Epidemiology (7) - Observational Studies

Case reports & series

Accounts of single occurrence of noteworthy health-related incident or collection of such events (e.g., case-series)

Types of analytical observational studies

Cross-sectional and ecological Case-control Cohort

Disadvantages of case reports

Data availability and accuracy is a challenge, esp. if retrospective Not based on systematic study Cases may not be generalizable Causes or associations may have other explanations

Key measure of cross-sectional study

Disease prevalence in those with and without exposures or at different exposure levels are compared

Types of descriptive observational studies

Disease surveillance Case report Case series

Disadvantages of ecological studies

Ecological fallacy: Applying exposure and outcome findings from the group level to the individual level Cannot establish a temporal relationship between exposure and outcome Confounding factors may not be known

Advantages of of case-control studies

Efficient for rare diseases Efficient for diseases with long induction and latent periods because you are starting with the outcome (the case) Can evaluate multiple exposures in relation to a disease Relatively quick and inexpensive

Cross-sectional studies

Exposure status and disease status of an individual are measured at one point in time

Cohort studies

First focus on an exposure: "Exposed" and "unexposed" are selected by the investigators to be followed longitudinally over time to observe difference in the incidence of the outcome (so it is exposure then determine the outcome)

Analytical

Have a comparison group (cross-sectional weakest in this respect—comparison gr. may be unstated and assumed as the larger population); sometimes called "natural experiments"

Cross-sectional studies are useful for

Health planning Hypothesis generating, not etiology (cause) proving

Advantages of case reports

Help identify new clinical issues and may lead to new hypotheses, ex.: Drug side effects and potential new uses (adverse and beneficial) Rare manifestations of a disease New characteristics of diseases Propose (not prove) new etiologies of diseases, e.g. Zika and Microcephaly Educational (maybe)

Disadvantages of cohort studies

Inefficient for evaluation of rare diseases If outcome has long latent period, study can take a long time Generally more expensive If retrospective, requires availability of records to minimize recall (memory) weaknesses Validity of results can be seriously affected by losses to follow-up

Disadvantages of cross-sectional studies

Lack of temporal sequence of exposure preceding disease, thus only describes an association Tends to include cases with long duration, which may have different characteristics and risk factors than series of incident cases Potential misclassification of disease status if disease has exacerbations and remissions (e.g. asthma, multiple sclerosis, lupus) or if disease is being treated (e.g. hypertension)

When to initiate case-control

Little is known about the disease Exposure data are difficult or expensive to obtain Rare disease (but you can identify cases of the disease, ex. Ebola) Disease with long induction and latent period Dynamic underlying population (too hard to track as required in cohort or RCT)

Relative risk

Measure of association used for deriving a causal inference

Pros/cons of prospective

More expensive, time consuming Not efficient for diseases with long latent periods Better exposure and confounder data—collect real-time as part of study Less vulnerable to recall bias

Advantages of cohort studies

Useful design when exposure is rare Examine multiple effects of single exposure (multiple outcomes) If prospective, minimize bias in ascertainment of exposure because you eliminate recall (memory) weaknesses or bias Can examine temporal relationship between exposure and disease (good for establishing causality) Allows direct measurement of incidence of disease in exposed and unexposed, thus calculation of Relative Risk

Matching

Without randomization, cases and controls may differ in characteristics Individual matching (pairwise or multiple)

Selection of case-control studies

Should lead to accurate classification of diseased and non-diseased individuals Homogeneous disease entity by strict diagnostic criteria, e.g. distinguishing cancer of the corpus uteri (body of uterus) from cancer of the cervix (neck of uterus) Applied uniformly

Observational studies

Used to study a wider range of exposures than experimental studies "Natural" experiments Mitigate many issues which are not feasible in experimental studies

Calculating odds ratio

ratio of the probability of the occurrence of an event to that of nonoccurrence.

Ecological studies

relate the frequency with which some characteristic and some outcome of interest occur in a group, perhaps in the same geographic area. Provide quick inexpensive answers to group phenomena

Prospective cohort study

the study population is identified at the beginning of the study and then followed through time.

Retrospective cohort study

the study population is identified in the past and then, typically through records, exposure and disease status are determined.

case-control study

A method of sampling a population in which cases of disease are first identified and enrolled, then a sample of the population that produced the cases (e.g., a comparable population) is identified and enrolled as the control group Exposures are then determined for individuals in each group

Disadvantages of case-control studies

Can investigate only one disease outcome (this is the Case) Cannot directly compute incidence rates of disease in exposed and unexposed (Odds ratio ≠ Relative Risk) Temporal relationship between exposure and disease may be difficult to establish Vulnerable to bias because retrospective (recall bias)

Pros/cons of retrospective

Cheaper, faster Efficient with diseases with long latent period Exposure data may be inadequate Potential for biased recall (remembering)

Advantages of cross-sectional studies

Often have reasonably good generalizability if sample size is large enough Many outcomes and risk factors can be assessed simultaneously Often conducted in a relatively short period of time Less costly than cohort and case-control studies

_____ study is used for diseases with a long latent period such as cancer

Retrospective

Ecological studies is a subset of _______ study

cross-sectional

If relative risk = 1

exposure is NOT associated disease

If relative risk < 1

exposure is associated with a decreased risk of disease (i.e. is protective)

If relative risk > 1

exposure is associated with an increased risk of disease

Descriptive

focus on person, place, time

Ecological studies look at

group characteristics

Calculating relative risk

incidence in exposed / incidence in unexposed

Selection of Controls

must be identified independently of exposure status Controls are a sample of the population that gave rise to the cases (member of control group who gets the disease "would" otherwise end up as a case in the study)

Case reports generally report

new or unique findings; Undescribed disease Adverse reaction Potential link between disease and exposure Treatment outcome

Key measure of case-control study

odds ratio