QUICK-REFERENCE BOX — Which Design Answers Which Question?
- Therapy / prevention? Randomized controlled trial (RCT).
- Prognosis / incidence? Cohort study.
- Rare-disease cause? Case-control study.
- Prevalence right now? Cross-sectional study.
- Best overall answer? Systematic review / meta-analysis of RCTs.
1. Overview
The Core Idea
Study design determines what you can legitimately conclude. Experimental designs — where the investigator assigns the exposure — support causal inference. Observational designs describe associations but are more vulnerable to bias and confounding. Knowing the design tells you, at a glance, how much weight a result deserves.
2. Why It Matters
Bedside Relevance
Neonatology changes practice on the strength of trials (surfactant, antenatal steroids, therapeutic hypothermia, caffeine). Being able to classify a study and judge its design is what lets you separate a practice-changing RCT from a hypothesis-generating case series.
3. Experimental Designs
The Randomized Controlled Trial
- The investigator randomly assigns the intervention, which balances known and unknown confounders across groups.
- Randomization + blinding + intention-to-treat analysis make the RCT the reference standard for therapy questions.
- Variants: parallel-group, crossover, cluster-randomized, and non-inferiority trials.
4. Observational Designs
| Design | Direction | Yields | Main weakness |
|---|
| Cohort | Exposure → outcome (forward) | Incidence, relative risk | Confounding; long/expensive if prospective |
| Case-control | Outcome → exposure (backward) | Odds ratio; efficient for rare outcomes | Recall and selection bias |
| Cross-sectional | Exposure & outcome at once | Prevalence | No temporality (chicken-or-egg) |
| Case series/report | Descriptive | Hypotheses | No comparison group |
5. Evidence Hierarchy
High to Low
- Systematic reviews / meta-analyses of RCTs
- Individual randomized controlled trials
- Cohort studies
- Case-control studies
- Cross-sectional studies
- Case series / case reports
- Expert opinion / bench research
6. Matching Design to Question
| Question type | Best design |
|---|
| Does this treatment work? | Randomized controlled trial |
| What is the prognosis / incidence? | Cohort study |
| What caused this rare disease? | Case-control study |
| How common is it now? | Cross-sectional study |
| What does all the evidence say? | Systematic review / meta-analysis |
7. Design Pitfalls
Watch For
- Cross-sectional data used to claim causation (no temporality).
- Case-control studies over-interpreted despite recall/selection bias.
- Uncontrolled case series presented as evidence of effectiveness.
- Observational associations treated as if they were randomized comparisons.
8. Key Pearls
High-Value Points
- Only randomization balances unknown confounders — its defining advantage.
- Case-control = start with the outcome (good for rare diseases) → odds ratio.
- Cohort = start with the exposure → relative risk and incidence.
- Cross-sectional = a snapshot → prevalence, no cause-and-effect.
- The design caps the conclusion — a great analysis can't rescue a weak design.
9. Common Mistakes to Avoid
Misreadings & Better Practice
Frequent errors in interpreting study design.
| Mistake | Why it's wrong | Better reading |
|---|
| "Cross-sectional shows X causes Y." | No temporality. | It shows association/prevalence only. |
| Reporting relative risk from a case-control study. | Case-control yields odds ratios. | Use OR (approximates RR if outcome rare). |
| Treating a case series as proof of efficacy. | No control group. | Hypothesis-generating only. |
10. Board-Style High-Yield Summary
Key Takeaways
- RCT (randomization) supports causal inference; observational studies show association.
- Cohort → RR/incidence; case-control → OR (rare outcomes); cross-sectional → prevalence.
- Evidence hierarchy tops out at systematic reviews/meta-analyses of RCTs.
- Match the design to the question type.
- Design determines the strength of the conclusion.
11. References
- 1.Grimes DA, Schulz KF. An overview of clinical research: the lay of the land. Lancet. 2002;359:57–61.
- 2.OCEBM Levels of Evidence Working Group. The Oxford Levels of Evidence.
- 3.Guyatt G, et al. Users' Guides to the Medical Literature.
Back to Clinical Guideline Hubs