Excessive prolongation of prothrombin time among patients treated with warfarin and admitted to the emergency room

Patients treated with warfarin are uncommonly admitted with excessive prolongation of international normalized ratio (INR) of prothrombin time (PT) to the emergency room (ER). Here we wanted to evaluate the reasons behind this prolongation, clinical and laboratory parameters, course and outcome of these patients.

Consecutive patients admitted to the ER with prolonged PT-INR>6 under warfarin treatment and who gave sufficient information regarding history of warfarin treatment, were followed prospectively. Patients or care-givers were asked about warfarin treatment, PT monitoring, current and new medications especially antibiotics and diet. Clinical and laboratory parameters, course and outcome were also documented. Age and gender matched patients with therapeutic INR level were used as a control group.

75 patients finished the study. The median INR level among the patients was 8.5 and a range of 6.1-59.5. The median duration of warfarin treatment was 12 months among the patients vs. 36 in the control group, p<0.001 and family physician was the sole decider on warfarin dose in 71 patients vs. 49 in the control group, p<0.100. 19 patients adhered to balanced diet vs. 34 in the control group, p=0.01. Dose increment of warfarin was the most common definite cause of prolonged PT followed by adding new medication especially amiodarone and recently started warfarin. Significantly more patients had evidence of bleeding than the control group but no difference in mean hemoglobin level or number of patients transfused with packed cells. No difference in the number of patients hospitalized or duration of hospitalization, but 5 in-hospital fatalities in the patients' group vs. none in the control group.

Excessive PT prolongation is mainly due to warfarin dose increment, adding new medication especially amiodarone and recently started warfarin. It is significantly more associated with bleeding and higher in-hospital mortality.