Why a sepsis bundle that worked in trials failed in eleven district hospitals: an implementation study

Abstract

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.

Keywords

• sepsis
• implementation research
• health systems
• quality improvement
• district hospitals

References

1. 1.Rhodes, A., et al. (2017). Surviving Sepsis Campaign guidelines 2016. Intensive Care Medicine, 43(3), 304–377.
2. 2.Damschroder, L. J., et al. (2022). The updated Consolidated Framework for Implementation Research. Implementation Science, 17, 75.
3. 3.Tong, A., Sainsbury, P., & Craig, J. (2007). COREQ: a 32-item checklist. International Journal for Quality in Health Care, 19(6), 349–357.
4. 4.Kruk, M. E., et al. (2018). High-quality health systems in the SDG era. The Lancet Global Health, 6(11), e1196–e1252.

Cite this article

Pillai, L., Ferreira, M., Qureshi, F. (2026). Why a sepsis bundle that worked in trials failed in eleven district hospitals: an implementation study. Lighthouse Journal of Public Health, 9(2), 88–114. https://doi.org/10.53456/ljph.2026.9.2.88