one month before starting MT in two different add-on rehabilitation intervention for improving motor occasions one week apart (to assess reproducibility), im- function in patients with stroke. The suppression of in- mediately before treatment and after treatment. VAS score appropriate proprioceptive input by visual information was also obtained at one-year follow-up. VAS was used to and/or the sensorimotor neural plasticity induced by assess pain severity at the hand and at the shoulder [22] in MT may help motor recovery [12, 24]; furthermore, the two separate conditions: at rest and during a maximal application of MT in people with stroke presenting com- squeeze left hand contraction [23]. plex regional pain syndrome (CRPS) type I showed a sig- nificant effect on pain relief [25, 26]. In our patient, the MT was used with the aim to reduce CPSP occurring Fig. 2 Example of upper limb movements performed by our patient during the mirror therapy training

Corbetta et al. Archives of Physiotherapy (2018) 8:4 Page 4 of 6 Table 1 Clinical findings before and after mirror therapy Baseline 1 Baseline 2 Pre MT Post MT Δ pre-post 12 months Δ pre-12 months follow-up follow-up Shoulder pain (VAS, cm) 7.5 7.2 6.7 5.5 1.2 0.2 6.5 Maximal Grip voluntary contraction 8.8 9.1 9.6 7.3 2.3 0.2 9.4 Hand pain (VAS, cm) 5 4.6 5.3 0.8 4.5 1.5 3.8 Maximal Grip voluntary contraction 6.8 7.3 7.2 3.3 3.9 3.3 3.9 Hand Strength (N) 22 20 21.3 25 3.7 n.a. Finger Strength (N) 5 4 4.6 4.6 0 n.a. Dexterity (sec) 23\" 20\" 21.5\" 24\" 2.5\" n.a. MT: mirror therapy; Δ: Difference pre-post, difference pre-12 months follow-up; n.a.: not assessed. Baseline assessme nts were performed about one month before starting MT in two different occasions one week apart. Pain was assesse d with the Visual Analogue Scale (VAS 0-10 cm), strength was assessed with Jamar and Pinch dynamometers (values are expressed in Newton, N), and finger dexterity was assessed with the Nine Hole Peg Test (values are expressed in seconds) after a haemorrhagic thalamic-capsular lesion in a pa- what is felt [32]. According to this hypothesis, the MT tient presenting pain and other sensory disturbances [7] combination of visual illusion and movement would rather than impairment of movement. lead the CNS to reach a “sensory congruence”, which in turn would contribute to pain reduction [32, 33]. The pathophysiology of the CPSP is still unclear and Interestingly, the patient’s awareness of the sensory different mechanisms involving the thalamus are sus- illusion does not reduce the attempt of the CNS to pected to underlie this phenomenon including deaffer- achieve the sensory coherence between visual and pro- entation of ascending pathways (leading to sensory prioceptive information [34]. This would be in accord- loss), disinhibition of its medial portion (leading to ance with Ramachandran et al. [11] who suggested hypersensitivity), and abnormalities in spinothalamic that pain mechanism is not influenced by the aware- function (leading to decreased or increased sensation ness of the trick. of temperature, especially cold) [4]. These mechanisms are usually exacerbated by excitotoxic and inflamma- The change of pain perception after MT extended also tory changes caused by the haemorrhagic lesion, the shoulder, suggesting that the whole upper limb under- resulting in a perception of pain even if it is not “acti- went a sensory re-organization. Nevertheless, a specificity vated” by noxious stimuli [27] (resulting in chronic of MT training and its long-term effect are suggested by pain). Furthermore, the altered balance between inhib- the fact that the MT training addressed to the shoulder ition and facilitation of sensory-motor brain areas has (spontaneously and steadily performed by the patient at been proposed as a possible underlying mechanism of home) specifically reduced shoulder pain at one-year central pain [4]. Particularly, a lesion of the lateral follow-up. thalamus has been identified as one of the most com- mon causes of CPSP [28]. One may speculate that in Limitations are present; the observation of a single our subject the lesion of lateral thalamus and lenticu- subject do not allow generalization: even if similar re- lar nucleus could induce an alteration in the func- sults are reported in studies enrolling people after stroke tional connectivity between basal ganglia and primary/ and presenting CPRP type I [25, 35]. During the treat- secondary somatosensory cortices which are involved ment, the patient was asked not to change her lifestyle in the sensory-discriminative dimension of pain, pain and not to introduce medications but the presence of intensity perception and nociceptive information pro- co-interventions was only orally reported by the patient. cessing. According to previous findings, MT could The positive response to MT and the reporting of pain optimize the altered balance between ispilesional and could have been influenced by the positive approach contralesional sensory-motor areas activation caused toward a new treatment. Lastly, a more comprehensive by the maladaptive reorganization of the somatosen- assessment of pain including other assessments than sory cortices, thus reducing pain perception [29, 30]. VAS would have improved the description of patient and Another hypothesis explaining pain reduction pro- of findings. In case of future trials assessing the effect of vided by MT after limb amputation relies on a sup- MT for reduction of CPSP, researchers should consider pression of sensory discrepancies between vision and to include the assessment of neuropathic pain through proprioception [11, 31]. In line with these findings, in the use of other instruments such as the painDETECT, our patient, after a short-term MT addressed to hand the Douleur Neuropathique en 4 Questions (DN4), the and forearm, the perception of pain changed. The Leeds Assessment of Neuropathic Symptoms and Signs brain continuously matches visual and kinaesthetic in- (LANSS) or the Neuropathic Pain Scale [36, 37]. These puts during movements, linking what is seen with questionnaires have high sensitivity and specificity for the detection of neuropathic pain, ranging from 83% to

Corbetta et al. Archives of Physiotherapy (2018) 8:4 Page 5 of 6 85% for the sensitivity, for DN4 and LANSS respectively, 8. Sposato LA, Sharma HA, Khan AR, et al. Thalamic cramplike pain. J Neurol and from 80% to 90% for specificity, for painDETECT Sci. 2014;336:269–72. and DN4 respectively [38]. 9. Nasreddine ZS, Saver JL. Pain after thalamic stroke: right diencephalic In conclusion, this case report suggests that MT after predominance and clinical features in 180 patients. Neurology. 1997;48: stroke may affect the perception of pain resulting from 1196–9. CNS lesions and it could be considered as an additional approach to treat people presenting CPSP. 10. Vestergaard K, Nielsen J, Andersen G, et al. Sensory abnormalities in consecutive, unselected patients with central post-stroke pain. Pain. 1995;61: Abbreviations 177–86. CNS: Central nervous system; CPSP: Central post stroke pain; CRPS: Complex Regional Pain Syndrome; MT: Mirror therapy; VAS: Visual analogue scale 11. Ramachandran VS, Altschuler EL. The use of visual feedback, in particular mirror visual feedback, in restoring brain function. Brain. 2009;132:1693–710. Funding None 12. Thieme H, Mehrholz J, Pohl M, et al. Mirror therapy for improving motor function after stroke. Cochrane Database Syst Rev. 2012;3:CD008449. Availability of data and materials All data generated or analysed during this study are included in this 13. Ramachandran VS, Rogers-Ramachandran D. Synaesthesia in phantom limbs published article. induced with mirrors. Proc Biol Sci. 1996;263:377–86. Authors’ contributions 14. Yavuzer G, Selles R, Sezer N, et al. Mirror therapy improves hand function in All authors designed the study, collected and interpreted the data, drafted or subacute stroke: a randomized controlled trial. Arch Phys Med Rehabil. 2008; critically revised the article and approved the final draft. 89:393–8. Ethics approval and consent to participate 15. Sutbeyaz S, Yavuzer G, Sezer N, Koseoglu BF. Mirror therapy enhances No ethical approval needed for the use of a treatment already in use with lower-extremity motor recovery and motor functioning after stroke: a people after a stroke. randomized controlled trial. Arch Phys Med Rehabil. 2007;88:555–9. Consent for publication 16. Giraux P, Sirigu A. Illusory movements of the paralyzed limb restore motor Written permission from the patient to participate and to publish results of cortex activity. NeuroImage. 2003;20:S107–11. this report is obtained. 17. Moseley GL, Gallace A, Spence C. Is mirror therapy all it is cracked up to be? Competing interests Current evidence and future directions. Pain. 2008;138:7–10. The authors declare that they have no competing interests. 18. Keith RA, Granger CV, Hamilton BB, Sherwin FS. The functional Publisher’s Note independence measure: a new tool for rehabilitation. Adv Clin Rehabil. 1987;1:6–18. Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. 19. Gregson JM, Leathley MJ, Moore AP, et al. Reliability of measurements of muscle tone and muscle power in stroke patients. Age Ageing. 2000;29: Author details 223–8. 1Laboratory of Analysis and Rehabilitation of Motor Function, San Raffaele Scientific Institute, Vita-Salute San Raffaele University, Via Olgettina 60, 20132 20. Bohannon RW, Smith MB. Interrater reliability of a modified Ashworth scale Milan, Italy. 2Neuroimaging Research Unit, Institute of Experimental of muscle spasticity. Phys Ther. 1987;67:206–7. Neurology, Division of Neuroscience, San Raffaele Scientific Institute, Milan, Italy. 3Department of Neurology and Neuroimaging Research Unit, Institute 21. Beebe JA, Lang CE. Relationships and responsiveness of six upper extremity of Experimental Neurology, Division of Neuroscience, San Raffaele Scientific function tests during the first six months of recovery after stroke. J Neurol Institute, Vita-Salute San Raffaele University, Milan, Italy. 4Humanitas Clinical Phys Ther. 2009;33:96–103. and Research Center, Humanitas University, Rozzano, Italy. 22. Dworkin RH, Turk DC, Wyrwich KW, et al. Interpreting the clinical Received: 7 September 2017 Accepted: 14 February 2018 importance of treatment outcomes in chronic pain clinical trials: IMMPACT recommendations. J Pain. 2008;9:105–21. References 1. Déjerine JJ, Roussy G. Le syndrome thalamique. Rev Neurol (Paris). 1906;14: 23. Bohannon RW, Schaubert K. Test-retest reliability of grip-strength measures obtained over a 12-week interval from community dwelling elders. J Hand 521–32. Ther. 2005;18:426–8. 2. Flaster M, Meresh E, Rao M, Biller J. Central postroke pain: current diagnosis 24. Altschuler EL, Wisdom SB, Stone L, et al. Rehabilitation of hemiparesis after and treatment. Top Stroke Rehabil. 2013;20:116–23. stroke with a mirror. Lancet. 1999;353:2035–6. 3. Yekutiel M. Sensory re-education of the hand after stroke. Whurr: 25. Cacchio A, De Blasis E, De Blasis V, Santilli V, Spacca G. Mirror therapy in Philadelphia PA; 2000. complex regional pain syndrome type 1 of the upper limb in stroke 4. Klit H, Finnerup NB, Jensen TS. Central post-stroke pain: clinical patients. Neurorehabil Neural Repair. 2009;23(8):792–9. characteristics, pathophysiology, and management. Lancet Neurol. 26. Pervane Vural S, Nakipoglu Yuzer GF, Sezgin Ozcan D, Demir Ozbudak S, 2009;8:857–68. Ozgirgin N. Effects of mirror therapy in stroke patients with complex 5. Hong JH, Choi BY, Chang CH, et al. The prevalence of central poststroke regional pain syndrome type 1: a randomized controlled study. Arch Phys pain according to the integrity of the spinothalamo-cortical pathway. Eur Med Rehabil. 2016;97(4):575–81. Neurol. 2012;67:12–7. 6. Lee IH, Kim YN, Son CS, et al. Clinical aspects of screening test tools for 27. Woolf C. Central sensitization: implications for the diagnosis and treatment central neuropathic pain in patients with thalamic stroke. J Phys Ther Sci. of pain. Pain. 2011;152:S2–1. 2011;23:749–52. 7. Andersen G, Vestergaard K, Ingeman-Nielsen M, Jensen TS. Incidence of 28. Sprenger T, Seifert CL, Valet M, Andreou AP, Foerschler A, Zimmer C, Collins central post-stroke pain. Pain. 1995;61:187–93. DL, Goadsby PJ, Tölle TR. Chakravarty MM assessing the risk of central post- stroke pain of thalamic origin by lesion mapping. Brain. 2012;135(Pt 8): 2536–45. 29. Foell J, Bekrater-Bodmann R, Diers M. Flor H mirror therapy for phantom limb pain: brain changes and the role of body representation. Eur J Pain. 2014;18(5):729–39. 30. Franz EA, Fu Y, Moore M, Winter T, Mayne T, Debnath R, Stringer C. Fooling the brain by mirroring the hand: brain correlates of the perceptual capture of limb ownership. Restor Neurol Neurosci. 2016;34(5):721–32. 31. Harris AJ. Cortical origin of pathological pain. Lancet. 2000;355:318–9. 32. Proske U, Gandevia SC. The kinaesthetic senses. J Physiol. 2009;587:4139–46. 33. Metral M, Chancel M, Brun C, et al. Kinaesthetic mirror illusion and spatial congruence. Exp Brain Res. 2015;233:1463–70. 34. Tesio L. Secondo Rapporto sull’ictus. Disabilità, riabilitazione, ricerca. Masson 2006. 35. Cacchio A, De Blasis E, Necozione S, Di Orio F, Santilli V. Mirror therapy for chronic complex regional pain syndrome type 1 and stroke. N Engl J Med. 2009;361(6):634–6.

