Study Designs in Public Health PDF

11/7/2024 Polytechnic University A. Observational Studies Faculty Of Medicine Investigator observes occurrence of condition/s in “self-assigned” groups of people; exposure not as...

11/7/2024 Polytechnic University A. Observational Studies Faculty Of Medicine Investigator observes occurrence of condition/s in “self-assigned” groups of people; exposure not assigned Carried out in more natural settings - “natural experiments” (6106223) Often most practical and feasible Lecture 6 Less control over study situation; results more 2024-2025 susceptible to distorting influence 2 1 1  11/7/2024 Did investigator Manipulation of exposure assign exposures? Experimental yes no Investigator Investigator observes exercises control occurrence of Experimental Observational over allocation of condition/s in “self”- Study Study exposure assigned groups of people Observational 3 4 2  11/7/2024 Did investigator Descriptive Study assign exposures? no Sets out to describe a situation Observational Study ex. Distribution of depression in a population in relation to sex, age, and other characteristics comparison group? – Often no a-priori hypothesis yes no Analytical Descriptive Study Study 5 6 3  11/7/2024 Analytical Study Directionality of investigation Must have clear and measurable hypothesis At least 1 dependent (outcome) variable and 1 independent (“exposure”) variable retrospective (case-control ) Sets out to test hypotheses or detect Exposure Outcome associations prospective (cohort) ex. Identify factors that explain higher rates of depression among women 7 8 4  11/7/2024 Design of a PROSPECTIVE Study Time 1. Case Report Studies Case studies describe in-depth the characteristics of one or a limited number of ‘cases’. Direction of inquiry A case report is a detailed report of the diagnosis, treatment, response to treatment, and follow-up after treatment of an individual patient. A case may be, for example, a patient, a health center, or a START WITH: village. D+ Such a study can provide quite useful insight into a problem. Exposed Case studies are common in social sciences, management sciences, D- and clinical medicine. POPULATION For example, in clinical medicine the characteristics of a still D+ unrecognized illness may be documented as a case study. This is often the first step toward building up a clinical picture of that illness. Unexposed D- 9 Study designs 10 5  11/7/2024 Case Reports 2. Case Series Case reports are considered the lowest level of human- A case series is group of case reports involving patients who were based/clinical evidence, but they are also the first line of given similar treatment. evidence, because they are where new issues and ideas emerge. Unique cases that cannot be explained by known diseases or syndromes. The Author describes some interesting observations that occurred to a small number of patients. Cases that show an important variation of a disease or condition. Multiple case reports showing similar findings will trigger more Cases that show unexpected events that may yield new or useful in-depth investigation. information Objective and brief report of a clinical characteristic or outcome from Most common in new or rare cases a group of clinical subjects. Generalization is not possible due to biased selection or unrepresentativeness of subjects, lack of control group, etc. Study designs 11 Study designs 12 6  11/7/2024 Case Series-CONT. Case Series & Case Reports It is a medical research descriptive study that tracks patients with a known exposure given similar treatment, or examines their medical records for exposure and outcome. It can be retrospective or prospective. Case series may be confounded by selection bias, which limits statements on the causality of correlations observed; For example, physicians who look at patients with a certain illness and a suspected linked exposure will have a selection bias in that they have drawn their patients from a narrow selection (namely their hospital or clinic). Study designs 13 Study designs 14 7  11/7/2024 In Case-Control studies – Retrospective – Question answered: “What happened?” 