A retrospective observational study of drotrecogin alfa (activated) in adults with severe sepsis: Comparison with a controlled clinical trial

Objective. To compare characteristics and outcomes of patients treated with drotrecogin alfa (activated) (DrotAA) in clinical practice to those treated in a phase III randomized controlled trial (PROWESS). Design: Observational data were collected retrospectively from patients who received DrotAA as part of physician-directed treatment. Setting. Intensive care units of five teaching institutions. Patients. Patients were >= 18 yrs old, had severe sepsis (confirmed/suspected infection with one or more sepsis-induced organ dysfunctions), and received DrotAA. Interventions. None. Measurements and Main Results. Baseline demographics, severity of illness, time from organ dysfunction onset to DrotAA treatment, daily assessment of organ dysfunction, serious bleeding events, and in-hospital mortality were reported. Timing from severe sepsis documentation to start of DrotAA infusion was categorized: day 0 (same calendar day); day 1 (next calendar day); and day >= 2 (second calendar day or later). Clinical practice patients (n = 274) were younger, had more comorbidities, had higher severity of illness (as measured by organ dysfunction or greater vasopressor/ventilator use), and received DrotAA later than PROWESS patients (all p < .05). Overall hospital mortality for clinical practice patients was 42%, compared with 37% for DrotAA-treated PROWESS patients with Acute Physiology and Chronic Health Evaluation 11 score >= 25. Mortality for day 0, day 1, and day >= 2:2 groups was 33%, 40%, and 52%, respectively. In PROWESS, the vast majority were treated on day 0 or day 1. Serious bleeding events during infusion were noted in 4.0% of clinical practice patients compared with 2.2% of PROWESS DrotAA-treated patients with Acute Physiology and Chronic Health Evaluation II score >= 25. Conclusions: Patients treated in clinical practice differed from those in PROWESS. Patients were younger, had more comorbidities, had greater severity of illness, and had longer mean time from severe sepsis onset to the start of DrotAA. Hospital mortality for patients treated within 1 day of severe sepsis onset was similar to DrotAA-treated PROWESS patients. While the low number of serious bleeding events precludes a definitive assessment, the observed incidence of serious bleeding events in clinical practice patients was numerically higher than in DrotAA-treated PROWESS patients.