Corbetta et al. Archives of Physiotherapy (2018) 8:4 Page 6 of 6 36. May S, Serpell M. Diagnosis and assessment of neuropathic pain. F1000 Med Rep. 2009;1:76. 37. Haanpää M, Attal N, Backonja M, Baron R, Bennett M, Bouhassira D, Cruccu G, Hansson P, Haythornthwaite JA, Iannetti GD, Jensen TS, Kauppila T, Nurmikko TJ, Rice AS, Rowbotham M, Serra J, Sommer C, Smith BH, Treede RD. NeuPSIG guidelines on neuropathic pain assessment. Pain. 2011;152(1):14–27. 38. Dach Eckeli F, Teixeira RA, Gouvêa AL. Neuropathic pain evaluation tools. Rev Dor São Paulo. 2016;17(Suppl 1):S20–2. Submit your next manuscript to BioMed Central and we will help you at every step: • We accept pre-submission inquiries • Our selector tool helps you to find the most relevant journal • We provide round the clock customer support • Convenient online submission • Thorough peer review • Inclusion in PubMed and all major indexing services • Maximum visibility for your research Submit your manuscript at www.biomedcentral.com/submit

Esakki et al. Archives of Physiotherapy (2018) 8:5 https://doi.org/10.1186/s40945-018-0046-z RESEARCH ARTICLE Open Access Rasch analysis of the patient-rated wrist evaluation questionnaire Saravanan Esakki1* , Joy C. MacDermid1,3, Joshua I. Vincent1, Tara L. Packham2, David Walton1 and Ruby Grewal3,4 Abstract Background: The Patient-Rated Wrist Evaluation (PRWE) was developed as a wrist joint specific measure of pain and disability and evidence of sound validity has been accumulated through classical psychometric methods. Rasch analysis (RA) has been endorsed as a newer method for analyzing the clinical measurement properties of self-report outcome measures. The purpose of this study was to evaluate the PRWE using Rasch modeling. Methods: We employed the Rasch model to assess overall fit, response scaling, individual item fit, differential item functioning (DIF), local dependency, unidimensionality and person separation index (PSI). A convenience sample of 382 patients with distal radius fracture was recruited from the hand and upper limb clinic at large academic healthcare organization, London, Ontario, Canada, 6-month post-injury scores of the PRWE was used. RA was conducted on the 3 subscales (pain, specific activities, and usual activities) of the PRWE separately. Results: The pain subscale adequately fit the Rasch model when item 4 “Pain - When it is at its worst” was deleted to eliminate non-uniform DIF by age group, and item 5 “How often do you have pain” was rescored by collapsing into 8 intervals to eliminate disordered thresholds. Uniform DIF for “Use my affected hand to push up from the chair” (by work status) and “Use bathroom tissue with my affected hand” (by injured hand) was addressed by splitting the items for analysis. After background rescoring of 2 items in pain subscale, 2 items in specific activities and 3 items in usual activities, all three subscales of the PRWE were well targeted and had high reliability (PSI = 0.86). These changes provided a unidimensional, interval-level scaled measure. Conclusion: Like a previous analysis of the Patient-Rated Wrist and Hand Evaluation, this study found the PRWE could be fit to the Rasch model with rescoring of multiple items. However, the modifications required to achieve fit were not the same across studies, our fit statistics also suggested one of the pain items should be deleted. This study adds to the pool of evidence supporting the PRWE, but cannot confidently provide a Rasch-based scoring algorithm. Keywords: PRWE, Rasch analysis, Patient-reported outcome measure, Distal radius fracture Background resulted in a proliferation of studies on commonly used Patient-reported outcome measures have become a outcome measures in upper extremity rehabilitation to cornerstone of evaluation in hand therapy and hand sur- evaluate measurement properties in a spectrum of con- gery [1, 2]. A well-developed patient-reported outcome texts and patient populations [5]. measure (PROM) can provide a clinically relevant evalu- ation of the patient perspective and status to inform Rasch analysis (RA) is a form of mathematical model- health care decisions [1, 2]. Classical test theory (CTT) ing employed to develop new outcome measures and ap- forms the basis for most evaluations of measurement praise the properties of existing instruments [6]. RA properties [3]. A core tenet of CTT is validity and reliabil- extends the measurement evaluation by critically evalu- ity results apply only to the sample studied [4]. This has ating discrete items and scores, which is an advantage over CTT [6]. The Rasch rating scale is based on item * Correspondence: [email protected] response theory (IRT) [7]. However, according to its de- 1School of Physical Therapy, Western University, London, ON, Canada veloper George Rasch, the central differentiating feature Full list of author information is available at the end of the article between the Rasch and IRT is the defining role of © The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.

