Abstract
Background
Screening for mental health problems has been shown to be effective to detect depression and initiate treatment in primary care. Current guidelines recommend periodic screening for depression and anxiety. This study examines the association of patient sociodemographic factors and clinic characteristics on mental health screening in primary care.
Design
In this retrospective cohort study, electronic medical record (EMR) data from a 14-month period from 10/15/2021 to 12/14/2022 were analyzed. Data were retrieved from 18 primary care clinics from the Corewell Health healthcare system in West Michigan. The main outcome was documentation of any Patient Health Questionnaire (PHQ-4/PHQ-9/GAD-7) screening in the EMR within the 14-month period at patient level. General linear regression models with logit link function were used to assess adjusted odds ratio (aOR) of having a documented screening.
Results
In total, 126,306 unique patients aged 16 years or older with a total of 291,789 encounters were included. The prevalence of 14-month screening was 79.8% (95% CI, 79.6–80.0). Regression analyses revealed higher screening odds for patients of smaller clinics (<5,000 patients, aOR 1.88; 95% CI 1.80–1.98 vs. clinics >10.000 patients), clinics in areas with mental health provider shortages (aOR 1.69; 95% CI 1.62–1.77), frequent visits (aOR 1.80; 95% CI, 1.78–1.83), and having an annual physical / well child visit encounter (aOR 1.52; 95% CI, 1.47–1.57). Smaller positive effect sizes were also found for male sex, Black or African American race, Asian race, Latinx ethnicity (ref. White/Caucasians), and having insurance through Medicaid (ref. other private insurance).
Discussion
The 14-month mental health screening rates have been shown to be significantly lower among patients with infrequent visits seeking care in larger clinics and available mental health resources in the community. Introducing and incentivizing mandatory mental health screening protocols in annual well visits, are viable options to increase screening rates.
Funder
Peter C. and Pat Cook endowed Clinical Research Fellow
NIH/NCCIH
Michigan State University
Publisher
Public Library of Science (PLoS)