Risk of Bias from Inclusion of Currently Diagnosed or Treated Patients in Studies of Depression Screening Tool Accuracy: A Cross-Sectional Analysis of Recently Published Primary Studies and Meta-Analyses

Danielle B Rice; Brett D Thombs

doi:10.1371/journal.pone.0150067

Risk of Bias from Inclusion of Currently Diagnosed or Treated Patients in Studies of Depression Screening Tool Accuracy: A Cross-Sectional Analysis of Recently Published Primary Studies and Meta-Analyses

PLoS One. 2016 Feb 26;11(2):e0150067. doi: 10.1371/journal.pone.0150067. eCollection 2016.

Authors

Danielle B Rice^{1

2}, Brett D Thombs^{1

2

3

4

5

6

7}

Affiliations

¹ Lady Davis Institute for Medical Research, Jewish General Hospital, Montréal, Québec, Canada.
² Department of Psychiatry, McGill University, Montréal, Québec, Canada.
³ Department of Epidemiology, Biostatistics, and Occupational Health, McGill University, Montréal, Québec, Canada.
⁴ Department of Psychology, McGill University, Montréal, Québec, Canada.
⁵ Department of Medicine, McGill University, Montréal, Québec, Canada.
⁶ Department of Educational and Counselling Psychology, McGill University, Montréal, Québec, Canada.
⁷ School of Nursing, McGill University, Montréal, Québec, Canada.

Abstract

Background: Depression screening can improve upon usual care only if screening tools accurately identify depressed patients who would not otherwise be recognized by healthcare providers. Inclusion of patients already being treated for depression in studies of screening tool accuracy would inflate estimates of screening accuracy and yield. The present study investigated (1) the proportion of primary studies of depression screening tool accuracy that were recently published in journals listed in MEDLINE, which appropriately excluded currently diagnosed or treated patients; and (2) whether recently published meta-analyses identified the inclusion of currently diagnosed or treated patients as a potential source of bias.

Methods: MEDLINE was searched from January 1, 2013 through March 27, 2015 for primary studies and meta-analyses on depression screening tool accuracy.

Results: Only 5 of 89 (5.6%) primary studies excluded currently diagnosed or treated patients from any analyses and only 3 (3.4%) from main analyses. In 3 studies that reported the number of patients excluded due to current treatment, the number of excluded patients was more than twice the number of newly identified depression cases. None of 5 meta-analyses identified the inclusion of currently diagnosed and treated patients as a potential source of bias.

Conclusions: The inclusion of currently diagnosed and treated patients in studies of depression screening tool accuracy is a problem that limits the applicability of research findings for actual clinical practice. Studies are needed that evaluate the diagnostic accuracy of depression screening tools among only untreated patients who would potentially be screened in practice.

MeSH terms

Cross-Sectional Studies
Depression / diagnosis*
Depressive Disorder / diagnosis*
Humans
Mass Screening
Meta-Analysis as Topic
Patient Selection*
Psychiatric Status Rating Scales
Psychometrics
Research Design*

Grants and funding

Ms. Rice was supported by a CIHR Frederick Banting and Charles Best Canada Graduate Scholarship. Dr. Thombs was supported by an Investigator Salary Award from the Arthritis Society. There was no specific funding for this study, and no funders had any role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. The authors had full access to the data and can take responsibility for the integrity of the data and the accuracy of the data analysis.