รวมเล่ม_PowerPoint_ว.ระบาดวิทยาฯ 2565 (Wanpen)

Steps for conducting a case-control study (retrospective study) Selection samples (case and control) Case group (diseased) Control group (non-diseased) The primary study base is the selection of cases - Representative of the population at risk of becoming a case. from the community by selecting cases or randomly selecting (sampling) cases from all existing patients - Have no health problems as in the community of interest. interested (outcome) The secondary study base, also known as the - Patients with characteristics most similar to the studied patients. case-defined study base, is the selection of cases from patients who come to health care facilities. or hospital.

Steps for conducting a case-control study (retrospective study) Data Collection 3.1 Study data 3.2 Study data 3.3 Observing, 3.4 Observing, 3.5 Observing, 3.6 Observing, from medical from other questioning, questioning, questioning, questioning, records about interviewing or interviewing or interviewing or interviewing or past illnesses/ records related surveying surveying exploring other surveying treatments to past illness/ information information about health that have been about the risks behaviors treatment. about recorded. family psychosocial environment factors

Data Analysis and Interpretation Exposure (Risk factor) Outcomes Total Prevalence Diseased Non-diseased Exposed a+b a / (a + c) Non-exposed ab c+d c / (a + c) cd a+b+c+d Total a+c b+d - The probability of exposed risk factor in case (P ex) = a / (a + c) The probability of non-exposed risk factor in case (P non) = c / (a + c) a / (a + c) Odds Ratio = ������ = c / (a + c) ������ Odds Ratio (OR) Interpretation ➢ A Odds Ratio of 1.0 indicates identical risk among the two groups. ➢ A Odds Ratio greater than 1.0 indicates an increased risk for the group in the numerator, usually the exposed group. ➢ A Odds Ratio less than 1.0 indicates a decreased risk for the exposed group, indicating that perhaps exposure actually protects against disease occurrence.

Steps for conducting a case-control study (retrospective study) Conclusion - The results obtained are only odds ratios. - A true relationship or relative risk cannot be concluded because this type of study was unable to determine the incidence rate of the exposed group and the non-exposed group.

The bias of a case-control study (retrospective study) 1. Recall bias - Some factors may have been forgotten - Incomplete medical records - The difference in the number of factors between case and control - Unclear Information on past exposures 2. Selection bias Often caused by the improper selection of the control group and the lack of cooperation from the control group in providing information.

Advantages & Disadvantages Advantages Disadvantages ➢ Cheaper ➢ Retrospective / more prone to ➢ Quicker / easier to conduct bias ➢ Good for diseases with long ➢ Can only assess one latency periods outcome/disease ➢ Can assess multiple exposures ➢ Good for rare diseases ➢ Cannot establish risk ➢ Cannot establish the incidence

Summary ➢ Case-control study is a retrospective design. ➢ Case-control studies must clearly define two groups at the start ( one with the outcome/disease and one without the outcome/disease) ➢ This design focused on looking back to assess whether there is a statistically significant difference in the rates of exposure to a defined risk factor between the groups.

Email: [email protected]



Chapter Contents Epidemiological Study Designs 2 Cohort study or Prospective study

Prospective or cohort study • A prospective or cohort study is • To study and test the Keywords an analytical epidemiological relationship between ➢ An analytical study, a longitudinal study that the factors that are follows over time a group of expected to cause epidemiological study similar individuals (cohorts) who disease and disease ➢ Observation differ with respect to certain progression. ➢ Follow-up of healthy exposures (factors) under study, to determine how these samples exposures (factors) affect rates of ➢ Starting from free of a certain outcome (disease). disease Definition Objectives ➢ Follow forward until the expected disease (outcome) arising

