Why a sepsis bundle that worked in trials failed in eleven district hospitals: an implementation study
Dr. Lakshmi Pillai, Dr. Marco Ferreira, Dr. Farhan Qureshi
Volume 9, issue 22026pp. 88–11438 citations
A sepsis care bundle with strong trial evidence was introduced across eleven district hospitals in two Indian states. Thirty-day mortality did not improve (adjusted OR 0.96, 95% CI 0.81 to 1.14). Rather than report a null result and stop, we conducted a parallel implementation study — 94 interviews, 340 hours of observation and time-stamped process data — to establish what happened. Bundle compliance was 34 per cent, and the binding constraint was not knowledge, motivation or protocol design but laboratory turnaround: the lactate result the bundle is built around arrived after the window in which it could change care in 71 per cent of cases.