Esakki et al. Archives of Physiotherapy (2018) 8:5 Page 2 of 11 specific objectivity, which presumes both individuals and of the participants [16]. Return to work status has been items can be rated [8]. Rasch uses probabilistic modeling used for many years by rehabilitation professionals as an to determine the degree to which items on a scale func- objective indicator of function, and is an increasingly tion as linear (interval-level) measurement of the latent popular measure of overall recovery from injury [17, 18]. construct, or domains of interest. Further, it models the predicted amounts of this latent construct within the in- Given the return to work is considered a vital indicator dividuals studied [9, 10]. While this interval-level of of recovery [19], current work status (compared to pre- measurement is a pre-requisite for much statistical ana- injury) may be an important person factor upon which lysis, many scales developed within CTT fail to meet this to evaluate the measurement properties of a question- interval measurement standard and are used for decision naire. While there is no reason to believe the difference making and statistical purposes, this may ultimately in- in wording of the instructions (the only difference be- fluence the validity of research findings [5]. tween the PRWHE and PRWE is ‘wrist/hand’ is substituted for ‘wrist’ on the PRWHE) would substan- The Patient-Rated Wrist Evaluation [PRWE] [11] (see tially influence measurement properties, there is an op- Additional file 1) is a patient-reported outcome measure portunity to replicate the Rasch analyses of the PRWHE intended to quantify perceptions of pain and disability using the PRWE. Furthermore, RA of PRWE data gath- evolving from wrist conditions. The PRWE has been ered in a different practice setting and from a different used in more than 150 studies and has been recom- patient population would provide an opportunity to con- mended as a core measure for evaluating outcomes in trast and compare the stability of findings in the differ- distal radius fracture (DRF) [12]. While the PRWE ques- ent practice environment. Although RA is thought to be tionnaire was originally developed for distal radius frac- less affected by participant samples, this has not always ture, the scoring instructions (but not the items) were been empirically supported [20, 21]. Taking these con- later modified as the Patient-Rated Wrist and Hand siderations together, there is a need to further explore Evaluation [PRWHE] [12, 13] to address not only wrist the measurement properties of the PRWE using the but also hand conditions. A recent systematic review Rasch paradigm with a different population. summarized 22 studies examined measurement proper- ties of the PRWE and found strong supporting evidence The purpose of this study is to utilize RA to evaluate for reliability and responsiveness [14]. The validity of the the PRWE in a cohort of persons following distal radius PRWHE version is supported for the use in patients with fracture: wrist and hand conditions by demonstrating similar re- sponsiveness with the components of the Disabilities of 1) To test the construct validity of the pain and dis- the Arm, Shoulder, and Hand (DASH) [15], but the sys- ability subscales of the PRWE by examining the unidi- tematic review noted a gap in clinically relevant indica- mensionality of the scales, and to evaluate the reliability tors like the minimally significant difference. While as defined by Rasch traditions. triangulating findings from different measurement models such as CTT and Rasch can lend confidence to 2) To examine the interval-level properties of the pain the properties of a scale, only a single study reported the and disability scales of the PRWE by examining the fit to properties of the instrument using RA. Packham and the Rasch model and ordering of item thresholds. MacDermid conducted RA on the PRWHE using 264 patient records representing a mixture of wrist and hand 3) To examine the potential for bias in PRWE score injuries and found good to excellent reliability of the based on age, gender and work status of respondents, scale [16]. The study found no significant differential and to explore solutions for minimizing any bias. item functioning, or scale differences between the people with injuries in the dominant hand when compared to Methods the injuries in the non-dominant hand. A key finding was best fit to the Rasch model was attained if the dis- Research design ability scale was split into the specific activities and usual Cross-sectional study using RA activities components for analysis [16]. Background re- scoring of some items was also required: however, the Instrument and procedures The 15-item PRWE ques- authors cautioned that revising the PRWHE based on a tionnaire evaluates the domains of pain and disability with single study would be premature. A critical limitation of three subscales: pain, specific activities, and usual activ- this study was some patients were represented multiple ities. The pain subscale has five items, rated as 0 = no pain times (at different time points in their recovery) within to 10 = worst pain ever. Functional interference, or disabil- the dataset. Another concern limiting the application of ity, is represented by six items on a specific activities sub- the findings in the study did not address the work status scale and four items on usual activities: these are rated on a 0–10 scale where the subject scores the amount of diffi- culty in performing the activity, with 0 = no difficulty in performing an activity and 10 ‘unable to perform the ac- tivity’. The final PRWE score represents equally weighted pain and disability (function) scores. We selected the 6-

Esakki et al. Archives of Physiotherapy (2018) 8:5 Page 3 of 11 month post-injury scores for analysis as a return to work each class for differential item functioning analysis [26]. status was also evaluated at this time point. In this instance, 4 class intervals were generated. Data collection Thresholds Category probability curves were used to identify disor- Sample size and characteristics dered thresholds, item misfit and inconsistent use of We conducted a secondary analysis of a convenience items. Disordered thresholds arise when the respondents (existing) cross-sectional data set consisting of 300 and find it difficult to differentiate between the item re- 82 patient’s PRWE scores 6 months post-injury, col- sponse options [27]. This occurs when there are too lected at the Hand and Upper Limb clinic at a large aca- many response options, or the selection options are demic healthcare organization in London, Ontario, similar to one another, confusing or open to misinter- Canada. Ethics approval for the original study was re- pretation. Disordered thresholds can be corrected either ceived from Western University ethics board. Men and by rescaling the tool or by collapsing the categories and women accounted for 32.5% and 67.5% of the sample re- revising the response option to improve the overall fit to spectively: mean age was 58.5 years. 67% (n = 256) were the model [27]. working and 33% (n = 126) were non-working popula- tion. 19.6% (n = 78) participants have left hand as dom- Fit statistics inance and 61.2% (n = 234) participants have DRF at the The following important fit statistics are inspected when dominant side hand (right or left). RA requires large analyzing the fit of the data to the Rasch model. samples to ensure adequate distribution of responses for analysis across all levels of the condition or construct of Unidimensionality interest: a minimum of 250 or at least 10 endorsements Unidimensionality is one of the main assumptions for the for each potential response category for each item have data to fit the Rasch model. The absence of any meaningful been suggested as standards for ensuring adequacy of pattern in the residuals reveals the presence of unidimen- sample size [21]. sionality [6]. A test proposed by Smith [28] examines the relationship between items and the first residual factor The paper copy of the PRWE questionnaire was com- identified by principal components analysis and uses these pleted by the patients, and the data were compiled in patterns to define 2 subsets of items. By then testing using SPSS for demographic examination and then imported paired t-tests, we could see if the person estimate derived into the RUMM2030 version 5.1(RUMM Laboratory Pty from these subsets significantly differs from that derived Ltd., Perth, Australia) [22] for RA. from all items. For the questionnaire to be unidimensional, the percentage of tests that are significant (P < 0.05) should RA be less than 5%. Final evaluation of unidimensionality is We followed the approach suggested by Lundgren Nils- completed only when all other scale adjustments have been son and Tennant [6] as described below. The 3 subscales completed [28]. of the PRWE were analyzed separately for sources of misfit to the model [23, 24]. Bonferroni correction [25] was applied throughout the analysis to reduce alpha er- rors due to multiple testing. Likelihood ratio (LR) test Local dependency The choice of Rasch model (Rating Scale vs. Partial RA employs Principal Components Analysis (PCA) of Credit) [24] was made by conducting and interpreting a the residuals to ensure the local independence of the likelihood ratio test that evaluates the likelihood that items [6]. An inter-item residual correlation > 0.2 above mathematical differences between polytomous response the average residual correlation was used as the thresh- options are equal. A significant LR suggests they are not old to indicate local dependency [6]. equal, and that the unconstrainted partial credit model should be used [24]. The local independence assumption can be violated in 2 ways: response dependency, and multidimensionality Class interval and distribution structure [29]. Response dependency occurs when the items are The size of class intervals was checked throughout the linked in such a way that the response on one item de- analyses to ensure equal distribution between the inter- termines the response on another. These dependencies vals. Class intervals are generated by the analysis soft- can be identified by the residual correlation matrix and ware after ranking the person location: the sample is rectified by combining the items into a ‘super item’ by then split into relatively equal class intervals to ensure combining 2 items or by deleting one of the dependent adequate representation of the key patient variables in items. Again, subsequent reanalysis is required to con- firm resolution of the dependency [8].

Esakki et al. Archives of Physiotherapy (2018) 8:5 Page 4 of 11 Item/person fit residuals uniform DIF occurs when the difference in performance In RA, the scale is always centered on zero logits, which varies with the level of the attributes. There is no definite represents the average item difficulty for the scale. Individ- procedure to rectify the non-uniform DIF, and therefore, ual item fit is then calculated relative to this point, or are the most common solution is for that item to be re- ‘fitted’ to the model. Person fit is then evaluated by con- written or removed from the questionnaire [6]. In this RA, sidering the mean location of persons in the sample. For a we tested DIF on working status, injured hand, dominant well-targeted measure, the mean location for persons hand, gender and age group variables. would be around the value of zero. When the mean is ap- proximately zero, and the standard deviation is close to Item difficulty Item difficulty for each of the subscales one, the item and persons (residuals) fit the model, and a was graphically represented by the generation of a hierarchical ordering of items (e.g., from low to high levels Wright map [10]. This allows the difficulty of the items of activity limitation) is achieved [4, 30]. Standardized fit to be compared and serves as a form of content valid- residuals for individual items were flagged as extreme if ation by looking for potential floor or ceiling effects. the values exceeded + 2.5. Results Item-trait interaction In this study, the likelihood ratio was statistically signifi- To analyze the property of invariance across the trait be- cant (P value < 0.05): therefore, we used partial credit ing measured, item-trait interaction is assessed using a parameters for the analysis [24]. There were no missing chi-square statistic. The chi-square statistic [31] com- data, and all 382 independent cases were determined to pares the difference in observed values with expected be valid by the RUMM 2030 software. The 3 subscales values across groups representing different ability levels were analyzed separately, as we presumed each subscale (called class intervals) across the trait to be measured represented a distinct latent trait. (e.g., pain). A significant chi-square value (< 0.05) indi- cates that the hierarchical ordering of the items varies Pain subscale across the trait, compromising the required property of Initial analysis of the 5 items on the pain subscale using invariance [31]. the partial credit model showed shows excellent Individ- ual item fit and indicated acceptable levels of discrimin- Reliability indices ation. Item 5 (How often do you have pain?) shows Person-Separation-Index (PSI) indicates the reliability of disordered threshold. The scale also shows significant the scale for estimating the amount of latent trait in any item-trait interaction (p < 0.001) and PSI of 0.84. Item 4, individual. This can also be interpreted as the ability of “Pain at its worst” showed non-uniform DIF across age the scale to identify differences among respondents [32]. groups. Local dependency was observed between item 1 A person-separation value of a minimum of 0.7 and and 5, where predictably, persons with pain at rest re- maximum of 0.95 is considered, in general, to be the ac- ported high frequencies of pain. Acceptable unidimen- ceptable level of PSI. Reliability of the fit characteristics sionality was observed. (Table 1: initial analysis). depends on the value of the PSI, with higher PSI indicat- ing higher reliability [32]. To improve the overall fit to the Rasch model item 5 was rescored by collapsing response categories from the Differential item functioning (DIF) original 0–10 responses based on the category probabil- DIF indicates potential sources of bias in-person measure- ity curves (Fig. 1) until the rescored curves show no dis- ments which result in misfit of the data to the Rasch ordered thresholds (F = 2.056, df 3, p = 0.105931), model. DIF occurs when distinct subgroups within the resulting in 8 categories (Table 2). Item 4 was deleted to sample population respond in divergent ways to the indi- eliminate the Non-Uniform DIF (Age group) (F = 2.290, vidual item even though they have equal levels of under- df 3, p = 0.078249). Re-analysis after rescoring shows lying characteristics [8]. DIF can be identified both that the person-item threshold map (Fig. 2) indicates graphically, by analyzing the item characteristic curves, that this subscale has good item coverage for wrist disor- and statistically using analysis of variance (ANOVA) [33]. ders related pain. Also, no local dependency was present DIF can occur in two forms: uniform and non-uniform. and unidimensionality was observed. Uniform DIF occurs when the difference in scoring per- formance remains constant across all respondent’s ability Specific activities subscale levels. Uniform DIF can be rectified by either combining Initial analysis showed there are no disordered thresholds items or by splitting items: subtest analyses are performed and the subscale demonstrated unidimensionality (Table 1: to verify the DIF was canceled by these adjustments. Non- initial analysis). Good Individual item fit, a good level of discrimination and high reliability of the scale was observed (Table 1: initial analysis). The fit challenges included uni- form DIF for item 4 (“Use my affected hand to push up