The nature of the cohort study Question of a Initially, the study The follow-up was used Disease / study population was for a long enough time to Outcome • Exposure/Factor divided into measure whether the ? individuals who did suspected factor was not have the disease. causing the disease. • Outcome/Disease • Exposed group • Incidence of the observe • •follow-up The pathogenesis follow-up exposure group of that factor: • (disease) no disease • : Induction period • : Incubation period • Incidence of the follow-up non-exposure • Non-exposed follow-up group group (disease)

The nature of the cohort study test the relationship between the exposure (factor) and disease observation without manipulation Study group follow-up Disease (+) No disease (-) Exposed to certain exposure (+) follow-up Disease (+) Comparison group No disease (-) Non-exposed to certain exposure (-) follow-up Initially the Study End of Study

Data collection and measurement of exposures and outcomes Measurement of Exposures Measurement of outcomes • Questionnaire • Questionnaire • Interview form • Biochemical or laboratory tests • Biochemical or laboratory tests • Physical examination • Physical examination • Special examination • Special examination • - x-ray • - x-ray • - EKG • - EKG • Death information • Surveying the environment • - death certificate • - Department of Provincial Administration • Illness information • - medical records • - disease registration record • - patient report (directly)

Data collection and calculation Study design Time Future Cohort study or Past Present Collecting personal Prospective study data and outcome - Collecting personal data and natural exposure Initially the Study End of Study Exposure Disease (+) No disease (-) Total Calculation outcome I ex = a / (a+b) Exposed (+) a b a+b I non = c / (c+d) Non-exposed (+) c d c+d Total a + c b+d a+b+c+d Incidence rate of exposure group (I ex) = a / (a + b) Calculation Incidence rate of non-exposure group (I non) = c / (c + d) Relative Risk or Risk Ratio = (I ex) / (I non) or ������ / (������ + ������) ������ / (������ + ������)

Outcomes: Incidence Rate of Both Groups 1. Cumulative • = จำนวนผทู้ ่ีเร่มิ ป่ วยในชว่ งระยะเวลำท่ีทำกำรศกึ ษำ Incidence (CI): จำนวนประชำกรเม่ือเรม่ิ ตน้ กำรศกึ ษำ∗ 2. Incidence Density (ID) • Incidence Density (ID) = จำนวน จำนวนผทู้ ่ีเร่มิ ป่วยในระยะเวลำท่ีทำกำรศกึ ษำ × ������ or Person-time Incidence person−time at risk∗ ตลอดระยะเวลำท่ีทำกำรศกึ ษำ * personS-timuemat risko=fผลtรวhมขeองระpยะeเวลrาsท่ีสoงั เกnต (-หtรอื iรmะยะเeวลาทoีส่ ัมfผสั tปจัhจยั eเสย่ี aง) ขtอ-งผrูถ้ iกู sสงั kเกตแต่ละราย population (Total person-time at risk)

Person-time at risk calculation…1 A study group: exposed group Example Investigators enrolled 500 men who were smokers in a study group and followed them annually for four years to determine the incidence rate of hypertension. After one year, one had a new diagnosis of hypertension, but 50 had been lost to follow-up. After two years, two had a new diagnosis of hypertension, and another 10 had been lost to follow-up. After three years, another five had new diagnoses of hypertension, and 20 had been lost to follow-up. After four years, another 6 had new diagnoses with hypertension, and 30 more had been lost to follow-up.

Person-time at risk calculation…2 Person-time at risk = Number Persons × Period of Time ENuxmabmerple Duration (Year) Lost FU. Incidence Person-time at risk Y 0 = 500 0 50 - 50x0 = 0 Y1 = 450 1 10 1 10x1 = 20 Y2 = 440 2 20 2 20x2 = 40 Y3 = 410 3 30 5 30x3 = 90 Y4 = 390 4 - 6 390x4 = 1,560 Numerator = number of new cases of hypertension = 1 + 2 + 5 + 6 = 14 Denominator = Person-time at risk = 0+20 + 40 + 90 + 1,560 = 1,710 Incidence Density (ID) or Person-time Incidence of Exposed Group = 14/1,710 = .008 / person-year (Or 8/1,000 person-years)

