BOSTON – A medication management strategy to minimize the effect of drug-drug interactions (DDIs) between warfarin and common antimicrobials resulted in significantly greater time within therapeutic range for anticoagulation, as well as a numerically smaller, but nonsignificant, number of bleeding events.
After implementation of a comprehensive inpatient and postdischarge guideline to manage DDIs between warfarin and 16 antibiotics, antivirals, or antifungal medications, patients’ in-hospital time within therapeutic range (TWTR) increased to 72% from 50% preimplementation (P = .043). Warfarin TWTR also improved across care transitions after the guidelines were implemented, rising to 70% from 46% (P = .012). No bleeding events occurred in the group studied after the guidelines were instituted, compared with four events in the comparator preguidelines group (P = .11).
Nghi Ha, PharmD, MPH, and his collaborators sought to determine whether formalizing a process to manage potentially dangerous antimicrobial-warfarin DDIs made a difference in achieving more TWTR for patients, as determined by international normalized ratio (INR) values. Dr. Ha, a clinical pharmacist at University of Michigan Health System, Ann Arbor, presented the results during a poster session at the annual meeting of the American Society for Microbiology.
Secondary outcome measures studied by Dr. Ha and his associates included the incidence of thrombosis or major bleeding events, as well as tracking documentation of medications and the anticoagulation plan in progress and discharge notes.
Patients were included if they were at least 18 years old, and if they were on 3 days or more of an antimicrobial with potential for DDI with warfarin. Patients who were also newly on other medications with the potential for significant DDI with warfarin were excluded to minimize the potential for confounding.
Dr. Ha and his collaborators characterized the study as a retrospective, single-center, quasi-experimental design of a pharmacist-run anticoagulation service. The study examined endpoints before and after implementation of comprehensive guidelines, and included 78 preguideline and 31 postguideline patients.
The guidelines drafted by the investigators and tested in their study included empiric adjustment of warfarin dosing for patients who were placed on an antibiotic with high potential to increase INR. These included many azoles and sulfamethoxazole/trimethoprim, for which initial warfarin doses were empirically reduced 20%-30% for patients whose INRs were therapeutic at the start of antimicrobial therapy. For ciprofloxacin, erythromycin, clarithromycin, and isoniazid, the guidelines recommended initial empiric warfarin dose reductions of 10%-15%.
Patients whose INRs were subtherapeutic at the beginning of therapy and who received these antimicrobials were continued on their maintenance warfarin dosing, but were monitored for rising INRs over the first 48 hours, for consideration of dosing adjustment as needed. Individuals with supratherapeutic INRs at the beginning of antimicrobial dosing had their warfarin doses reduced or held by a more aggressive algorithm based on their initial INR, and based on the potential of the antimicrobial to increase INR.
On discharge, patients were either reverted to their previous warfarin regimen if they had been stable on that regimen, or had their inpatient warfarin dosing increased by 10%-20%.
Drugs that were deemed to have moderate potential to increase INR included doxycycline, levofloxacin, moxifloxacin, quinupristin/dalfopristin, telaprevir, boceprevir, and simeprevir. For these medications, the protocol recommended no initial dose adjustment, but recommended monitoring of INR to consider a dose reduction if needed. On hospital discharge, patients who had been on these medications were to resume their previous warfarin dosing.
Antimicrobials with potential to decrease INR included nafcillin, for which the protocol recommended empiric warfarin dose increases of 25%-50%, starting 3-5 days after nafcillin was begun. Patients on rifampin or rifabutin were to increase their warfarin by 20%-30%, also 3-5 days after beginning the antibiotics. Patients on ritonavir alone, or any protease inhibitor given for HIV along with ritonavir, were closely monitored, but no empiric dosing adjustments were made.
Patients with initial subtherapeutic INRs had dosing increased by 30%-50% for nafcillin and 20%-30% for rifampin and rifabutin. A stepped algorithm for dose adjustment or withholding was also developed for these medications to treat patients with initial supratherapeutic INRs. Patients on these medications were to resume their previous warfarin dosing, with monitoring and adjustment if they had not been previously stable.
Documentation of antimicrobial-warfarin DDI in the anticoagulation service discharge summary improved significantly once the guidelines were implemented (40% compared with 14%, P = .02). There was not a significant improvement in DDI documentation in daily progress notes.
The comprehensive intervention included the formulation of guidelines and requirements to document the medication interaction in the medical chart. Other interventions included training for clinical pharmacists and the development of pocket cards and flyers to educate team members about the new guidelines. The electronic health record had triggers built and implemented to prompt consideration of warfarin/antimicrobial DDIs as well.