Graduate Research Assistant University of North Carolina at Chapel Hill Chapel Hill, United States
Background: Outpatient parenteral antimicrobial therapy (OPAT) programs enable patients with serious infections to continue treatment after hospital discharge. These patients are often medically complex and require monitoring and dose/medication adjustments to treatment over time to reduce risk of adverse drug events.
Objectives: To estimate 10-week cumulative incidence of unplanned discontinuation from OPAT using the Aalen-Johansen estimator, and to compare unanticipated stopping across infection diagnoses and antimicrobial medications.
Methods: We conducted a cohort study of patients enrolled in the University of North Carolina Medical Center OPAT program between March 2015 and December 2022. We prospectively collected data for the planned date of OPAT completion for each patient (entered at treatment start in the EHR). Patients were followed from their OPAT regimen initiation (day 0) until their observed treatment stop date, using a grace period of 5 days to confirm discontinuation (rather than a pause). We classified each observed discontinuation as “planned,” “unanticipated early stop,” or “unanticipated prolonged treatment” based on a +/-5 day grace period around the planned OPAT completion date. We estimated cumulative incidence of each type of discontinuation using Aalen-Johansen estimator to account for competing events (e.g., early stopping precludes both planned stops and prolonged treatment). We compared unanticipated early stopping and prolonged treatment by infection diagnosis and initial antimicrobial medication.
Results: The cohort included 2057 unique patients who experienced 2350 OPAT courses; 59.7% were male and median age was 56 years (IQR 45-66). Osteomyelitis was the most common infection (43.7%); 30.9% of initial treatment regimens included vancomycin. Median planned OPAT duration was 36 days (5th-95th centiles 15-44). Overall cumulative incidence of unanticipated early discontinuation was 11.5%; across infection diagnoses, cumulative incidence ranged 6.7% (skin/soft-tissue infection) to 17.4% (combined diabetic foot infection and osteomyelitis); and across initial treatments, from 4.6% (vancomycin+rifampin) to 15.9% (other vancomycin combinations). Cumulative incidence of unplanned prolonged therapy occurred was 20.4%, ranging across infection diagnoses from 5.4% (endocarditis) to 30.5% (skin/soft-tissue infection) and 16.1% (vancomycin+cephalosporin) to 31.2% (piperacillin-tazobactam).
Conclusions: One-third of patients discontinued OPAT >5 days earlier or later than planned. Unplanned discontinuation varied considerably across infection and initial medication regimen, which belies potential opportunities to avoid adverse drug reactions and side effects that often burden patients in OPAT.
