Promoting tailored healthcare: improving methods to investigate subgroup effects in treatment response when having multiple individual participant datasets

Johannes Reitsma, Maroeska Rovers

Individuals differ in their response to treatment. Personalized or patient-centered healthcare involves tailoring therapeutic decisions to individuals based on patient and disease characteristics. The current widespread initiatives for sharing individual participant data (IPD) will create unique opportunities to investigate whether subgroups differ in their response to treatments. Benefits of using IPD rather than traditional meta-analysis arise from standardization of subgroups and outcomes across studies, the higher validity of subgroup findings through proper adjustment for differences in patient and study characteristics, and the increased possibilities to search for subgroups based on multiple characteristics.

In our recent papers in PlosMed, JAMA and BMJ we however identified several remaining challenges in IPD meta-analyses leading to diversity and uncertainty on how to perform and interpret subgroup analyses based on IPD. These gaps include: (i) lack of insight and evidence why and when various distinct statistical subgroup approaches lead to conflicting results; (ii) uncertainty how to combine IPD and aggregate data; (iii) unclear how to incorporate differences in design and population between studies; (iv) uncertainty how to best calculate the probability that a new trial will find the subgroup of interest. In view of the expected rise in open access data, guidance on appropriate methodology is urgently needed to improve the use and uptake of IPD meta-analyses and their findings.

Our overall aim is to evaluate and improve prevailing approaches, and to develop novel methods where needed, for investigating and interpreting subgroup effects in treatment response when multiple IPD sets are available. It should lead to more credible evidence about whether a relevant subgroup effect exists.



Debray TPA, de Jong VMT, Moons KGM, Riley RD. Evidence synthesis in prognosis research. Diagnostic and Prognostic Research 2019:3;13.

Riley RD, Moons K, Snell KIE, Ensor J, Hooft L, Altman D, Hayden J, Collins GS, Debray TP. A guide to systematic review and meta-analysis of prognostic factor studies. BMJ 2019:364;k4597.

Nguyen TL, Debray TP. The use of Prognostic Scores for Causal Inference with General Treatment Regimes. Stat Med 2019:38;2013-29.

van Hoorn R, Tummers M, Booth A, Gerhardus A, Rehfuess E, Hind D, Bossuyt PM, Welch V, Debray TP, Underwood M, Cuijpers P, Kraemer H, van der Wilt GJ, Kievit W. The development of CHAMP: a checklist for the appraisal of moderators and predictors. BMC Med Res Methodol 2017:173.

Audigier V, White IR, Jolani S, Debray TPA, Quartagno M, Carpenter J, van Buuren S, Resche-Rigon M. Multiple imputation for multilevel data with continuous and binary variables. Stat Sci 2018:33;160-83.

The Netherlands Organisation for Health Research and Development

Source of Funding

The Netherlands Organisation for Scientific Research supports a strong system of sciences in the Netherlands by encouraging quality and innovation in science. Our conviction is that scientific research contributes to our prosperity and well-being and that it provides for our growing need for knowledge: for facing societal challenges, for economic development and to better understand ourselves and the world.

Lorem Ipsum is simply dummy text of the printing and typesetting industry. Lorem Ipsum has been the industry's standard dummy text ever since the 1500s, when an unknown printer took a galley, Lorem Ipsum is simply dummy text of the printing and typesetting industry.