Meta-analytic case-control study — pooling evidence from case-control designs
Meta-Analysis of Case-Control Studies · Also known as: pooled case-control analysis, case-control meta-analysis, meta-analytic case-control design, systematic pooled case-control
A meta-analytic case-control study systematically identifies, critically appraises, and quantitatively synthesizes data from multiple independent case-control studies examining the same exposure-disease relationship. By pooling odds ratios across studies, it yields a more precise and generalizable estimate of association than any single study can provide, while formally quantifying heterogeneity across populations, settings, and study periods.
Read the full method
Sign in with a free account to read this section.
Method map
The neighbourhood of related methods — select a node to explore.
When to use it
Use a meta-analytic case-control study when multiple independent case-control studies have examined the same exposure-disease pair and you need a precise, synthesized effect estimate — particularly for rare diseases where individual studies are underpowered. It is well suited to pharmacoepidemiology, occupational exposure research, and nutritional epidemiology. Do not use it as a substitute for a primary study when no adequate case-control literature exists; do not pool studies that differ fundamentally in case definition, exposure measurement, or control source without pre-specified heterogeneity analyses; and avoid it when only two or three studies are available, as the pooled estimate will be unstable and funnel-plot asymmetry tests will be unreliable.
Strengths & limitations
- Produces a more precise pooled odds ratio than any individual case-control study by increasing effective sample size.
- Formally quantifies consistency of evidence across settings, populations, and time periods.
- Identifies sources of heterogeneity that individual studies lack power to detect, generating hypotheses for future research.
- Transparent and reproducible when conducted according to PRISMA-O/MOOSE reporting guidelines.
- Can detect publication bias and small-study effects through funnel plot and statistical tests.
- The pooled estimate inherits the confounding and recall bias present in the underlying case-control studies — garbage in, garbage out.
- Substantial clinical and methodological heterogeneity across primary studies can make pooling misleading even when statistical heterogeneity is low.
- Publication bias systematically inflates pooled effect sizes; negative or null case-control studies are less likely to be published.
- Aggregation of study-level data (rather than individual participant data) prevents adjustment for individual-level confounders not uniformly reported across studies.
- Quality and completeness of reporting in older case-control studies often limits what can be extracted and synthesized.
Frequently asked
How is a meta-analytic case-control study different from a standard meta-analysis of randomized trials?
The statistical pooling machinery is similar, but the underlying study design is observational. Case-control studies report odds ratios rather than risk ratios or mean differences, and they carry systematic biases — recall bias, selection bias in control choice, and unmeasured confounding — that do not exist in randomized trials. These biases are not eliminated by pooling; they are inherited and potentially amplified. Quality appraisal tools (Newcastle-Ottawa Scale rather than Cochrane RoB) and sensitivity analyses are therefore especially important.
Should I use a fixed-effects or random-effects model?
Random-effects models are almost always more appropriate for meta-analyses of case-control studies because the underlying populations, exposure definitions, and control sources vary across studies, making the assumption of a single true effect size (fixed-effects) untenable. The DerSimonian-Laird estimator is widely used but can underestimate variance; REML-based estimators or the Hartung-Knapp-Sidik-Jonkman correction are preferable when the number of studies is small.
What is an acceptable I² for pooling?
I² quantifies the proportion of total variability attributable to between-study heterogeneity rather than chance. There are no universal cut-offs that determine whether pooling is appropriate — even a low I² does not guarantee clinical homogeneity. Inspect the forest plot for direction and magnitude consistency, assess clinical plausibility, and conduct pre-specified subgroup analyses to explore heterogeneity rather than relying solely on I².
Can I pool adjusted and unadjusted odds ratios together?
Mixing crude and adjusted ORs is generally inadvisable because they estimate different quantities and are subject to different degrees of confounding control. Where possible, extract the most fully adjusted OR from each study, note which confounders were controlled, and consider sensitivity analyses that include or exclude studies with minimal adjustment to assess the impact of confounding on the pooled estimate.
What reporting guideline applies?
The MOOSE (Meta-analysis Of Observational Studies in Epidemiology) statement and the PRISMA-O (Preferred Reporting Items for Systematic Reviews and Meta-Analyses of Observational Studies) extension are the primary reporting standards. Following these checklists ensures transparency in the search strategy, quality appraisal, data extraction, and presentation of results, facilitating peer review and replication.
Sources
- Shapiro, S. (1994). Meta-analysis/Shmeta-analysis. American Journal of Epidemiology, 140(9), 771-778. DOI: 10.1093/oxfordjournals.aje.a117324 ↗
- Stroup, D. F., Berlin, J. A., Morton, S. C., Olkin, I., Williamson, G. D., Rennie, D., ... & Thacker, S. B. (2000). Meta-analysis of observational studies in epidemiology: a proposal for reporting. JAMA, 283(15), 2008-2012. DOI: 10.1001/jama.283.15.2008 ↗
How to cite this page
ScholarGate. (2026, June 3). Meta-Analysis of Case-Control Studies. ScholarGate. https://scholargate.app/en/epidemiology/meta-analytic-case-control-study
Which method?
Set this method beside its closest kin and read them side by side — the library lays the books on the table; the choice is yours.
- Case-control studyEpidemiology↔ compare
- Matched case-control studyEpidemiology↔ compare
- Nested case-controlEpidemiology↔ compare
- Prospective Case-Control StudyEpidemiology↔ compare