Conducting Meta-analyses and Systematic Reviews in Long COVID: Challenges and Lessons Learned 4353

Abstract

Objectives

To identify and address the methodological challenges and lessons learned in conducting meta-analyses and systematic reviews in the emerging field of long coronavirus disease 2019 (COVID).

Design

This study results from a larger systematic review in response to methodological difficulties, specifically related to participant inclusion/exclusion criteria, sample sizes and study comparability issues. Forty-one studies were included.

Setting

The study synthesized evidence from a wide range of settings, including hospital and outpatient clinics.

Participants

The intended participants were individuals suffering from long COVID, identified through variable inclusion/exclusion criteria across studies, with a focus on cognitive and psychological symptoms. Challenges were noted in the heterogeneous application of these criteria and in the variable sample sizes across studies.

Interventions

Five studies were included that reported on interventions aiming at alleviating cognitive and psychological symptoms in long COVID, although very few randomized controlled trials were identified, highlighting a gap in the literature.

Main Outcome Measures

Outcome measures varied across studies but generally included assessments of cognitive and psychological functioning through neuroimaging, neurophysiology, and clinical scales.

Results

Significant challenges were observed, including issues related to the Population, Intervention, Comparison, Outcomes framework, such as unequal sample sizes and inappropriate comparisons between large samples of neurotypical participants and small samples of long COVID participants. Additionally, the inclusion/exclusion criteria for long COVID were inconsistently applied, with few studies adhering to Centers for Disease Control and Prevention or World Health Organization guidelines.

Conclusions

These findings underscore the complexity of researching long COVID, particularly in synthesizing evidence across studies with varied methodologies, sample sizes, and clinical definitions. The findings highlight the need for standardized criteria for defining neuropsychological symptoms associated with long COVID and for designing studies with methodologies that can be more readily compared and aggregated in systematic reviews and meta-analyses. Guidelines for future research projects on long COVID will be discussed in more detail.

Disclosures

Eva Pettemeridou is acting as the Communications and Social Media Officer of the International Interdisciplinary Special Interest Group of the American Congress of Rehabilitation Medicine. All authors receive a salary from a joint research grant conducting work on long COVID.