Concept Architecture
Concept
Theoretically, Case-Control Design is an observational epidemiological study design in which individuals with a specified outcome or disease (cases) are compared with individuals without the outcome (controls) to investigate prior exposure to potential risk factors. The concept is founded on observational study methodology, probability theory and statistical inference. It exists to estimate the association between exposures and outcomes, particularly for rare diseases or conditions with long latency periods where prospective follow-up would be inefficient.
Mathematically, Case-Control Design is represented by comparing the odds of prior exposure between cases and controls. The principal measure of association is the odds ratio, which estimates the relative odds of exposure among cases compared with controls. Logistic regression extends this framework by estimating adjusted odds ratios while controlling for potential confounding variables. Because sampling is based on outcome status, direct estimation of disease incidence or relative risk is generally not possible.
In practice, Case-Control Design begins with identifying eligible cases and selecting an appropriate control group from the same underlying population. Historical exposure information is then collected from medical records, registries, interviews or administrative databases. Statistical analyses estimate crude and adjusted odds ratios, which are frequently used in health economics to quantify disease risk, estimate epidemiological parameters for decision models and evaluate associations between healthcare interventions and outcomes.
Purpose
Used to investigate associations between prior exposures and disease outcomes by comparing exposure histories among individuals with and without the outcome of interest, particularly when studying rare diseases or outcomes with long latency periods.
Mathematical Formulae
Primary Formula
Odds Ratio:
OR = (a ? d) / (b ? c)
where:
- a = exposed cases
- b = exposed controls
- c = unexposed cases
- d = unexposed controls
Supporting Formulae
Logistic regression model:
log(p / (1 ? p)) = ?? + ??X? + ??X? + ? + ??X?
Estimated odds ratio for exposure:
OR = e^??
Related Mathematical Methods
- Logistic Regression
- Maximum Likelihood Estimation
- Odds Ratio Estimation
- Conditional Logistic Regression
- Matched Case-Control Analysis
- Confounding Adjustment
- Stratified Analysis
Example
A study investigates whether previous exposure to a high-cost immunosuppressive therapy is associated with a rare adverse event. The study includes 200 cases with the adverse event and 400 matched controls without the event. Among the cases, 80 were previously exposed, while 120 were not. Among the controls, 100 were exposed and 300 were not.
OR = (80 ? 300) / (100 ? 120) = 24,000 / 12,000 = 2.0
The odds of previous exposure are therefore estimated to be twice as high among cases as among controls. This estimate may subsequently inform parameterisation of a health economic decision model.
Excel Implementation
| Function | Example Formula | Health Economics Application |
|---|---|---|
| PRODUCT | =PRODUCT(B2,E2) | Calculate the numerator of the odds ratio. |
| PRODUCT | =PRODUCT(C2,D2) | Calculate the denominator of the odds ratio. |
| IFERROR | =IFERROR((B2*E2)/(C2*D2),"") | Calculate the odds ratio while handling zero counts. |
| EXP | =EXP(B10) | Convert logistic regression coefficients into odds ratios. |
| LN | =LN(F2) | Calculate the natural logarithm of the odds ratio for statistical analyses. |
VBA (Optional)
VBA can automate calculation of odds ratios, confidence intervals and summary tables for multiple case-control analyses.
Sources
- Rothman KJ, Greenland S, Lash TL. Modern Epidemiology.
- Breslow NE, Day NE. Statistical Methods in Cancer Research. Volume I: The Analysis of Case-Control Studies.
- Hosmer DW, Lemeshow S, Sturdivant RX. Applied Logistic Regression.
- Agresti A. Categorical Data Analysis.
- NICE. Health Technology Evaluation Manual.
- Drummond MF, et al. Methods for the Economic Evaluation of Health Care Programmes.