Esakki et al. Archives of Physiotherapy (2018) 8:5 Page 5 of 11 Table 1 Summary fit statistics for individual subscales of the PRWEa Analysis Item fits residual Person Fit residual Item-trait interaction Unidimensionality PSIa Per C < 5%a Mean SD Mean SDa Chi-square (df)a P PAIN SUBSCALE Initial −0.41 1.72 −0.45 0.94 23 (20) 0.49 6.9% 0.89 Final −0.53 1.17 −0.44 0.91 18 (20) 0.54 2.5% 0.84 SPECIFIC ACTIVITIES SUBSCALE Initial −0.60 2.13 −0.45 1.07 54 (30) 0.00 4.7% 0.81 Final −0.47 1.87 −0.37 1.03 32 (25) 0.04 - (since items were split for DIF) 0.80 USUAL ACTIVITIES SUBSCALE Initial −0.41 1.72 −0.45 0.94 23 (20) 0.28 1.5% 0.78 Final −0.43 1.58 −0.45 1.00 35 (25) 0.07 - (since items were split for DIF) 0.81 aSource of misfit to the Rasch model; SD = Standard deviation; df = Degrees of freedom; per C < 5% = proportion of t-tests that were significant at level of signifi- cance of 0.05; 95% CI = 95% confidence interval; PSI = Person separation index; PRWE – Patient-Rated Wrist Evaluation For the data to satisfy Rasch model requirements: The mean is expected to be approx. Around zero (Can range between ± 2.5); S.D. should be approx. 1; Chi-square value is expected to be small and statistically non-significant; For a measure to be unidimensional per C < 5% should be less than 0.05; if it is higher than 0.05 then look into the lower limit the 95% confidence interval if it is less than 0.05 then the measure is unidimensional PSI (Person separation index) should be greater than 0.70 for the summary statistics to be reliable; from a chair”) (F = 15.769, df 1, p = 0.000091) and item 6 became non-significant indicating an acceptable fit of (“Use bathroom tissue with my affected hand”) (F = 0.183, the data to the Rasch model (Fig. 3), and the analysis df 1, p = 0.669405) by both work status and injured hand. showed unidimensionality (Table 1: final analysis). None of the items exhibited DIF for gender or dominant hand. Local dependency was observed between the item 1 Usual activities subscale (“Turn a door knob using the affected hand”) and item 2 The usual activities subscale analysis demonstrated that (“Cut meat using a knife in my affected hand”). the scale was unidimensional with 95% confidence interval (0.0150) and reliability was good (PSI = 0.78) (Table 1: ini- To improve the overall fit of the specific activities sub- tial analysis). No DIF was observed for the injured hand, scale to Rasch model, the following actions were taken. dominant hand, and gender. However, it shows misfit to Initially, to deal with DIF, we split the item 4 for the the Rasch model with disordered thresholds for two of the work status as “Yes” and “No” for working and not four items (items 3 and 4). Item 2 “Household work working and split the item 6 for the injured hand as into (cleaning, maintenance)” showed uniform DIF for work right: left: both. This essentially creates a different status, while non-uniform DIF for age group was observed PRWE scoring system for persons in each of these cat- for item 4 “Recreational activities”. egories. Then item 1 and 2 were bundled (treated statis- tically as a single item) to address the local dependency. To increase the overall fit of the scale to the Rasch Re-analysis of the altered scale confirmed that DIF and model, the items (3 and 4) with disordered thresholds local dependency were resolved. Chi-square residual were collapsed to reorder into 7 intervals. (Table 2). Fig. 1 Pain item 5 (Pain frequency) category probability curve

Esakki et al. Archives of Physiotherapy (2018) 8:5 Page 6 of 11 Table 2 Table showing the structure of scores for individual items of the PRWE Item 0 1 2 3 4 5 6 7 8 9 10 PAIN SUB SCALE Pain - At rest 0 1 2 3 4 5 6 7 8 9 10 Pain - When doing a task with repeated wrist/hand movement 0 1 2 3 4 5 6 7 8 9 10 Pain - When lifting a heavy object 0 1 2 3 4 5 6 7 8 9 10 Pain - When it is at its worst 0 1 2 3 4 5 6 7 8 9 10 How often do you have pain?a 01233445567 SPECIFIC ACTIVITIES SUB SCALE Turn a door knob using my affected hand 0 1 2 3 4 5 6 7 8 9 10 Cut meat using a knife in my affected hand 0 1 2 3 4 5 6 7 8 9 10 Fasten buttons on my shirt 0 1 2 3 4 5 6 7 8 9 10 Use my affected hand to push up from a chair 0 1 2 3 4 5 6 7 8 9 10 Carry a 10 lb. object in my affected hand 0 1 2 3 4 5 6 7 8 9 10 Use bathroom tissue with my affected hand 0 1 2 3 4 5 6 7 8 9 10 USUAL ACTIVITIES SUB SCALE Personal activities (dressing, washing) 0 1 2 3 4 5 6 7 8 9 10 Household work (cleaning, maintenance) 0 1 2 3 4 5 6 7 8 9 10 Work (your job or everyday work)a 01233444556 Recreational activitiesa 01233444556 aRescored items; PRWE – Patient-Rated Wrist Evaluation After rescoring, summary statistics showed no local de- Fig. 4). The person-item threshold map illustrated the high pendency on the scale. To eliminate the DIF, item 2 was level of recovery seen in the sample of 6 months post re- split for work status as “Yes” and “No” for working and covery in DRF sample (pink bars). The scale has good not working (F = 4.054, df 1, p = 0.045085), and item 4 coverage across the range of abilities, and that’s illustrated was split for the age group as 0–35 years: 36–50 years: by the figures (blue bars) (Fig. 5). 51–65 years: 65 years plus for the better distribution of per- sons within this category (F = 1.693, df 3, p = 0.168899). Discussion The final analysis demonstrated the data to fit the Rasch The result of this RA adds additional evidence for accept- model, increasing the reliability of the subscale (PSI = 0.86) able measurement properties to that derived from classical and decreased chi-square value (Table 1: final analysis, test methods on the psychometric properties of the Fig. 2 Person Item threshold (Pain subscale)

Esakki et al. Archives of Physiotherapy (2018) 8:5 Page 7 of 11 Fig. 3 Person-item threshold distribution (Specific activities) PRWE. As with many Rasch analyses conducted on in- the optimal number of categories would then vary by struments that were not developed using Rasch, some question. Therefore, background rescoring is a simple so- modifications were required to achieve interval level scal- lution and is commonly used to reduce items responses to ing or fit to the Rasch model. However, these adjustments fewer categories “behind the scenes” without disturbing indicated an adjusted PRWE could provide interval level the original construction or outward appearance of the scaling and appropriate targeting for a DRF population. item that is in common use [34]. During the initial steps of our RA, we found three disor- The three subscales of the PRWE were considered sep- dered item thresholds from the pain (1 item) and usual ac- arately to accommodate the unidimensionality assump- tivity (2 items) subscales. Similar findings have been tion of the Rasch model. Unidimensionality was observed in the previous study on the measurement prop- observed in all the three subscales. This suggests each erties of the PRWHE (3 items) [16] and the analysis of the subscale represents a unique construct and would sup- similarly structured Patient-Rated Elbow Evaluation port comparisons between these components, as well as (PREE) (17 items) [33]. Other work has suggested disor- comparisons based on the total PRWE score. This is dered thresholds in 0 to 10 scales may reflect difficulty for concordant with the previous PRWHE analysis [16], respondents in finding meaningful distinctions between which was unable to fit the disability component to the 11 different response categories. Nonetheless, the concept model as a unit, and needed to look at the subscale level of 0 to 10 is commonly used and easy for patients to in order to see fit to the model. Confirmatory factor ana- understand. It is not clear if overtly rewriting items to lysis of this dataset could potentially validate this per- have fewer categories would be beneficial, especially since spective and is warranted. While it is common to use a Fig. 4 Usual activities category probability curve