Person-time at risk calculation…3 Comparison group: non-exposed Example Investigators enrolled 500 men who were non-smokers in a comparison group and followed them annually for four years to determine the incidence rate of hypertension. After one year, no one had a new diagnosis of hypertension, but 10 had been lost to follow- up. After two years, one had a new diagnosis of hypertension, and another 20 had been lost to follow-up. After three years, another one had new diagnoses of hypertension, and 5 had been lost to follow-up. After four years, another 2 had new diagnoses with hypertension, and 50 more had been lost to follow-up.

Person-time at risk calculation…4 Person-time at risk = Number Persons × Period of Time ENuxmabmerple Duration (Year) Lost FU. Incidence Person-time at risk Y 0 = 500 0 10 - 10x0 = 0 Y1 = 450 1 20 0 20x1 = 20 Y2 = 440 2 5 1 5x2 = 10 Y3 = 410 3 50 1 50x3 = 150 Y4 = 390 4 - 2 425x4 = 1,700 Numerator = number of new cases of hypertension = 0 + 1 + 1 + 2 = 4 Denominator = Person-time at risk = 0+20 + 10 + 150 + 1,700 = 1,880 Incidence Density (ID) or Person-time Incidence of non-exposed Group = 4/1,880 = .002 / person-year (Or 2/1,000 person-years)

Relative RisRkeolar tRivisek RRaistkio((คR่าRค)ว(าคมา่ คเวสาีย่ มงเสสี่ยมั งพสัมัทพธัท)์ ธ)์ Relative Risk (RR) = ������������������������ ������������ ������������������������������������������ (������������������������������������������������������ ������������������������������������������������������������) ������������ ������������������������������������������ ������������������������������ ������������������������ ������������ ������������������������������������������ (������������������������������������������������������ ������������������������������������������������������������) ������������ ������������������������������������������������������������ ������������������������������ (������������������−������������������������������������������) ������������������������ ������������ ������������������������������������������ ������������ ������������������������������������������ ������������������������������ (I ex) = 14/1,710 = .008 (8/1,000 person-years) ������������������������ ������������ ������������������������������������������ ������������ ������������������−������������������������������������������ (I non) = 4/1,880 = .002 (2/1,000 person-years) Relative Risk or Risk Ratio = (I ex) / (I non) , = .008/.002 x 1,000 = 4 The RR of 4.0 means that people who smoke are 4 times more likely to develop hypertension than those who do not smoke. Relative Risk or Risk Ratio (RR) Interpretation ➢ A risk ratio of 1.0 indicates identical risk among the two groups. ➢ A risk ratio greater than 1.0 indicates an increased risk for the group in the numerator, usually the exposed group. ➢ A risk ratio less than 1.0 indicates a decreased risk for the exposed group, indicating that perhaps exposure actually protects against disease occurrence.

The biases of the cohort study design 1. Selection bias 2. Information bias 3. Confounding bias 1.1 Choosing an 2.1 Misclassification of 3.1 Bias sauriscihngadsuaringgef,osllmowo-ukpin, gcan inappropriate study exposures or risk factors be solved by clarification and group can be solved by understanding of the selected sampling 2.2 Inaccurate classification of samples and by emphasizing the outcome or disease importance of cooperation during 1.2 Inappropriate participation in the study. selection of 2.3 Lost of follow-up 3.2 Study design and data collection comparison groups can biases can be solved by assigning be solved by using an both groups to have the same internal comparison interference factor characteristics. group. These biases can be solved by good planning



➢ This is a prospective follow-up analytical study without any action on the study interests. ➢ Typically examines multiple health effects of exposure after observation of a group of people with factors and a group of people without factors and followed for a period of time to see if the “incidence rate” of those people with factor factors differs from the comparison group that does not have the factors studied. ➢ Follow-up must be long enough to measure whether the hypothesized factor is pathogenic, and must be at least equal to the induction period of the factor or the incubation period.

