Types of Study

Question 1

Descriptive Study

Answer 1

Measures the frequency of health outcomes and their distribution (prevalence)

Uses routine data and surveys, understanding prevalence and burden of health issue

Who, what, where, when

Question 2

Analytical study

Answer 2

Aims to investigate which exposures may be responsible for health outcomes (detect or confirm hypothesised associations)
- Can be ecological, cross-sectional, cohort or case-control

Question 3

Cross-sectional Study

Answer 3

Compares outcome and exposure measured simultaneously (at one point in time)

Measure: Prevalence

Strengths: Easy to collect data - rapid and inexpensive; Good for fixed exposures; Can collect data on multiple variables and outcomes

Weaknesses: Cannot prove causality, Information bias (reverse causality, re-call), Confounding, Long latency periods

Question 4

Ecological Study

Answer 4

Compares the frequency of exposure and outcome at a population level (aggregate group data). Population level analysis (rather than individual).

Measure: Dependent on method of data collection (prevalence, risk or incidence rate)

Strengths: easy to collect routine data (rapid and inexpensive); allows identification of associations for investigation; compares groups, locations and time periods

Weaknesses: Ecological fallacy (cannot be used to infer causality at an individual level); confounding (lack of info); information bias (misclassification), selection bias (data)

Question 5

Cohort Study

Answer 5

Group of individuals who share a common characteristic are followed up over time to measure exposure to a risk factor and measure incidence of outcome. Can be prospective or retrospective.

Measure: Incidence risk or rate

Strengths: temporality (best design to infer causality); can identify multiple outcomes and new exposures; low probability of selection bias and confounding; can look at dose-response

Weaknesses: Loss to follow-up (needs large sample); logistically difficult, time-consuming and expensive; ethical conflict if adverse outcomes

Question 6

Case-Control Study

Answer 6

Defines study group by outcome (cases and controls) and compare differences in previous exposure. Can use matching and nested controls.

Measure: Odds ratio of exposure

Strengths: rapid and inexpensive (compared to cohort studies); useful for rare outcomes, genetics and latency; can investigate multiple exposures (retrospective)

Weaknesses: prone to selection bias and information bias; confounding (due to insufficient data), overmatching, prevalence can create reverse causality

Question 7

Intervention Study

Answer 7

Measures that association between outcome and exposure for a specific intervention (prevention or treatment)

Measure: Dependent on the methods used to collect data (incidence risk rate - cohort; prevalence - cross-sectional)

Strengths: Temporality (exposure prior to outcome); reversibility (can remove or reduce exposure to minimise or eliminate outcome); equal distribution of confounding factors (randomised); allocation concealment reduces bias

Weaknesses: Expensive and time-consuming (large sample, follow-up); selection bias at enrolment (perceived risk); bias due to non-compliance

Question 8

Detail the difference between randomised and non-randomised control trials

Answer 8

Non-randomised control trials:

• Can be used to evaluate interventions by comparison with a historical, geographical or opportunistic control group
• Useful when intervention is complex, there is a need for large-scale evidence or RCT is unethical
• Subject to bias and confounding

RCTs:
- Requires use of contemporary comparison group (control arm) and is randomly allocated (exposure and control)
- Random allocation via: simple, systematic, blooded, stratified, matched.
Strength: controls adjust for background effects; equal distribution of confounding factors
Limitations: complex (confounders, modifiers, bias); time-consuming and resource intensive

Question 9

Detail some of the ethical issues related to the use of intervention studies

Answer 9

• Withholding our withdrawing intervention from comparison group (if already shown to be safe and effective)
• Sufficient clinical or public health importance must be balanced against risk and harm
• Ethics vary significantly across countries and over time