Esakki et al. Archives of Physiotherapy (2018) 8:5 Page 8 of 11 Fig. 5 Usual activities Person-item threshold single summary score in reporting outcomes in clinical transferring that is often reported by older adults aged studies, it has become more apparent that pooling differ- 65 and above [37]. However, work status and age are ent constructs can have drawbacks. Recent recommen- correlated in this sample which reflects the epidemiology dations around the use of core measures suggest that of DRF, since most older adults aged 65 and above pain and disability subscales should be considered as would not be working [38]. So, it is difficult to determine separate constructs [33]. Our RA approach is consistent which factors drive the differential item functioning. with that recommendation. Given that trials will con- Gender or dominant hand did not exhibit DIF, which tinue to prefer to use a single summary outcome meas- shows that this scale has good construct validity. A dom- ure, it is advisable that such studies also examine inance effect has been reported in some patient-reported subscale differences in outcome to avoid inaccurate in- outcome measures, such as the DASH [15], however, the terpretation of the impact of interventions. PRWE instructions refer to “the affected hand”, whereas the DASH refers to difficulty at the person-level: this In the pain subscale, we found non-uniform DIF was ex- may account for the differences seen in the importance hibited based on age group. This misfit should be consid- of dominance as an outcome mediator. The local de- ered in the context that previous pain studies indicate the pendency observed between the “Turn a door knob perception of pain differs from person to person and also using the affected hand” and “Cut meat using a knife in pain tolerance is reduced as people age [35, 36]. A mean- my affected hand” has not been reported in other Rasch ingful pattern of local dependency was observed between analyses of the PRWHE, PREE and is not intuitive. It items 1 “pain at rest” and 5 (pain frequency). If the pain is may be that difficult with a hand grip that occurs follow- reported at rest, the person is also likely to have pain more ing DRF links these items in this sample. Since the prior frequently. But these local dependency issues were accom- RA included different diagnoses, this link may not have modated when subtest analysis was performed, demon- occurred in a more heterogeneous sample. Because it strating the impact of this correlation is mitigated by has not been reported in previous studies, the depend- always considering both items together in the context of ency may not generalize beyond this sample. the scale. In terms of clinical implications, it is important to note these adjustments we made to fit the PRWE to the In the usual activities subscale, “Household work (clean- Rasch model are to support the accuracy of interval level ing, maintenance)” was the source of misfit. The analysis scoring [33] for research comparisons to other interval shows uniform DIF for the work status: this may be due level variables such as grip strength. to the differences in the perception of the household work among the people who are working and who are not, and In the specific activities subscale, uniform DIF was ob- their contribution to household tasks. People who hold served by work status for “use my affected hand to push paid employment outside the home may contribute rela- up from a chair.” It may be that people in the workplace tively less to household work when compared to people have more possibilities of using the injured hand while who stay at home [39]. This could explain the reason for pushing up from the chair when compared to the non- observing a uniform DIF. This finding was similar to RA working participants thus calibrating this differently. results reported for the PREE, where uniform DIF was ob- Similarly, in usual activities subscale, differential item served on “Household work (cleaning, maintenance)” functioning by age groups may reflect the difficulty in

Esakki et al. Archives of Physiotherapy (2018) 8:5 Page 9 of 11 Table 3 Transformation Matrix for Converting Raw Ordinal Level from the usual activities subscale for gender. In our sam- Scores to Interval-level Scores, Using the Revised Scoring Where ple, non-uniform DIF was observed between different age the Scale is Out of a Maximum of 10 Points. This conversion can groups for “Recreational activities”, perhaps due to the be used only with the modified PRWE questionnaire age-based differences in participation as the type of activ- ities, intensity of the recreational activities, and the value ROW SCORE (ORDINAL) LOGIT LOCATION INTERVAL-LEVEL SCORE of participation in recreation also differs between the age groups [38, 40]. Younger adults often engage in high- 0 −5.49 0.06 intensity recreational activities while older adults may tend to do mild to moderate intensity recreational activ- 1 −4.49 4.34 ities, but may perceive the level of intensity differently. Recreational activities in older adults are most likely to 2 −3.79 7.38 participate in activities such as watching television or lis- tening to the radio and leisure walking [37, 38]. This may 3 −3.30 9.50 explain why the participants answered this question differ- ently. Since age is one of the most commonly reported el- 4 −2.91 11.18 ements in clinical research studies, the distributions of this may need to be considered when interpreting the 5 −2.58 12.60 patient-reported outcomes using the PRWE in clinical studies of persons with wrist conditions. This also reflect- 6 −2.28 13.86 ive of the nature of the DRF population, where fractures in younger persons are most likely high velocity related to 7 −2.01 15.03 sports and MVA, in comparison to older adults where low-velocity injuries such as falls from standing height 8 −1.76 16.11 predominate [41]. 9 −1.53 17.13 The strengths of the current study are its high PSI values and the excellent power of fit with a sample size of 382 pa- 10 −1.30 18.09 tients. Although we used available data from a DRF popula- tion for this secondary analysis, we were able to examine 11 −1.09 18.99 for DIF based on gender, age, work status and hand domin- ance and affected side. In Rasch tradition, item and person 12 −0.90 19.84 measures are not considered sample-dependent if the data can be shown to fit the Rasch model after adjustment for 13 −0.71 20.64 DIF. We were able to build on previous work using RA to examine the predominantly similar outcome measure the 14 −0.54 21.38 PRWHE [16] while addressing the limitations of that previ- ous study. While we anticipated measurement properties 15 −0.37 22.09 would be similar given the small variations between the scales, we now have empirical data confirming this. More 16 −0.22 22.75 importantly, our study addresses previously unreported measurement properties including the impact of the work- 17 −0.08 23.37 ing status of the participants on the performance of the PRWE. Return to work is a considered a vital indicator of 18 0.05 23.95 recovery [18, 19], this study utilized the current and pre- injury work status as a person factor for analyzing the 19 0.18 24.50 measurement properties of PRWE. The ANOVA supports that even though the working and the non-working popula- 20 0.30 25.02 tions answered the question differently, still the mean total PRWE score of people who have returned to work (− 1.47 21 0.41 25.52 logits) and those not working (− 1.48 logits) have no statis- tical difference (F (1,379) = 0.033, p = 0.86). 22 0.52 26.00 The limitations of the current study are that the data 23 0.63 26.46 were collected only from one location and at a single re- covery time point. As this data represents participant 24 0.73 26.91 status at 6 months post-injury, we might expect a floor 25 0.84 27.35 26 0.94 27.79 27 1.04 28.23 28 1.15 28.68 29 1.25 29.14 30 1.36 29.62 31 1.48 30.13 32 1.60 30.65 33 1.73 31.22 34 1.87 31.82 35 2.03 32.48 36 2.19 33.20 37 2.38 34.02 38 2.61 34.97 39 2.88 36.15 40 3.24 37.69

Esakki et al. Archives of Physiotherapy (2018) 8:5 Page 10 of 11 bias, as persons are generally recovered and may have Publisher’s Note very low scores on the PRWE. However, this analysis suggests even at this lower end of scores, the data still fit Springer Nature remains neutral with regard to jurisdictional claims in published the Rasch model. While RA supposes that results should maps and institutional affiliations. be transferable across patient populations [9], our study does not address if similar findings would occur across Author details diverse patient populations, or across a DRF population 1School of Physical Therapy, Western University, London, ON, Canada. at different time points in recovery. Finally, the rescor- 2School of Rehabilitation Science, McMaster University, Hamilton, ON, ing, splitting and bundling of items required to achieve Canada. 3The Hand and Upper Limb Centre, St Joseph’s Health Centre, model fit is likely not operational for general clinical London, ON, Canada. 4Department of Surgery, University of Western Ontario, practice, and future work should confirm our findings London, ON, Canada. before undertaking the development of systems to facili- tate Rasch scoring, such as specific conversion tables Received: 26 June 2017 Accepted: 8 February 2018 (Table 3) or digital apps. References Conclusion 1. MacDermid JC. Patient-reported outcomes state-of-the-art hand surgery and In conclusion, this Rasch analysis suggests the psychomet- ric measurement and interval level properties of the future applications. Hand Clin. 2017;30:293–304. PRWE are robust, reinforcing previous studies on PRWE/ 2. Goldhahn J, Beaton D, Ladd A, MacDermid JC, Hoang A. Recommendation PRWHE’s psychometric properties using both classical test theory and Rasch approaches. The PRWE conformed for measuring clinical outcome in distal radius fractures: a core set of to many of the fit expectations of the Rasch model, but domains for standardized reporting in clinical practice and research. Arch not all. With modifications, we were able to fit all the Orthop and Trauma Surg. 2014;134:197–205. items to the Rasch model. 3. DeVellis RF. Classical test theory. Med Care. 2006;44:50–9. 4. Prieto L, Alonso J, Lamarca R. Classical test theory versus Rasch analysis for Additional file quality of life questionnaire reduction. Health Qual Life Outcomes. 2003;1:27. 5. Sica N, Chachamovich E, Pio M, Fleck DA, Tennant A. An introduction to Additional file 1: Patient Rated Wrist Evaluation. (DOCX 18 kb) Rasch analysis for psychiatric practice and research. J Psychiatr Res. 2013;47: 141–8. Abbreviations 6. Pallant JF, Tennant A. An introduction to the Rasch measurement model: an ANOVA: Analysis of Variance; CTT: Classical Test Theory; DASH: Disabilities of example using the hospital anxiety and depression scale (HADS). Br J Clin the Arm, Shoulder, and Hand; DIF: Differential Item Functioning; DRF: Distal Psychol. 2007;46:1–18. Radius Fracture; IRT: Item Response Theory; LR: Likelihood Ratio; MVA: Motor 7. Reckase MD. The past and future of multidimensional item response theory. Vehicle Accidents; PCA: Principal Components Analysis; PREE: Patient-Rated Appl Psychol Meas. 1997;21:25–36. Elbow Evaluation; PROM: Patient-Reported Outcome Measure; PRWE: The 8. Nilsson ÅL, Tennant A. Past and present issues in Rasch analysis: the Patient-Rated Wrist Evaluation; PRWHE: Patient-Rated Wrist and Hand functional independence measure (FIM TM) revisited. J Rehabil Med. 2011; Evaluation; PSI: Person Separation Index; RA: Rasch Analysis 43:884–91. 9. Tennant A, McKenna SP, Hagell P. Application of Rasch analysis in the Acknowledgements development and application of quality of life instruments. Value Heal. Not applicable 2004;7:22–6. 10. Wright BD, Masters GN. Rating scale analysis. Rasch measurement. 5835 S. Funding Chicago: MESA Press; 1982. No funding were received for this study 11. MacDermid JC. Development of a scale for patient rating of wrist pain and disability. J Hand Ther. 1996;9:178–83. Availability of data and materials 12. Taylor J, Kersten P. The patient-rated wrist and hand Evaluation : a systematic Please contact author for data requests review of its validity and reliability. NZ J Physiother. 2014;42:141–7. 13. MacDermid JC. The Patient-Rated Wrist Evaluation (PRWE) User Manual. Authors’ contributions https://srs-mcmaster.ca/wp-content/uploads/2015/05/English-PRWE-User- JM and JV collected the data, JV and TP helped in analyzing the data, DV Manual.pdf. Accessed 20 Jan 2018. and RG helped to draft the manuscript. SE participated in the design of the 14. Mehta SP, MacDermid JC, Richardson J, MacIntyre NJ, Grewal R. A study and performed the statistical analysis and draft the manuscript. All systematic review of the measurement properties of the patient-rated wrist authors read and approved the final manuscript. evaluation. J Orthop Sports Phys Ther. 2015;45:289–98. 15. MacDermid JC, Tottenham V. Responsiveness of the disability of the arm, Ethics approval and consent to participate shoulder, and hand (DASH) and patient-rated wrist/hand evaluation (PRWHE) Not applicable in evaluating change after hand therapy. J Hand Ther. 2004;17:18–23. 16. Packham T, MacDermid JC. Measurement properties of the patient-rated Consent for publication wrist and hand evaluation: Rasch analysis of responses from a traumatic Not applicable hand injury population. J Hand Ther. 2013;26:216–24. 17. Butler JR, Baldwin ML, Johnson WG. Managing work disability: why first Competing interests return to work is not a measure of success. ILR Review. 1995;48:452–69. The authors declare that they have no competing interests 18. Mackenzie EJ, Shapiro SAM, Smith RT, Siegel JH, Moody M, Pitt A. Factors influencing return to work following hospitalization for traumatic injury. Am J Public Heal. 1987;77:329–34. 19. Shames J, Treger I, Ring H, Giaquinto S. Return to work following traumatic brain injury: trends and challenges. Disabil Rehabil. 2007;29:1387–95. 20. Chang WC, Chan C. Rasch analysis for outcomes measures: some methodological considerations. Arch Phys Med Rehabil. 1995;76:934–9. 21. Hagell P, Westergren A. Sample size and statistical conclusions from tests of fit to the Rasch model according to the Rasch unidimensional measurement model (RUMM) program in health outcome measurement. J Appl Meas. 2016;17:416–31. 22. Andrich D, Luo G. Sb. RUMM 2030. RUMM Laboratory Pty Ltd: PERTH; 2010.

