Epidemiology approach

Asking for Preferences

Reading File

Finding Sources

Reading File

Loading Image

I now have comprehensive content from Park's Textbook. Let me compile the full structured answer.

Epidemiological Study Designs

Source: Park's Textbook of Preventive and Social Medicine

Epidemiological study designs are broadly divided into three major categories: Descriptive, Analytical (Observational), and Experimental.

1. DESCRIPTIVE EPIDEMIOLOGY

Descriptive studies describe the distribution of disease in a population in terms of Person, Place, and Time. They do not test hypotheses - they generate them.

Uses:

Provide data on the magnitude of disease (morbidity, mortality rates)
Provide clues to disease aetiology and help formulate aetiological hypotheses
Provide background data for planning and evaluating preventive/curative services
Describe variations in disease occurrence by time, place, and person

Types:

Case reports and case series
Ecological (correlational) studies
Cross-sectional surveys

2. ANALYTICAL EPIDEMIOLOGY

In contrast to descriptive studies that look at entire populations, analytical studies focus on the individual within the population. The object is not to formulate, but to test hypotheses. Although individuals are evaluated, the inference is to the population from which they are selected.

The two main analytical study designs are:

A. Case-Control Study (Retrospective)

A case-control study has three distinct features:

Both exposure and outcome (disease) have already occurred before the start of the study
The study proceeds backwards from effect to cause
It uses a control or comparison group to support or refute an inference

Design (Fig. 8 from Park's):

Design of a Case Control Study - proceeds backwards from cases and controls to determine past exposure

Start with: Cases (people with disease) + Controls (people without disease), both drawn from the same population
Ask: Were they exposed or not exposed in the past?
Measure of association: Odds Ratio (OR) - an estimate of Relative Risk

When to use: Rare diseases, quick results needed, limited resources, generates or tests a specific hypothesis first.

B. Cohort Study (Prospective)

Design (Fig. 8 from Park's):

Design of a Cohort Study - follows people forward from exposure to disease outcome

Start with: People without the disease
Divide into: Exposed vs Not Exposed
Follow forward in time to see who develops disease
Measure of association: Relative Risk (RR), Attributable Risk (AR), Incidence Rates

Comparison groups in cohort studies:

Internal comparisons: Sub-groups within a single cohort classified by degree of exposure (e.g., cigarettes/day)
External comparisons: An outside control group (e.g., smokers vs. non-smokers)
General population rates: Experience of exposed group compared to general population rates

Classic example - Doll & Hill (1951): Sent questionnaires to 59,600 British doctors about smoking habits, formed two cohorts (smokers and non-smokers), followed for 4 years and 5 months. Results confirmed the association between smoking and lung cancer.

Classic example - Framingham Heart Study (1948): Followed residents of Framingham, Massachusetts, for 20 years to study relationships between risk factors (serum cholesterol, BP, weight, smoking) and cardiovascular disease.

Key Measures from Cohort Studies

Measure	Formula	Meaning
Relative Risk (RR)	Incidence in exposed / Incidence in non-exposed	Strength of association; RR = 1 means no association
Attributable Risk (AR)	Incidence (exposed) - Incidence (non-exposed)	Portion of disease attributable to the exposure
Population-Attributable Risk	Incidence (total population) - Incidence (non-exposed)	Disease burden in the population due to the exposure

Example (smoking & lung cancer, Doll & Hill):

RR = 10 (smokers are 10x more likely to develop lung cancer)
AR = (10 - 1) / 10 × 100 = 90% (90% of lung cancer in smokers is attributable to smoking)

Case-Control vs. Cohort: Key Differences

Feature	Case-Control	Cohort
Direction	Effect → Cause (retrospective)	Cause → Effect (prospective)
Starting point	The disease	People exposed to risk factor
Test question	Does suspected cause occur more in cases vs. controls?	Does disease occur more in exposed vs. unexposed?
Use	First approach; exploratory; rare diseases	Testing a precisely formulated hypothesis
Sample size	Smaller	Larger
Time	Quick results	Long follow-up, delayed results
Measure	Odds Ratio (estimate of RR)	Incidence rates, RR, AR
Cost	Relatively inexpensive	Expensive

3. EXPERIMENTAL EPIDEMIOLOGY

Experimental studies are similar in approach to cohort studies except the conditions are under the direct control of the investigator. The investigator deliberately applies or withdraws the suspected cause/intervention in an experimental group while making no change in a control group, then compares outcomes.

This sharply contrasts with observational studies (descriptive, case-control, cohort), where the epidemiologist only observes the natural course of events.

Aims:

To provide scientific proof of aetiological/risk factors
To measure effectiveness and efficiency of health services for prevention and treatment

Two types:

Randomized Controlled Trials (RCTs) - strict random allocation
Non-randomized/quasi-experimental trials - depart from strict randomization for practical reasons

Randomized Controlled Trial (RCT)

Design (Fig. 9 from Park's):

Design of a Randomized Controlled Trial - population is selected, exclusions made, then randomized into experimental and control groups for manipulation, follow-up, and assessment

Basic steps in conducting an RCT:

Drawing up a protocol (aims, objectives, selection criteria, sample size, allocation procedures, treatment schedule, evaluation plan)
Selecting reference and experimental populations
Randomization (ensures comparability of groups, eliminates selection bias)
Manipulation or intervention (apply treatment/vaccine/preventive measure)
Follow-up
Assessment of outcome

Experimental studies carry all the advantages and disadvantages of cohort studies, plus three additional challenges: cost, ethics, and feasibility.

Sub-types of experimental studies (based on study design):

Concurrent parallel: Two groups studied simultaneously - one receives intervention, other does not
Cross-over: Each participant receives both the treatment and control at different times (serves as their own control)

Summary Hierarchy (Levels of Evidence)

Experimental
  └── RCT (highest evidence for causation)
  └── Non-randomized trials

Analytical (Observational)
  └── Cohort Study (prospective)
  └── Case-Control Study (retrospective)
  └── Cross-Sectional Study

Descriptive
  └── Ecological studies
  └── Case reports / Case series
  └── (lowest for establishing causation)

The progression in epidemiological investigation typically follows: Descriptive → Hypothesis generation → Case-Control (test hypothesis) → Cohort (confirm) → RCT (prove causation/efficacy).

- Park's Textbook of Preventive and Social Medicine, pp. 83-94

This is a shared conversation. Sign in to Orris to start your own chat.