Email: [email protected]



Principle of • Conducted at a single point in time or over a short period of time (snapshot of the population) • Exposure status and disease status are measured at one point in time or over a period. • Providing the prevalence of a disease • Comparison of prevalence among exposed and non- exposed (analytic cross-sectional study) • It does not involve manipulating variables

Formulating The magnitude of effect / Hypothesis & Point estimation Comparing the prevalence of exposed Precision / Interval estimation / and non-exposed Statistical significance groups Parameter Target Population

Process of 2. Study the whole group 3. Randomly select a suspected to be a factor certain number of sample 1. Study by observing the causing disease or health phenomenon of illness or sizes. problems. health problem at a particular point of time. 5. Comparing disease occurrence 4. Measure the factors between the exposed and non- that are expected to exposed group, Is it different or not? affect the occurrence of the disease, and measure the prevalence of existing disease Note: Analytical cross-sectional studies can only obtain correlations from studies at the point of interest.

ObjectivesObjectives of the - To examine the presence or absence of an outcome and the presence or absence of exposure at a specific point of time. - To provide information about the prevalence of disease or health problems of a current population. - To measure the association between exposure and a disease, condition, or outcome within a defined population.

Method of study - random sampling of samples Simple Random Stratified Random Cluster Sampling Systematic Multi-stage Sampling Sampling Sampling Random Sampling - drawing (lottery) - random number table (random table) - spinning wheel -artificial intelligence Source: Source: Source: Source: Source: https://research- https://www.simplypsychology.org/s https://www.scribbr.com/methodol https://www.voxco.com/blog/ https://www.qualtrics.com/experi methodology.net/sampling-in- tratified-random-sampling.html ogy/stratified-sampling/ stratified-sampling-vs-cluster- encemanagement/research/syste primary-data-collection/multi- sampling/ matic-random-sampling/ stage-sampling/

Data Collection 1. Interview, 2. Physical 3. Laboratory Tests 4. Medical Questionnaires, Examination or or Special Reports, Assessment forms Observation Documents Examinations

Measurement of disease frequency (Measure of morbidity..การป่วย..) 1. Prevalence rate Odds ratio Interpretation ➢ A odds ratio of 1.0 indicates identical It is the proportion of persons in a population who have a particular risk among the two groups. disease or attribute at a specified ➢ A odds ratio greater than 1.0 point in time or over a specified period of time. indicates an increased risk for the group in the numerator, usually the = XK exposed group. ➢ A odds ratio less than 1.0 indicates a The value of ������ (constant) can be 100, 1000, decreased risk for the exposed group, 10,000 indicating that perhaps exposure actually protects against disease occurrence.

Data Analysis: OR using a 2x2 table 2x2 table Exposure Outcomes Total Diseased Non-diseased Exposed (+) a+c Non-exposed (-) ac b+d bd Total a+b c+d n OR = (odds of disease in exposed) / (odds of disease in the non-exposed) Formula Odds of exposure in cases = a/c Odds of exposure in controls = b/d Odds Ratio = (a/c) / (b/d)

Examples of data analysis and conclusions Calculate the prevalence of cardiovascular Exposure Outcomes Total disease in the smokers’ group and the Diseased Non- cardiovascular disease prevalence rate in Exposed (+) 89 the non-smokers’ group. Non-exposed (-) diseased 90 14 75 179 Total 3 87 17 162 Results Odds of exposure in cases (P ex) = 14 / 89, = 0.157 or 15.7% Odds of exposure in controls (P non) = 3 / 90, = 0.033 or 3.3% Prevalence Ratio, Odds Ratio = 15.7 / 3.3, = 4.8 Prevalence Difference = 15.7 - 3.3, = 12.4% An odds ratio of 4.8 means that the prevalence of cardiovascular disease in smokers is 4.8 times as high as in non- smoker people. A prevalence difference of 12.4% means that the difference in smoker rate among people with cardiovascular disease and those without disease is 12.4 per 100.

