Impact of pilot/feasibility studies in meta-analyses of acupuncture: a meta-epidemiological study

Session Type
Overviews of reviews and scoping reviews
Won J1, Choi S2, Son B3, Lee B4
1Department of Meridian & Acupoint, College of Korean Medicine, Dong-eui University, Korea South
2Department of Obstetrics & Gynecology, College of Korean Medicine, Dong-eui University, Korea South
3Department of Internal Medicine, College of Korean Medicine, Dong-eui University, Korea South
4KM Science Research Division, Korea Institute of Oriental Medicine, Korea South

Background: Pilot/feasibility studies are assumed to be associated with inflated effect size of meta-analyses. And whether acupuncture is used as an adjunctive therapy or not may influence the effect size.
Objectives: This study aims to explore the changes of effect sizes in meta-analyses when adopting different criteria regarding inclusion of pilot/feasibility studies and compare the effect size following usage pattern of acupuncture (alone or as an adjunctive therapy) across therapeutic areas.
Methods: Searches were to identify systematic reviews with meta-analyses of acupuncture from January 2017 to December 2020. The summary effect sizes (ES) were extracted from each meta-analysis. Individual studies included in the meta-analyses were classified into one of the following four categories: sample size (N ≤ 100, N > 100) and usage (acupuncture alone or as an adjunctive therapy). The effect size was re-calculated following the sample size except of small sized study and compared with that of original study. Fixed and random effects models and meta-regressions were estimated. The potential independent variables identified by relevant literatures (types of outcome, acupuncture and conditions) were extracted and analysed.
Results: Until March 2023, searches are finished and relevant studies are under review. Preliminary findings will be reported.
Conclusions: Results of this study can demonstrate the impact of pilot/feasibility studies and usage pattern of acupuncture for effect sizes of meta-analyses.
Patient, public and/or healthcare consumer involvement: none