A Randomized Trial Of A Personalized Clinical Decision Support Intervention To Improve Statin Prescribing In Patients With Atherosclerotic Cardiovascular Disease (PCDS Statin)

Presented at the ACC.23/WCC by: Prof. Salim Virani, MD, PhD - Houston, TX, USA

Introduction and methods

Use of statins, and specifically of high-intensity statin (HIS), remains low in ASCVD patients, due to factors such as statin-associated side effects (SASE) and therapeutic inertia.

The aim of this study was to investigate whether personalized reminders improve use of HIS in patients with ASCVD.

A cluster-RCT was conducted in which an intervention that included reminders was investigated. The intervention was developed by constructing machine learning algorithms using structured and unstructured data via natural language processing (NLP) to identify SASEs and by performing qualitative interviews to determine patient perspectives on SASEs and clinical information needs.

The intervention included reminders individualized to each patient and sent to their primary care clinicians 2-7 days before the patient’s visit or outside of the primary care visit. No more than 3 unsigned alerts were sent out to clinicians before they received more reminders to prevent alert fatigue.

Usual care was access to patient dashboard with statin guideline compliance assessment.

27 Primary care clinics including 36,641 patients were randomized to intervention or usual care. The trial started August 2021 and ended November 2022.

Primary outcome was pre-post change in HIS use, and secondary outcome was change in statin use between intervention and usual-care sites.

Main results

• Baseline HIS use was 53.6% and 55.9% at the intervention and usual care sites, respectively. At the end of the study, this was 55.2% and 53.7%, respectively, with a between group difference of 3.8%, 95%CI:3.7-3.9% (OR for HIS use with intervention 1.06, 95%CI:1.02-1.11).
• In the intervention arm, absolute change in HIS was 10.1% for those who received a reminder vs. -0.18% in those who did not receive a reminder (number needed to remind=18).
• Increase in HIS in the intervention arm was 9.1% and 10.9% in patients with or without SASE, respectively (P=0.02).
• Statin use decreased at intervention (81.1% to 78.7%) and usual care sites (82% to 76.8%).
• Proportion of patients with proportion of day covered (PDC) ≥0.8 increased in the intervention sites from 60.8% to 67.2% and in the usual care sites from 54.4% to 58% (between group difference of 2.8%, OR for PDC ≥0.8=1.38, 95%CI:1.32-1.45).

Conclusion

In this multisite study, structured data and NLP were leveraged to provide individualized information to clinicians. These centrally-processed individualized reminders to clinicians resulted in an increase in HIS use and statin adherence in patients with ASCVD.

- Our reporting is based on the information provided at the ACC.23/WCC -

The findings of the PCDS Statin study were simultaneously published in Circulation