Implementing an EMR-Based Referral for Smoking Quitline Services with Additional Provider Education, a Cluster-Randomized Trial.

BACKGROUND: Despite evidence of their effectiveness, free smoking quitlines are underused. The best way to educate providers about and encourage use of quitlines is not established. We examined if electronic medical record (EMR)-integrated best practices alerts (BPAs) with or without additional provider education resulted in increased quitline referrals. METHODS: Waitlist-controlled, cluster-randomized trial of primary care practices assigned to three arms. Providers in participating sites received a new EMR-based BPA for quitline referral and additional education outreach visits, the BPA alone, or usual care. The study was conducted in 2 phases: phase 1 from April 17 to October 16, 2017, and phase 2 from November 9, 2017, to May 8, 2018. In phase 2, the usual-care sites were randomized to either of the two intervention arms. The unit of randomization was primary care practice site. All in-office, primary care provider visits with smokers were included. The primary outcome was referral to the quitline. Secondary outcomes included patient acceptance and enrollment in quitline services. RESULTS: Twenty-two practice sites were enrolled. Smoking prevalence at sites ranged from 4.4 to 23%. In phase 1, the BPA-plus-education arm had 5636 eligible encounters and 405 referrals (referral rate 7.2%) while the BPA-only arm had 6857 eligible encounters and 623 referrals (referral rate 9.1%). The usual-care arm had 7434 encounters but no referrals. Comparing the BPA-plus arm to the BPA-only arm, the odds ratio of referral was 0.76 (CI 0.3-1.8). In phase 2, the combined BPA-plus-education sites had 8516 eligible encounters and 475 referrals (rate 5.6%). The BPA-only sites had 9134 eligible encounters and 470 referrals (rate 5.2%). The odds ratio comparing the 2 groups in phase 2 was 1.06 (0.5-2.2). CONCLUSIONS: An EMR-based BPA can improve the number of referrals to quitline services, though more work is needed to improve providers' use of quitlines and low patient acceptance of services. Trial Registration NIH Clinicaltrials.gov identifier: NCT03229356.