Esakki et al. Archives of Physiotherapy (2018) 8:5 Page 11 of 11 23. Suhr DD. Exploratory or confirmatory factor analysis. http://www2.sas.com/ proceedings/sugi31/200-31.pdf. Accessed 20 Jan 2018. 24. Masters G. A Rasch model for partial credit scoring. Psychometrika. 1982;47: 149–74. 25. Armstrong RA. When to use the Bonferroni correction. Ophthal Physiol Opt. 2014;34:502–8. 26. Walton DM, Wideman TH, Sullivan MJL. A Rasch analysis of the pain catastrophizing scale supports its use as an interval-level measure. Clin J Pain. 2013;29:499–506. 27. Parkitny L, Henry J, Walton D, Oliveira L, Costa P, Refshauge KM. Rasch analysis supports the use of the depression, anxiety, and stress scales to measure mood in groups but not in individuals with chronic low back pain. J Clin Epidemiol. 2012;65:189–98. 28. Smith EV Jr. Detecting and evaluating the impact of multidimensionality using item fit statistics and principal component analysis of residuals. J Appl Meas. 2002;3:205–31. 29. Andrich D, Kreiner S. Quantifying response dependence between two dichotomous items using the Rasch model. Appl Psychol Meas. 2010;31:181–92. 30. Tennant A, Pallant J. Introductory Rasch analysis: a workbook. Leeds, UK: Department of Rehabilitation Medicine, University of Leeds; 2011. 31. Linacre JM, Wright BD. Chi-square fit statistic. Rasch Measure Trans. 1994;8:360. 32. Wright BD. Interpreting reliabilities. Rasch Meas Trans. 1998;11:602. 33. Vincent JI, MacDermid JC, King GJW, Grewal R. Rasch analysis of the patient- rated elbow evaluation questionnaire. Health Qual Life Outcomes. 2015;13:84. 34. Bond T, Fox C. Applying the Rasch model: fundamental measurement in the human sciences. 3rd ed. Mahwah: Routledge; 2015. 35. Lautenbacher S, Kunz M, Strate P, Nielsen J, Arendt-Nielsen L. Age effects on pain thresholds, temporal summation and spatial summation of heat and pressure pain. Pain. 2005;115:410–8. 36. Cole LJ, Farrell MJ, Gibson SJ, Egan GF. Age-related differences in pain sensitivity and regional brain activity evoked by noxious pressure. Neurobiol Aging. 2010;31:495–503. 37. Alexander NB, Galecki AT, Grenier ML, Nyquist LV, Hofmeyer MR, Grunawalt JC, Medell JL, Fry-Welch D. Task-specific resistance training to improve the ability of activities of daily living-impaired older adults to rise from a bed and from a chair. J Am Geriatr Soc. 2001;49:1418–27. 38. Kenny GP, Yardley JE, Martineau L. Physical work capacity in older adults: implications for the aging worker. Am J Ind Med. 2008;51:610–25. 39. Bianchi SM, Milkie MA, Sayer LC, Robinson JP. Is anyone doing the housework? Trends in the gender division of household labor. Soc Forces. 2002;79:191–228. 40. Dzaja I, MacDermid JC, Roth J, Grewal R. Functional outcomes and cost estimation for extra-articular and simple intra-articular distal radius fractures treated with open reduction and internal fixation versus closed reduction and percutaneous Kirschner wire fixation. Can J Surg. 2013;56:378–84. 41. Koo KO, Tan DM, Chong AK. Distal radius fractures: an epidemiological review. Orthop Surg. 2013;5:209–13. Submit your next manuscript to BioMed Central and we will help you at every step: • We accept pre-submission inquiries • Our selector tool helps you to find the most relevant journal • We provide round the clock customer support • Convenient online submission • Thorough peer review • Inclusion in PubMed and all major indexing services • Maximum visibility for your research Submit your manuscript at www.biomedcentral.com/submit

Seebacher et al. Archives of Physiotherapy (2018) 8:6 https://doi.org/10.1186/s40945-018-0045-0 RESEARCH ARTICLE Open Access Exploring cued and non-cued motor imagery interventions in people with multiple sclerosis: a randomised feasibility trial and reliability study Barbara Seebacher1* , Raija Kuisma1, Angela Glynn1 and Thomas Berger2 Abstract Background: Motor imagery (MI) is increasingly used in neurorehabilitation to facilitate motor performance. Our previous study results demonstrated significantly improved walking after rhythmic-cued MI in people with multiple sclerosis (pwMS). The present feasibility study was aimed to obtain preliminary information of changes in walking, fatigue, quality of life (QoL) and MI ability following cued and non-cued MI in pwMS. The study further investigated the feasibility of a larger study and examined the reliability of a two-dimensional gait analysis system. Methods: At the MS-Clinic, Department of Neurology, Medical University of Innsbruck, Austria, 15 adult pwMS (1.5–4.5 on the Expanded Disability Status Scale, 13 females) were randomised to one of three groups: 24 sessions of 17 min of MI with music and verbal cueing (MVMI), with music alone (MMI), or non-cued (MI). Descriptive statistics were reported for all outcomes. Primary outcomes were walking speed (Timed 25-Foot Walk) and walking distance (6-Minute Walk Test). Secondary outcomes were recruitment rate, retention, adherence, acceptability, adverse events, MI ability (Kinaesthetic and Visual Imagery Questionnaire, Time-Dependent MI test), fatigue (Modified Fatigue Impact Scale) and QoL (Multiple Sclerosis Impact Scale-29). The reliability of a gait analysis system used to assess gait synchronisation with music beat was tested. Results: Participants showed adequate MI abilities. Post-intervention, improvements in walking speed, walking distance, fatigue, QoL and MI ability were observed in all groups. Success of the feasibility criteria was demonstrated by recruitment and retention rates of 8.6% (95% confidence interval, CI 5.2, 13.8%) and 100% (95% CI 76.4, 100%), which exceeded the target rates of 5.7% and 80%. Additionally, the 83% (95% CI 0.42, 0.99) adherence rate surpassed the 67% target rate. Intra- rater reliability analysis of the gait measurement instruments demonstrated excellent Intra-Class Correlation coefficients for step length of 0.978 (95% CI 0.973, 0.982) and step time of 0.880 (95% CI 0.855, 0.902). Conclusion: Results from our study suggest that cued and non-cued MI are valuable interventions in pwMS who were able to imagine movements. A larger study appears feasible, however, substantial improvements to the methods are required such as stratified randomisation using a computer-generated sequence and blinding of the assessors. Trial registration: ISRCTN ISRCTN92351899. Registered 10 December 2015. Keywords: Multiple sclerosis, Physiotherapy, Motor imagery, Rhythmic cueing, Walking, Fatigue, Quality of life, Motor imagery ability, Feasibility, Reliability two-dimensional gait analysis * Correspondence: [email protected] 1School of Health Sciences, University of Brighton, Robert Dodd Building, 49 Darley Road, Eastbourne BN20 7UR, UK Full list of author information is available at the end of the article © The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.

