Predictors of warfarin-associated adverse events in hospitalized patients: Opportunities to prevent patient harm.

Background: The optimum international normalized ratio (INR) monitoring frequency for hospitalized patients receiving warfarin is unknown.

Objective: Assess relationship between daily versus less frequent INR monitoring and overanticoagulation and warfarin-related adverse events.

Design: Retrospective cohort study using Medicare Patient Safety Monitoring System data.

Setting: Randomly selected acute care hospitals across the United States.

Patients: Patients hospitalized from 2009 to 2013 for pneumonia, acute cardiac disease, or surgery who received warfarin.

Interventions: None.

Measurements: (1) Association between frequency of INR monitoring and an INR ≥6.0 or warfarin-related adverse event. (2) Association between the rate of change of the INR and a subsequent INR ≥5.0 and ≥6.0.

Results: Among 8529 patients who received warfarin for ≥3 days, for 1549 (18.2%) the INR was not measured on 2 or more days. These patients had higher propensity-adjusted odds ratios (ORs) of having a warfarin-associated adverse event (OR: 1.48, 95% confidence interval [CI]: 1.02-2.17) for cardiac patients and surgical patients (OR: 1.73, 95% CI: 1.20-2.48), with no significant association for pneumonia patients. Cardiac and pneumonia patients with 1 day or more without an INR measurement had higher propensity-adjusted ORs of having an INR ≥6.0 (OR: 1.61, 95% CI: 1.07-2.41 and OR: 1.92, 95% CI: 1.36-2.71, respectively). A 1-day increase in the INR of ≥0.9 occurred in 621 patients (12.5%) and predicted a subsequent INR of ≥6.0 (positive likelihood ratio of 4.2).

Conclusion: Daily INR measurement and recognition of a rapidly rising INR might decrease the frequency of warfarin-associated adverse events in hospitalized patients.