Electronic risk alerts to improve primary care management of chest pain: a randomized, controlled trial

Abstract

Background: The primary care evaluation of chest pain represents a significant diagnostic challenge.

Objective: To determine if electronic alerts to physicians can improve the quality and safety of chest pain evaluations.

Design and participants: Randomized, controlled trial conducted between November 2008 and January 2010 among 292 primary care clinicians caring for 7,083 adult patients with chest pain and no history of cardiovascular disease.

Intervention: Clinicians received alerts within the electronic health record during office visits for chest pain. One alert recommended performance of an electrocardiogram and administration of aspirin for high risk patients (Framingham Risk Score (FRS) ≥ 10%), and a second alert recommended against performance of cardiac stress testing for low risk patients (FRS < 10%).

Main measures: The primary outcomes included performance of an electrocardiogram and administration of aspirin therapy for high risk patients; and avoidance of cardiac stress testing for low risk patients.

Key results: The majority (81%) of patients with chest pain were classified as low risk. High risk patients were more likely than low risk patients to be evaluated in the emergency department (11% versus 5%, p < 0.01) and to be hospitalized (7% versus 3%, p < 0.01). Acute myocardial infarction occurred among 26 (0.4%) patients, more commonly among high risk compared to low risk patients (1.1% versus 0.2%, p < 0.01). Among high risk patients, there was no difference between the intervention and control groups in rates of performing electrocardiograms (51% versus 48%, p = 0.33) or administering aspirin (20% versus 18%, p = 0.43). Among low risk patients, there was no difference between intervention and control groups in rates of cardiac stress testing (10% versus 9%, p = 0.40).

Conclusions: Primary care management of chest pain is suboptimal for both high and low risk patients. Electronic alerts do not increase risk-appropriate care for these patients.

Trial registration: ClinicalTrials.gov NCT00674375.

Figure 1

CONSORT diagram of patient and physician eligibility and randomization. Physicians were randomized within health centers according to volume of patients with chest pain evaluated in the prior 6 months.

Figure 2

Active electronic reminders were delivered to physicians during office encounters, and facilitated electronic ordering of recommended tests.