Seebacher et al. Archives of Physiotherapy (2018) 8:6 Page 2 of 19 Background effective than metronome-cued MI in improving fatigue and quality of life (QoL). However, we did not measure Introduction the participants’ ability to imagine movements or their ac- Multiple Sclerosis (MS) is a chronic disease of the tual gait synchronisation with music beat. We also did not central nervous system, which leads to destruction of include a non-cued MI group and thus could not compare the protective myelin nerve sheaths and accumulating between the effects of cued and non-cued MI. In other disability. People with MS have impairment in motor, words, there are various areas of uncertainty in knowledge sensory, visual and other body systems [1]. Evidence concerning the mechanisms of the rhythmic-cueing and shows that 40–80% of patients report fatigue [2, 3] MI interventions in persons with MS, which have not yet which, together with walking impairment, contributes to been addressed. For example, the contributions from the a limitation in their walking endurance [4] and MI practice and the music and verbal cueing, respectively, independence in daily life activities [5]. Therefore, it is to the functional improvements have not been investi- considered essential to develop novel rehabilitation stra- gated. To our knowledge, only one study explored the ef- tegies to improve walking and fatigue, which negatively fect of five weeks of MI practice on fatigue, walking speed impact walking abilities. and QoL in 20 people with MS and mild to moderate disability [20]. Their findings showed that fatigue and QoL Motor imagery significantly improved, but the walking speed improve- Apart from physical training, motor imagery (MI) has ment was not significant. increasingly been used in neurorehabilitation to enhance motor relearning in people with neurological disorders Rhythmic auditory stimulation has been successfully [6]. Defined as mental rehearsal of a movement, the used to improve walking performance [21]. Hence, we motor action is not executed [7, 8]. Motor areas of the hypothesised that rhythmic auditory cueing of the MI brain responsible for movements’ physical execution are would serve as an external timekeeper to the imagined activated, although to a lesser degree, during imagined steps. Thereby, like in real walking [22], imagery of walk- movements [6–9]. Both real and imagined movements ing along a regular and hence predictable auditory are associated with similar motor preparation processes rhythm might be processed quite accurately, mainly [7, 8]. In addition, imagined and actual movements have because the imagined steps are produced slightly ahead previously been shown to share similar temporal profiles of the cue [21]. This coupling process between external [8, 10]. Mental chronometry, that is temporal con- rhythm and body adjustment is referred to as rhythmic gruence, is the similar time duration of imagined and entrainment which, for example, causes the synchronisa- real motor actions [10, 11]. Consequently, the capacity tion of gait with the tempo of music beat [21, 22]. to preserve the temporal features of a movement during Against the background of these findings, we suggested MI has been associated with a person’s MI ability [11, that rhythmic cueing might induce entrainment and en- 12]. Mental chronometry studies showed that the MI ac- hance the MI ability in our participants. To the best of curacy and its temporal organisation were impaired in our knowledge so far, no study has evaluated the mecha- participants with MS versus controls; these deficits in nisms of differently cued and non-cued MI for walking MI ability were associated with cognitive impairment, rehabilitation in people with MS. Thus, we plan to con- but were independent from motor functioning [13–16]. duct a randomised controlled trial (RCT) to investigate Reduced MI accuracy and timing are also related to de- the effects and mechanisms of differently cued and non- pression [14]. However, rhythmic auditory cueing has cued MI on walking, fatigue, QoL, MI ability and gait been found to promote the MI ability in people with synchronisation with music beat in people with MS. The MS, by optimising their MI duration and movement present feasibility study was used to gather preliminary amplitudes during an upper limb task [17]. results and examine the feasibility of the aforesaid RCT. Rationale Aims To our knowledge, only one study explored the effects of The aims of this feasibility study were: rhythmic-cued MI in 15 people with stroke [18]. Kim et al. observed an improvement in walking performance after 1) to explore the success of the feasibility criteria: kinaesthetic rhythmic-cued MI practice, compared to vis- recruitment and retention rates and adherence rate; ual MI practice without auditory cueing [18]. Similarly, re- sults from our previous study showed significant 2) to explore the safety of the interventions, adverse improvements in walking performance after four weeks of events and participant acceptability of the music- and metronome-cued kinaesthetic MI practice interventions; with additional verbal cueing in people with MS and mild to moderate disability [19]. Music-cued MI was more 3) to obtain preliminary information on change in walking speed and walking distance induced by the two types of music-cued and non-cued MI;

Seebacher et al. Archives of Physiotherapy (2018) 8:6 Page 3 of 19 4) to acquire preliminary information on change in opportunity to ask questions. They had at least twenty- fatigue produced by the two types of music-cued four hours to consider their participation and to consult and non-cued MI; family members or others. After one week, potential par- ticipants who agreed to telephone contact, were called 5) to collect preliminary data on the change in QoL by the researcher asking if they would like to participate generated by the two types of music-cued and non- in the study. Written informed consent was obtained by cued MI; the researcher. Participants were then randomly allo- cated using a 1:1:1 ratio, and they were asked to draw a 6) to evaluate the baseline MI ability and obtain sealed envelope, prepared by a researcher not involved preliminary information on its change; in this study, with numbers ´1′, ´2′ or ´3′ which corre- sponded with the three study groups. The randomisation 7) to assess the reliability and repeatability of the was restricted insofar as only 15 envelopes were pro- quantitative gait measurement instruments used to vided. Using this procedure, the physiotherapist involved assess synchronisation with music beat. in the recruitment could not influence group allocation. Methods Intervention The intervention of this study consisted of music- and Study design and location verbally-cued MI (MVMI group), music-cued MI (MMI A three-group parallel randomised controlled single- group) and non-cued MI (MI group). After the rando- centre feasibility trial was conducted. The adjunct CON- misation and prior to the intervention, study participants SORT checklist for pilot and feasibility studies is available were familiarised with rhythmic-cued MI or non-cued as Additional file 1 [23]. All measurements were taken at MI by the researcher (BS), as suggested in previous stu- the MS Clinic of the Clinical Department of Neurology, dies [26, 27]. These authors have proposed the PET- Medical University of Innsbruck, Austria. TLEP approach to MI in neurorehabilitation, involving the “Physical, Environmental, Task, Timing, Learning, Participants and recruitment Emotional, and Perspective” components of MI [28]. Recruitment used unselected consecutive sampling and The PETTLEP checklist is based on neuroscientific fin- restricted randomisation from 3rd March to 14th April dings and has been developed by Holmes and Collins 2016. Due to the piloting character of this study and (2001, page 60) for performance improvements in time constraints, 15 participants were recruited into athletes. The PETTLEP elements concern the physical, three groups. The MS-Clinic currently cares for approxi- or bodily, position of the practitioner including arousal, mately 2500 people with MS, 339 were screened during the imagined environment, the imagined task, the MI recruitment, of those 174 were eligible. A CONSORT timing, the learning or changes induced by the MI, the flow diagram for pilot and feasibility studies [23] is emotions or affective states, which are related to the MI shown in Fig. 1. task, and the MI perspective. The PETTLEP ideas were applied to the context of the present study to enhance Inclusion Criteria were: people with any MS phenotype the effectiveness of the intervention. according to McDonald’s criteria [24], mild to moderate disability (Expanded Disability Status Scale, EDSS 1.5 to Firstly, participants were informed about the concept 4.5) [25], aged 18 years or over, any ethnicity, German of (cued) MI and its application in sports and neuroreh- speaking (questionnaires, MI familiarisation, instructions). abilitation. Secondly, participants also learned about the different perspectives (internal and external) and modes Exclusion Criteria were: concomitant diseases which (visual, kinaesthetic) which can be adopted during MI may affect rhythmic cued MI or walking (e.g. orthopaedic [29, 30]. In the visual mode, the persons imagine disorders, untreated hearing loss), a relapse of MS within watching themselves (internal perspective) or other the last three months, recent start of physiotherapy treat- people (external perspective) moving, and in the kinaes- ment or change of medication which is known to affect thetic mode, the persons experience or feel themselves walking within the last two months, known pregnancy, moving [31]. When employed in clinical practice, people overt symptoms or signs of depression or cognitive dys- use both perspectives when asked to imagine a move- function diagnosed and documented by the Innsbruck MS ment. Therefore, participants had the opportunity to try Clinic. A relapse during the intervention period would them out for themselves and become aware of their have led to the exclusion from the study. preferred mode or perspective. The researcher placed emphasis on internal, kinaesthetic MI, which was Eligible individuals were identified during their usual adopted for the intervention and MI assessment of this visit to the Outpatient MS Clinic by their neurologist study and the internal, visual MI which was only used (TB). After that, they received the recruitment flyer from the reception staff. They also obtained the participant information sheet and detailed verbal information on the study from the researcher who is a physiotherapist (BS). Oral and written information was presented in com- prehensible lay language. Eligible individuals had the

Seebacher et al. Archives of Physiotherapy (2018) 8:6 Page 4 of 19 Fig. 1 CONSORT flow diagram for pilot and feasibility studies for the MI ability testing. During the training, partici- distance they had reached and indicate when they imag- pants were asked for MI content characteristics such as ined reaching the same point. The time was measured, the mode and perspective they were using, for the envi- and feedback was given to the participants. If desired, ronment or for movement aspects they were imagining they could repeat the imaginative task as many times as (learning). Moreover, the duration of the actual and they wanted. The intervention is presented in Table 1 imagined walking performance was compared to moni- and was based on the Template for Intervention tor the mental process [26]. Participants were asked to Description and Replication (TIDieR template) [32]. walk a six-metre distance along the marked hallway while the time was measured. After that, they were Selection of the music style, beat patterns and tempo asked to imagine themselves walking the same six-metre were based on relevant literature in the field [21]. Music beat in the MVMI and MMI groups was in 2/4 or 4/4 m