Related Concepts (2)
Library
Publications
1
Good Practices for Real-World Data Studies of Treatment and/or Comparative Effectiveness: Recommendations from the Joint ISPOR-ISPE Special Task Force on Real-World Evidence in Health Care Decision Making — Berger, Sox, Willke, Brixner, Eichler, Goettsch, Madigan, Makady, Schneeweiss, Tarricone, Wang, Watkins & Mullins, Vol. 20, No. 8 ed., 2017 (Value in Health)
The joint ISPOR-ISPE recommendations on good procedural practice for real-world data studies (observational studies and registries) used to inform healthcare decisions — study registration, replicability and stakeholder involvement — the reference for RWE credibility in HTA.
Journal ArticleView source →
Frequently Asked Questions (6)
What is a case-control design?
An observational design comparing individuals who have experienced an outcome, the cases, against similar individuals who have not, the controls, on prior exposure.
Source: Rothman KJ, Greenland S, Lash TL. Modern Epidemiology. 3rd ed. Lippincott Williams & Wilkins; 2008.
Why is a case-control design efficient for rare outcomes?
A case-control study starts from the outcome, gathering people who already have a condition and comparing their past exposures with those of similar people who do not. This makes it efficient for rare outcomes, because it deliberately assembles enough cases rather than waiting for them to arise, which a forward-looking study would need a huge sample and long follow-up to do. The efficiency comes at the cost of relying on retrospective exposure information, which can be less reliable. It works backward from effect to cause. Rothman and colleagues (2008) describe this design.
Source: Rothman et al. 2008
How does a case-control study work?
A case-control study works by identifying cases, individuals with the outcome of interest, and controls, comparable individuals without it, then ascertaining and comparing their prior exposures to determine whether an exposure is more common among cases than controls. The comparison yields a measure of association, typically the odds ratio. Selecting appropriate controls that represent the population from which the cases arose is important. Because it starts with the outcome and looks back at exposure, the case-control design efficiently studies factors associated with rare or slowly developing outcomes.
Source: Rothman, Greenland & Lash 2008
What are the advantages of a case-control design?
The advantages of a case-control design are its efficiency for studying rare outcomes, since it starts with cases who have the outcome rather than waiting for it to occur; its suitability for outcomes with long latency, avoiding lengthy follow-up; its ability to examine multiple exposures for a single outcome; and its relatively low cost and quick conduct. By beginning from the outcome and looking backward at exposures, it can investigate associations that would require very large or long cohort studies. These advantages make the case-control design valuable for certain questions, particularly rare diseases.
Source: Rothman, Greenland & Lash 2008
What are the limitations of a case-control design?
The limitations of a case-control design include its susceptibility to bias: selection bias in choosing cases and especially controls, which must represent the source population; recall bias, since exposure is often ascertained retrospectively and may be remembered differently by cases and controls; and difficulty establishing the time order of exposure and outcome. It usually yields an odds ratio rather than direct risk and cannot easily give incidence. These limitations mean case-control studies are designed carefully, particularly in control selection and exposure measurement, and interpreted with attention to potential biases.
Source: Rothman, Greenland & Lash 2008
How does a case-control design differ from a cohort design?
A case-control design differs from a cohort design in direction: a case-control study starts with the outcome, comparing exposures in those with and without it, looking backward, while a cohort study starts with exposure, following groups forward to see who develops the outcome. Case-control studies are efficient for rare outcomes and yield odds ratios, whereas cohort studies suit rare exposures, can measure incidence and multiple outcomes, and establish time order more clearly. So the two differ in starting point, direction, efficiency for different situations, and the measures they provide.
Source: Rothman, Greenland & Lash 2008
Trust Record
Verified by Dr Darrin Baines
British health economist
Professional identity: darrinbaines.org
Verification date: 12 Nov 2025
Content version: 1.0.0
Canonical Identity
- Persistent URI
- https://healtheconomics.wiki/concept/case-control-design
- Term code
- HE-ES-CTM-012
Stable URI · Machine-readable · Resolvable · CC BY 4.0