Concept Architecture
Concept
Theoretically, Matching-Adjusted Indirect Comparison (MAIC) is a population-adjusted comparative effectiveness method used to estimate relative treatment effects when head-to-head clinical trials are unavailable and individual patient data (IPD) are available for one treatment but only aggregate data are available for the comparator. It is based on causal inference and propensity score weighting, whereby patients in the IPD trial are reweighted so that their baseline characteristics match published aggregate characteristics from the comparator trial. The method exists to reduce bias arising from cross-trial differences in patient populations.
Mathematically, MAIC is represented as a weighting optimisation problem in which patient-specific weights are estimated so that the weighted means of selected baseline covariates equal the corresponding aggregate means reported for the comparator study. Relative treatment effects are then estimated using the weighted pseudo-population, with uncertainty quantified using weighted regression methods and effective sample size.
In practice, MAIC is implemented by selecting clinically relevant effect modifiers and prognostic variables, estimating balancing weights, assessing covariate balance and calculating weighted treatment effects. In health economics it is widely applied in health technology assessment to support indirect treatment comparisons, cost-effectiveness analyses and reimbursement submissions when direct comparative evidence is unavailable.
Purpose
Used to reduce cross-trial population differences and estimate comparative treatment effects for economic evaluation and health technology assessment when head-to-head evidence is unavailable.
Mathematical Formulae
Primary Formula
w? = exp(x??�)
where � is chosen such that:
?w?x? / ?w? = x?*
where:
w? = patient weight
x? = vector of baseline covariates for patient i
x?* = published aggregate covariate means from the comparator trial
Supporting Formulae
Effective Sample Size:
ESS = (?w?)� / ?w?�
Weighted treatment effect:
?? = Weighted estimate obtained using the MAIC weights
Related Mathematical Methods
Propensity Score Weighting
Method of Moments
Logistic Weight Calibration
Inverse Probability Weighting
Anchored Indirect Comparison
Unanchored Indirect Comparison
Network Meta-Analysis
Example
An individual patient dataset contains 500 patients receiving Treatment A, while only published aggregate results are available for Treatment B. Patients receiving Treatment A are reweighted so that age, disease severity and prior treatment history match the published characteristics of the Treatment B trial. After weighting, the effective sample size decreases from 500 to 318, reflecting the reduction in independent information. The weighted analysis estimates a hazard ratio of 0.78 for Treatment A versus Treatment B, which is subsequently used within a cost-effectiveness model.
Excel Implementation
| Function | Example Formula | Health Economics Application |
|---|---|---|
| EXP | =EXP(A2) | Calculate exponential patient weights |
| SUMPRODUCT | =SUMPRODUCT(weights,covariates)/SUM(weights) | Calculate weighted covariate means |
| SUM | =SUM(range) | Calculate total patient weights |
| SUMSQ | =SUMSQ(weights) | Calculate effective sample size |
| Solver | Optimise � values | Estimate balancing weights that match aggregate covariates |
VBA (Optional)
VBA can automate patient weighting, covariate balance assessment, effective sample size calculation and repeated MAIC analyses across multiple treatment comparisons.
Sources
- Signorovitch JE, Sikirica V, Erder MH, et al. Matching-Adjusted Indirect Comparisons: A New Tool for Timely Comparative Effectiveness Research. Value in Health. 2012.
- Phillippo DM, Ades AE, Dias S, et al. Methods for Population-Adjusted Indirect Comparisons in Health Technology Appraisal. Medical Decision Making.
- NICE. Health Technology Evaluations: The Manual.
- ISPOR Good Practice Reports.
- Briggs A, Claxton K, Sculpher M. Decision Modelling for Health Economic Evaluation.
Related Concepts (2)
Library
Publications
1
Interpreting Indirect Treatment Comparisons and Network Meta-Analysis for Health-Care Decision Making: ISPOR Task Force on Indirect Treatment Comparisons Good Research Practices, Part 1 — Jansen, Fleurence, Devine, Itzler, Barrett, Hawkins, Lee, Boersma, Annemans & Cappelleri, Vol. 14, No. 4 ed., 2011 (Value in Health)
The ISPOR good-practice guidance on interpreting indirect treatment comparisons, network and mixed treatment comparisons for decision making — terminology, assumptions, validity and how to critically appraise an ITC/NMA when head-to-head trial evidence is unavailable.
Journal ArticleView source →
Frequently Asked Questions (6)
What is a matching-adjusted indirect comparison?
A method comparing treatments from separate trials with different populations by reweighting one trial's patient data to match the other trial's reported characteristics.
Source: Signorovitch et al. 2010
Why does a matching-adjusted indirect comparison reweight one trial's patients?
When two treatments are compared indirectly through a common comparator, the comparison can be distorted if the trials enrolled different kinds of patients. A matching-adjusted indirect comparison corrects for this by reweighting the individual patients in one trial so that their average characteristics match those reported for the other trial, before the comparison is made. This aligns the populations on measured factors, reducing the bias a naive indirect comparison would carry. It requires patient-level data for one trial and only summary data for the other. Signorovitch and colleagues (2010) developed this method.
Source: Signorovitch et al. 2010
How does a matching-adjusted indirect comparison work?
A matching-adjusted indirect comparison works by using the individual patient data from one trial to reweight its patients so that their average characteristics match those reported for the comparator trial, typically through a weighting analogous to propensity weighting on the effect-modifying and prognostic variables. The reweighted trial's outcomes are then compared with the comparator trial's reported results, giving an indirect estimate adjusted for the population differences. This balances the populations on the observed characteristics used, so the comparison reflects a common population, reducing bias from differences between the original trial populations.
Source: Signorovitch et al. 2010
When is a matching-adjusted indirect comparison used?
A matching-adjusted indirect comparison is used when treatments must be compared indirectly, direct head-to-head evidence is unavailable, and the trials differ in their populations in ways that could bias a standard indirect comparison, provided individual patient data are available for at least one trial and summary data for the other. It is common in health technology assessment when comparing a new treatment with a competitor. So MAIC is applied to adjust for population differences in indirect comparisons, improving comparability when the necessary individual patient data are available for one of the treatments.
Source: Signorovitch et al. 2010
What are the limitations of a matching-adjusted indirect comparison?
The limitations of a matching-adjusted indirect comparison include that it can adjust only for characteristics that are measured and reported, so unobserved differences between the populations remain a source of bias; that reweighting reduces the effective sample size, lowering precision, especially when the populations differ greatly; and that it typically adjusts one population to the other rather than to a common target, complicating interpretation. Its validity depends on including the relevant effect modifiers. So MAIC improves on unadjusted indirect comparisons but remains subject to residual bias and reduced precision, and is interpreted accordingly.
Source: Bucher et al. 1997
How does a matching-adjusted indirect comparison differ from a standard indirect comparison?
A matching-adjusted indirect comparison differs from a standard indirect comparison in that it uses individual patient data to reweight one trial's population to match the other's characteristics, adjusting for differences between the populations, whereas a standard indirect comparison, such as the Bucher method, links trials through a common comparator using only aggregate results, assuming the populations are similar. MAIC therefore addresses population differences that a standard comparison ignores, reducing bias when the trials differ, but it requires individual patient data and reduces precision. So MAIC is a population-adjusted refinement of the standard indirect comparison.
Source: Signorovitch et al. 2010
Trust Record
Verified by Dr Darrin Baines
British health economist
Professional identity: darrinbaines.org
Verification date: 20 Nov 2025
Content version: 1.0.0
Canonical Identity
- Term code
- HE-ES-CER-020
Stable URI · Machine-readable · Resolvable · CC BY 4.0