Seebacher et al. Archives of Physiotherapy (2018) 8:6 Page 5 of 19 Table 1 Intervention chart ITEM NO ITEM DESCRIPTION 1 BRIEF NAME MVMI Group MMI Group MI Group Music- and verbally cued MI Music-cued MI Non-cued MI 2 WHY PETTLEP approach to MI (Holmes and Collins 2001) 3 WHAT - Study CDs or dropbox link including the audio mix and download to smartphone, laptop, tablet or MP3-player MATERIALS - 4 sessions on each CD, one for each week - Headphones or earphones could be used if desired Study CD - Kinaesthetic MI instructions - Kinaesthetic MI instructions - Kinaesthetic MI instructions Content - Instrumental music in 2/4 or 4/4 m - Instrumental music in 2/4 or 4/ 4m - Emphasis of every first beat, or every first and third beat by rhythmic verbal cues (e.g. “toe-off” or “step- step”) For example, music titles used in week two were: Unheilig, Der Berg (Intro), 82 bpm; Black, No cueing Wonderful Life, 106 bpm; Malcolm Arnold, The River Kwai March, 120 bpm; Uriah Heep, Lady in black, 86 bpm; Abba, Dancing queen, 101 bpm; Toto, Africa, 100 bpm; DJ Bobo, I’m living to love you, 110 bpm; Katy Perry, Firework, 120 bpm. Availability After completion of the main study, a download of the 3 study CDs will be available upon request from the corresponding of CDs author. 4 WHAT - MI introduction, familiarisation and training: in lay language; description of the concept of MI; its application and effects PROCEDURES in sports and neurorehabilitation; principles of neuroplasticity; MI perspectives (internal and external) and modes (visual, kinaesthetic). - Measurement of actual and imagined walking duration over a 6-m distance to monitor the mental process - Performance feedback for participants and repeated training if desired - Weekly phone calls for support and adherence reports - Additional introduction to rhythmic auditory stimulation plus its use in neurorehabilitation - Rhythmic-cued MI familiarisation PETTLEP Elements Position - Practise at any time of the day when alert (Physical) - Seated in an upright body position - Shoulders relaxed - Avoid tightening the muscles or moving - Eyes closed - Normal breathing Environment - Practice in a quiet place at home - Imagine walking indoors (long hallway similar to that in the MS Clinic) and walking outdoors (on a straight and familiar path) Tasks - Take long strides - Take giant strides - Roll your feet on the ground and feel your body weight on your soles - Touch the ground with your heels first - Raise the front of your feet - Raise your knees - Pace - Place/feel your weight on your feet - Place/feel your weight on your legs - Stamp your feet while walking - Walk effortlessly, almost as if you were floating - Walk forcefully and energetically as if you were an athlete

Seebacher et al. Archives of Physiotherapy (2018) 8:6 Page 6 of 19 Table 1 Intervention chart (Continued) ITEM NO ITEM DESCRIPTION - March as if you were in the army - Walk in an extremely upright posture such as when balancing a sachet, filled with rice, on your head - Feel the swinging of your arms while walking - Feel the swinging of your legs while walking Timing External timing was provided: “imagine yourself External timing was provided: Timing was internal and depended on walking in time with the music and verbal cues” “imagine yourself walking in time the tempo and intensity of the with the music” walking tasks. - Tempo (cadence) was between 80 and 120 steps/min - Slow, medium and fast music pieces alternated, with a general progression in the tempo - The imagined walking tempo was consistent with the music beat at 80–120 bpm. Learning - See familiarisation - Additionally, weekly phone call support was provided Emotion MI instructions included motivational and arousal enhancing aspects (e.g. walk forcefully and energetically as if you were an athlete). See instructions under Tasks. Motivational instrumental music was used with the MI Perspective Kinaesthetic MI from an internal, first-person perspective 5 WHO The intervention including the preparation of the CDs was provided by the researcher (BS), an experienced physiotherapist with PROVIDED 11 years of musical training. 6 HOW - MI introduction, familiarisation and training: individually or in small groups (2–3 participants) and depending on the group they were allocated to - Monitoring of mental process: individually - Weekly phone calls: individually 7 WHERE - MI introduction, familiarisation, training and monitoring of mental process: at MS Clinic Innsbruck, Department of Physiotherapy - Cued MI practice: At participants’ homes MI practice: At participants’ homes 8 WHEN AND 17 min, 6 times a week, for 4 weeks HOW MUCH 9 TAILORING Same intervention for all participants Same intervention for all Same intervention for all participants participants 10 No modifications No modifications No modifications MODIFICATIONS 11 HOW WELL - Intervention adherence was assessed using a participant diary and also during weekly phone calls and at post-intervention PLANNED - Recording in excel sheets was performed by the researcher (physiotherapist) who instructed participants 12 HOW WELL The adherence rate was median 5 (range 4, 6) times per week or 83% (95% confidence interval 0.42, 0.99). ACTUAL Abbreviations: MI: motor imagery; bpm: beats per minute and in addition, in the MVMI group, every first beat, or weekly to support them with the MI and as a reminder every first and third beat were stressed and emphasised of the practice. The phone calls were made by the re- by rhythmic verbal cues [21, 22]. Appropriate rhythmical searcher, who introduced the participants to the (cued) sequences were cut and mixed (GarageBand, Apple Inc.) MI practice and explained all the procedures. Questions with the MI instructions of walking. Karaoke music that were asked during the phone calls are added as pieces were selected from a wide range of musical styles, Additional file 2. including rock, pop, folk music, dance, techno and marching music, hard rock and film music. Data collection Demographic (gender, age) and MS disease specific data Participants were asked to practice MI six times a (current EDSS) were extracted from patients’ charts, week, once a day for seventeen minutes over a four week study specific assessment data were collected pre and period. After each week and in all groups, the audio mix post intervention by one physiotherapist who is not a was changed to facilitate adherence and to retain atten- member of the MS Clinic. The German version of all tion with the MI [21], so that four mixes, designed in assessments was used. Baseline and post-intervention the same way, were on one CD. Participants were called

Seebacher et al. Archives of Physiotherapy (2018) 8:6 Page 7 of 19 assessments were performed at the physiotherapy de- meaningful change, or minimal clinically important dif- partment of the MS Clinic, always in the mid-afternoon, ference (MCID), in walking [36]. in view of daytime fluctuations in walking abilities and fatigue. Participants were allowed to rest at any time Walking distance was measured by the 6-Minute Walk during the instructions and assessments. A history of de- Test (6MWT) [37]. The 6MWT was carried out as pression and cognitive dysfunction before onset or its recommended by the American Thoracic Society- worsening or new occurrence after diagnosis of MS was Guidelines. Learmonth et al. (2013) demonstrated that a rigorously asked and documented by the same treating change of 20% and above in walking distance represents neurologist (TB). Clinical definitions of depression and the minimal detectable change (MDC) [38], reflecting cognitive dysfunction were used rather than formal the smallest real difference which exceeds the measure- neuropsychological testing. Depression was defined as a ment error [39, 40]. Other research reported that a state of low mood and loss of activity along with charac- change in walking distance of 20% and above is clinically teristic symptoms such as sadness, anxiety, awkward- meaningful [41]. Based on clinical judgement, a 20% ness, loss of appetite, insomnia, up to suicidal thoughts. change in walking distance is relevant for the patients in Cognitive dysfunction was defined by report and clinical their daily lives. Good to excellent psychometric proper- assessment of characteristic symptoms such as impair- ties of the 6MWT have been demonstrated in an MS ment in orientation, memory, attention, learning, population [37]. 6MWT reference values for healthy language, visuospatial skills, calculating, planning or any women and men aged 20 to 50 are mean (standard other executive function. deviation, SD) 593 ± 57 m and 638 ± 44 m, respectively, with differences depending on height and age [42]. Primary outcomes Females and males with MS and an EDSS below 4.0 Feasibility walked mean (SD) 380.1 ± 156.0 m and 459.5 ± 133.8 m, Feasibility of conducting a full-scale RCT was evaluated. respectively; women and men with an EDSS of up to 6.5 The criteria for feasibility success were: a) a target re- walked 322.2 ± 156.4 m and 362.8 ± 169.2 m, with varia- cruitment rate of 5.7% out of 174 eligible patients (or 10 tions connected to age, cardiorespiratory function and participants per month), b) a target retention rate of balance [43]. 80% and c) a target minimum adherence rate of 67% (4 practice sessions per week out of a maximum of 6). Walking aids for both the T25FW and 6MWT were used if required and were documented and kept consis- tent during the two assessments. Safety, adverse events and acceptability Fatigue Participants were asked to report any adverse event such Fatigue was assessed using the Modified Fatigue Impact as falls, excessive fatigue, psychological distress and/or Scale (MFIS). The MFIS is a modified version of the other safety related occurrences. Severe adverse events Fatigue Impact Scale [44] and one part of the Multiple would have led to early study termination. During the Sclerosis Quality of Life Inventory (MSQLI) [45]. It is a weekly phone calls, participants were asked for their 21 item Likert scale that evaluates, via self-report, the ef- feedback on the study procedures, which were recorded fects of fatigue on physical, cognitive and psychosocial in Excel files and reported. Participant adherence with functioning, with higher numbers indicating greater the interventions was noted by participants in a diary. fatigue. All items can be answered by five categories Adherence and the acceptability of the interventions (never, rarely, sometimes, often and almost always; range were reported narratively. Acceptability referred to 0–4) resulting in a total score from 0 to 84. The MFIS kinaesthetic MI, melodies, beat tempo and the verbal has an excellent reliability and moderate to high validity cueing. Questions that were asked by the researcher are and responsiveness in people with MS [46, 47]. Based on available as additional information; see Additional file 2. previous studies, the cut-off value for MS-related fatigue was set at ≥38 points on the MFIS total score [48]. Secondary outcomes Responsiveness of the MFIS, as expressed by the Walking speed and walking distance smallest detectable change (SDC) is − 16.2 points on the Walking speed was assessed by the Timed 25-Foot Walk total score, − 8.9 points on the physical subscale, − 8.0 (T25FW) [33]. The T25FW is the most commonly re- points on the cognitive subscale and − 2.3 points on the ported short walking test, which has excellent validity, psychosocial subscale [47]. reliability and responsiveness in people with MS [34] and was administered according to instructions in the Quality of life Multiple Sclerosis Functional Composite [35]. There is a QoL was assessed with the Multiple Sclerosis Impact Scale consensus in the literature that a change of 20% and (MSIS-29). The MSIS-29 is a 29-item disease-specific, above in walking speed corresponds to a clinically