PRP070: Depression screening outcomes related to launch of best practice advisory (BPA) in EMR at an academic family medicine clinic

Aman Luthra, MD; Elizabeth Mukherjee; Gurneet Matharoo, MD


Context: PHQ2/9 is a module commonly employed in primary care setting as a screening tool for depression. Depression being the most prevalent psychiatric condition seen in primary care, thus outcomes are highly dependent on appropriate therapy, follow up and proper documentation.
Objective: We aimed to look at the utility of PHQ2/9 questionnaire at an academic Family Medicine clinic in improving patient outcomes through use of a Best Practice Advisory (BPA) launch in EMR EPIC.
Study Design: A qualitative informatics study conducted at an Academic Family Medicine clinic.
Population studied: A randomized chart review was performed of 281 clinic patients >18 years old with documented PHQ2/9 scores.
Intervention: Data was grouped according to pre and post launch date of BPA, which aimed to improve depression screening documentation beginning 8/7/2019.
Outcome measures: Data was utilized to determine whether BPA launch fostered a positive correlation for the following parameters: documentation of PHQ2/9 scores or depression symptoms in clinic note, interventions selected on the EPIC “screening” tab, mental health problem list update, and arrangement of behavioral health/psychiatry follow up within 3 months. Analysis was performed using pre and post proportions. Chi-squared test was used to evaluate confidence intervals for proportions which allowed to further determine statistical significance. A p value of <0.05 was used for statistical significance.
Results: Analysis of 73/281 charts revealed a 15.9% increase in documentation of PHQ9 scores or symptoms (Pre 29% vs. Post 44.9 %, X2 0.24, p 0.6); a 10% increase in documentation of interventions selected on screening tab (Pre 25% vs Post 35%, X2 0.37, p 0.5); an 7.9% increase in mental health problem list update (Pre 65.6% bs Post 73.5%, X2 1.79, p 0.2); and an 8% decrease in arrangement of behavioral health/psychiatry follow up within 3 months (Pre 54% vs Post 46%, X2 2.52, p 0.1).
Expected Outcomes: Even though not statistically significant, a positive correlation was observed for three out of four of the metrics we aimed to improve. With increase in sample size, we believe statistical significance may be achieved. This study could be pivotal in improving depression intervention through the use of bio-informatics.

Social Media


11400 Tomahawk Creek Parkway
Leawood, KS 66211