Presentation approaches for enhancing interpretability of patient-reported outcomes measures in meta-analyses: A systematic survey of Cochrane reviews

Article type
Authors
Zeng L1, S Devji TSD2, Wang Y2, Yao L2, Han M3, K Nampo F4, Granholm A5, Tadayon B2, Manja V6, EI-Khechen H2, C Johnston B2, Bhandari M2, Thabane L2, L Patrick D7, H Guyatt G2
1West China Second University Hospital, Sichuan University
2McMaster University
3Chosun University
4Federal University of Latin-American Integration
5Copenhagen University Hospital
6University of California Davis
7University of Washington
Abstract
Background:
Interpreting the magnitude and importance of pooled effects in meta-analysis can be challenging, particularly for patient-reported outcomes (PROs) measured using questionnaires with which clinicians have limited familiarity.

Objectives:
We are evaluating Cochrane systematic review authors’ approaches to calculation, presentation and interpretation of PRO measures (PROMs) results in meta-analyses.

Methods:
Our methodological survey includes Cochrane systematic reviews that report at least one statistically significant result for at least one PRO measured by a PROM reported as a continuous outcome. We are including 200 eligible Cochrane systematic reviews as planned in our published protocol. We started from Cochrane systematic reviews published in April 2019 and screened backwards, stopping at October 2015 when we achieved the target sample size. We are documenting authors’ approach to calculating pooled effects when studies used different PROMs for the same construct (standardized mean difference, natural units of the most widely used instrument, dichotomization with relative effects, dichotomization with absolute effect, minimally important difference [MID] units) and their approach to interpreting the importance or magnitude of the pooled effects (e.g. reference to the MID). We are assessing whether an MID was applied in the calculation, presentation and interpretation of PRO measures.

Results:
We plan to finish data analysis in June.

Discussion:
Our methodological survey will inform the systematic review community regarding the current practice of summarizing and presenting effect estimates for PROMs presented as continuous variables in Cochrane systematic reviews. We anticipate an underuse of the most informative available methods, and possible deficiencies in interpretation. We anticipate our results will influence recommendations on reporting, conduct and interpretation of PROMs and subsequent practice.

Patient or healthcare consumer involvement: Not involved