Comment on—The effectiveness of positive psychological interventions for patients with cancer: A systematic review and meta-analysis

Abstract

We have carefully read the recent paper ‘The effectiveness of positive psychological interventions for patients with cancer: A systematic review and meta-analysis’ by Tian et al. (Tian et al., 2024) in the Journal of Clinical Nursing and have some concerns regarding the ‘3.7 Data Synthesis’ section.

The authors stated:Heterogeneity was determined using the I² value statistic test. If I² ≤50%, heterogeneity between studies was considered low and a fixed-effects model was used; and if I² >50%, heterogeneity between studies was high and a random-effects model was used.

In reality, for data analysis, due to the variations of complex interventions, we should only use a random-effects model (Borenstein et al., 2010; Dettori et al., 2022). The strategy of starting with a fixed-effect model and then moving to a random-effects model if the test for heterogeneity is significant relies on a flawed logic and should be strongly discouraged. The selection of a model should be based solely on the question of which model fits the distribution of effect sizes and thus takes account of the relevant source(s) of error. When studies are gathered from the published literature, the random-effects model is generally a more plausible match. The fixed-effect model is based on the assumption that all studies in the meta-analysis share a common (true) effect size. Under the fixed-effect model, the summary effect is an estimate of the effect which is common to all studies in the analysis. Under the random-effects model, the summary effect is an estimate of the mean of a distribution of true effects.

Additionally, the Cochrane Handbook for Systematic Reviews of Interventions Version 6.4, 2023 (Higgins et al., 2023) indicates that the decision between fixed- and random-effects meta-analyses has been the subject of much debate, and no universal recommendation is provided. The handbook outlines six key considerations for making this choice, with the third point stating:

‘Under any interpretation, a fixed-effect meta-analysis ignores heterogeneity. If the method is used, it is therefore important to supplement it with a statistical investigation of the extent of heterogeneity.’

In section 4.5.2 of the ‘Hope’ chapter (Figures), the author employed a fixed-effects model. By utilizing Review Manager 5.4.1 and the dataset provided by the author, we conducted an analysis using a random-effects model. The results indicate a slight variance in the test for overall effect, as depicted in Figures 1 and 2, which could lead to an imprecise estimation of effect size and subsequently affect our confidence in the evidence.

Therefore, we suggest that using a random-effects model is a more reasonable approach. We welcome the author to engage in potential follow-up discussions with us.

None.

None.