OBJECTIVES : Health economic data are often difficult to compare across studies because of the differences in healthcare systems, researchers’ perspectives, and data collection. This raises questions about whether meta-analyses (MAs) can be useful for summarizing such evidence. To explore these issues, we conducted a systematic literature review (SLR) to assess where MAs have been used to synthesize health economic data and challenges associated with this approach.

METHODS : EMBASE and MEDLINE were searched for SLRs and/or MAs published in English from database inception to May 2019 and reporting on health economic outcomes associated with any treatment. SLRs summarizing data only narratively were excluded. Included studies were assessed for whether they considered, attempted or conducted MAs, reported barriers to such analysis, and any overall trends on its use.

RESULTS : Sixteen studies published between 2008 and 2018 met the inclusion criteria. Of these, 11 had completed MAs, while the other five had only considered/attempted such analysis. There were no obvious trends on whether MAs were used to summarize and/or present health economic data. The most commonly meta-analyzed data related to length of inpatient stay (n=8 studies), inpatient treatment (n=6), and hospitalizations/admissions (n=3). Data related to outpatient visits, readmissions, emergency room visits, and costs were meta-analyzed in one study each. All five studies that considered but did not undertake an MA cited heterogeneity or variability in reporting of data to explain this.

CONCLUSIONS : Decision-makers depend heavily on health economic data in reaching conclusions about reimbursement. Therefore, they may benefit from more precise estimates of such data that might be provided by MAs. However, given the paucity of evidence identified by this SLR, future research is needed to determine whether or how MAs can best be used in this setting to provide insights beyond those available from standard qualitative and narrative synthesis.