3. Case control studies – Matching needed for controls (Age, gender and locality) – Might be difficult to differentiate from Case Series (Both are after the fact) 15 16 8  11/7/2024 Case control study 17 18 9  11/7/2024 Odd Ratios Case Control Total Smoker 60(a) 100 (b) 160 Non Smoker 40 (c) 300 (d) 340 Total 100 400 500 Study designs 19 Study designs 20 10  11/7/2024 Strengths of case-control studies Disadvantages (Case-Control) Well suited to study etiology of rare outcomes, Case and control selection troublesome e.g. cancer, congenital malformations. Subject to bias Can easily study multiple exposures Direct estimation of incidence is not Efficient if long delay between exposure and possible outcome If the incidence of exposure is high, it is Require fewer individuals (i.e. relatively difficult to show the difference between inexpensive) cases and controls 21 22 11  11/7/2024 4. Cohort Studies Limitations of case-control studies Step 1 in a prospective study design: identify Not well suited to study multiple outcomes relevant group/s of people and collect The time sequence of exposure and outcome information about their exposure history can be unclear Step 2: Follow these people over time and Does not provide data on absolute risk measure the incidence of outcome/s of interest Relies on information about past exposures that Retrospective and prospective cohort studies may be prone to bias depend on the direction of a follow-up inquiry 23 24 12  11/7/2024 Design of a COHORT Study 2*2 typical table Time Direction of inquiry Outcome Status D+ present absent Exposed D- POPULATION exposed D+ Exposure Unexposed D- Status not 25 exposed 26 13  11/7/2024 CHD Design of a RETROSPECTIVE Study + - Total + 84 2916 3000 Time Smoking - 87 4913 5000 Direction of inquiry Relative Risk or Risk Ratio START WITH: E+ CASES Incidence in exposed+ = 84/3000 = 28.0/1000 (with Incidence in exposed- = 87/5000 = 17.4/1000 E- disease) POPULATION RR = 28.0÷17.4 = 1.61 E+ CONTROLS Risk for CHD among smokers is 1.6 times (without higher than among non-smokers E- disease) 27 Study designs 28 14  11/7/2024 Strengths of cohort studies Relative Risk Well suited to study rare exposures Can easily study multiple outcomes Provides direct measure of risk of outcome among RR > 1 = positive association exposed and unexposed persons RR = 1 = no association Not dependent on recall of past exposures RR < 1 = negative association Begins with “healthy” persons thereby preventing 'selective survival' bias 30 Study designs 29 15  11/7/2024 Limitations of cohort studies Main limitation of the prospective Inefficient for rare outcomes Not well suited to study multiple exposures design is time & cost Assessment of exposure status may influence participant’s Especially when studying chronic diseases behavior Definitions of exposures and outcomes may change over time that may only become apparent years after Difficult if long delay between exposure and outcome the exposure/s of interest or may require Lost-to-follow-up bias years of exposure to “cause” the outcome 31 32 16  11/7/2024 Retrospective Historical Prospective Design Historical and concurrent cohort studies Alternative strategy to “concurrent” prospective design to reduce time/costs PAST PRESENT FUTURE Requires identifying a defined cohort from some time in the past Historical cohort Follow-up period is the time since determination of groups Follow-up exposure status until present (or future) assembled Incidence/risk measures can be estimated in the same manner as in a concurrent prospective study Concurrent cohort groups Follow-up 33 assembled 34 17  11/7/2024 Historical Prospective Design Historical Prospective Design Particularly useful when the exposure under Example: thyroid cancer investigation is “unique” in some way. e.g., risk among people exposed occurred only in the past, occurred in specific to the 1986 Chernobyl groups of people nuclear-reactor accident Often applied to the study of acute environmental exposures. 35 36 18  11/7/2024 Historical Prospective Design Historical Prospective Design Possible sources of information about Sometimes difficult to obtain a exposure status: comprehensive list of persons who experienced the outcome (less problematic Industrial worker records in a mortality study) Military records Often difficult to obtain information about Insurance companies or health care provider other exposures in the cohort companies Registries of persons receiving specific medical treatment 37 38 19  11/7/2024 Did investigator assign exposures? 3. Cross-sectional (prevalence) study no Observational Study comparison group? Usually involves a representative (random) sample of a dynamic population yes no Examines exposures and outcomes Analytical Descriptive Study Study simultaneously direction? Based on prevalence data Case- Cross- Cohort sectional Control Study Study 39 Study 40 20