Abstract
Background:
Non-steroidal anti-inflammatory drugs (NSAIDs) are associated with increased morbidity and mortality in patients with congestive heart failure (CHF). Current guidelines recommend discontinuation of NSAIDs in all patients with CHF, but in clinical practice, many patients remain with active prescriptions. We sought to reduce prevalence of active NSAID prescriptions in a veteran patient population with CHF by implementing an electronic health record (EHR) alert advising against NSAID prescriptions.
Methods:
This single-center quality improvement project was initiated at the Miami Veterans Affairs Medical Center. In patients with any diagnosis of heart failure, when a provider attempted to initiate or renew an NSAID prescription, an EHR alert was activated warning of the potential harms. Providers were required to acknowledge the alert prior to electronic signature. NSAIDs activating the alert included celecoxib, ibuprofen, diclofenac, and naproxen. The primary outcome of interest was the number of patients with CHF and active NSAID prescriptions, assessed 6 months before and after alert implementation. Analysis of the combined long-term secondary outcomes of hospitalization for acute decompensated heart failure and all-cause mortality is ongoing. Relative risk reductions with statistical significance determined by p<0.05 were calculated for both primary and secondary outcomes.
Results:
A total of 144 patients were included in this study. In the 6 months preceding alert implementation, NSAIDs were discontinued in 30.9% (17/55) of patients. At 6 months follow-up after EHR alert initiation, NSAIDs were discontinued or left to expire in 65.2% (58/89) of patients in which the EHR alert was activated. The relative risk of patients with CHF being prescribed NSAIDs was significantly reduced by 49.6% (relative risk=0.504; 95% confidence interval [0.361-0.704], p=0.0001). After intervention, death was reported in 3.2% of patients persisting on NSAID therapy, compared to 1.7% of patients that had NSAIDs discontinued (p=0.65).
Conclusions:
Implementation of an EHR alert advising of the harm of NSAIDs in patients with CHF in a veteran population has resulted in a statistically significant decrease in the number of active NSAID prescriptions. Further study with larger patient populations and extended follow-up will help determine whether these findings are sustainable and lead to a clinically significant reduction in mortality and hospitalizations.