ISPOR Europe 2018
Barcelona, Spain
November, 2018
Multiple Diseases/No Specific Disease
Research on Methods (RM)
Statistical Methods (SM)
Hatswell A1, Freemantle N2, Baio G2
1Delta Hat Limited and University College of London, Nottingham, UK, 2University College London, London, UK

Where cross trial comparisons are required, differences in patient populations have the potential to bias comparisons. Whilst methods such as propensity scoring exist to match studies where individual level data are available, until recently, no such analysis was possible where only aggregate level data were accessible for the historical study. Matching Adjusted Indirect Comparison (MAIC) attempts to address this issue by weighting patients in the contemporary study for which individual patient characteristics are available, to match aggregate characteristics with those observed in the historical study.


We conducted a simulation study with a large number of patients (10,000) on control and intervention, with outcomes determined by 12 covariates. 6 of these were in balance, with 6 on average more favourable in the intervention arm - with the intervention also assumed to have a positive effect. Various scenarios were tested, and the level of error from the ‘true’ difference calculated both for naïve comparisons between the simulated arms, and the error remaining after MAIC.


MAIC did eliminate the bias in our simulation compared to a naïve comparison, however the error around the estimate showed a distribution i.e. although patient characteristics gave an overly favourable impression of the interventions effectiveness, MAIC was as likely to underestimate the true effect, as it was to overestimate it. Whilst MAIC estimates were not completely accurate, they represented a marked improvement (approximately 60%) on naïve comparisons under ideal conditions.


Methods such as MAIC allow comparisons between studies with no control arms, or with differences in control arms. Understanding the mechanisms and resulting outputs however are key to their appropriate use. In this study we show that MAIC reduces the error in comparisons, but is not a perfect method. Careful consideration of the implicit assumptions, and interpretation of results is warranted.