Depression Screening Tool Accuracy Individual Participant Data Meta-analyses: Data Contribution was Associated with Multiple Factors

Yin Wu,Ying Sun,Yi Liu,Brooke Levis,Ankur Krishnan,Chen He,Dipika Neupane,Scott B Patten,Pim Cuijpers,Roy C Ziegelstein,Andrea Benedetti,Brett D Thombs,Scott B. Patten,Roy C. Ziegelstein,Brett D. Thombs

DOI: https://doi.org/10.1016/j.jclinepi.2023.08.006

IF: 7.407

2023-08-24

Journal of Clinical Epidemiology

Abstract:Objective To examine the proportion of eligible primary studies that contributed data, study characteristics associated with data contribution, and reasons for non-contribution using diagnostic test accuracy individual participant data meta-analysis (IPDMA) datasets from the DEPRESsion Screening Data project. Study Design and Setting We reviewed dataset contributions from four IPDMAs. A multivariable logistic regression model was fitted to evaluate study factors associated with data contribution. Results Of 456 eligible studies from four included IPDMAs, 295 (65%) contributed data. More recent year of publication and higher journal impact factor were associated with greater odds of data contribution. Studies conducted in Europe (excluding the United Kingdom), Oceania, Canada, the Middle East, Africa, Central or South America (reference=the United States), recruited in inpatient care or non-medical settings (reference=outpatient), reported screening accuracy results, or drew negative conclusions (reference=positive conclusions) were more likely to contribute data. Studies of the Geriatric Depression Scale (reference=the Patient Health Questionnaire) or lacking funding information were negatively associated with data contribution. Over 80% of non-contributions were due to authors being unreachable or data being unavailable. Conclusion The study identified factors associated with data contribution that may support future research to promote data contribution to IPDMAs.

public, environmental & occupational health,health care sciences & services