The biases of the cross-sectional analytic study 1. Selection bias Solution 1.1 Self-selection 1. Selection bias 1.2 Late look bias (occurs when one - A random sampling of samples test detects disease earlier) 2. Information bias 1.3 Healthy worker effect - Definition - Orientation & Emphasizing the importance of 2. Information bias: response cooperating / non-response bias < 70% Biases

Advantages & Disadvantages Advantages Disadvantages • Rapid, inexpensive, and can • Unable to sort out what came from first provide analytic clues exposure or outcome • Less prone to error about • Prone to sample distortion bias. exposure recall bias • Not suitable for rare diseases, diseases that recur that last a short time, and acute disease outbreaks

Summary An analytical cross-sectional study is an epidemiological study that observes the occurrence of a disease or health problem without any intervention. Health problems in both exposed and non-exposed groups during the specified period. The correlation was then analyzed and the outcomes were compared between the two groups. The results of this study were able to infer the relationship between factors and outcomes. However, it is not possible to determine what is the cause and what is the consequence of the disease/or health problem.





Chapter Contents Overview of Epidemiological Study Designs 1 Descriptive Study / Observational Study

Overview study design in Epidemiology Definition Exposure means any factor that a population has, is exposed to, or is contacted to, which may contribute to both positive and negative effects. [a protective factor, harmful (a risk factor), or has no effect at all] Outcome is a condition that is expected to result from exposure. • This can be good (positive) or bad (negative). • Other names include dependent variable or response 3

Sample of Exposure & Outcome (Disease) Outcome (Disease) Outcome (Disease) Source: Panithee Thammawijaya, 2557

Overview study design in Epidemiology Epidemiology Study Type and Objective 1. 0bservational study 2. Experimental design For situation analysis For identifying the For studying efficacy and and to postulate (potential) cause: effectiveness, evaluation hypothesis: of programs 1.2 Analytic 1.1 Descriptive epidemiologic epidemiologic study 2.1 Clinical trial study > Cross-sectional study 2.2 Field trial > Ecological study > Case control study 2.3 Community trial (Correlational study) > Cohort study > Cross-sectional study (Prevalence study)

According to the COVID-19 pandemic in Thailand, Who Where When How Why

Descriptive Study Healthy focus on patient Patient Aims Magnitude and Severity Distribution: Time, Place, Person Hypothesis Formulation 7

Descriptive study / 0bservational study ➢ Descriptive studies are observational studies that describe the patterns of disease occurrence in relation to variables such as place, time, and person. ➢ A descriptive study is one in which information is collected without changing the environment (i.e., nothing is manipulated).

Descriptive study: WH questions ✓WHAT ✓HOW : HOW much of the problem Incident rate, Attack rate, Prevalence rate ✓WHO : Characteristics of the patients ✓WHEN : Time line, Epidemic curve ✓WHERE : Place of incidence/prevalence or linking people together 9

Distribution characteristics of diseases or health problems Personal characteristicsPERSONThe distribution of a Geographic location that are usually availableTIMEdisease or health problem or place of residence, for descriptive PLACEaccordingly hour, day,work, school such as epidemiology include month or year such as age, gender, race and / date and time of / Elephant filariasis is most ethnic group, exposure to risk factors common in the southern socioeconomic status, / Date of onset of illness part of the country, occupation, religion, and / Incubation period of the marital status and disease / Allergic dermatitis is lifestyle. / Duration of illness found more in people who . / Disease-free period work in factories with chemicals.

Descriptive study process 1. Determination of 2. Planning for 3. Data collection problems and scope conducting all activities of the study of study 6. Reporting and 5. Discussion and 4. Data Analysis and disseminating the conclusion of the interpretation results of the study study