Updating a Sepsis Bundle: Synthesis and Implementation
Your 450-bed hospital’s sepsis committee is revising its bundle. Baseline door-to-antibiotic time is 45 minutes, bundle compliance is 68%, and sepsis mortality is 18%. The team reviews the Surviving Sepsis Campaign guideline (strong recommendation for immediate antimicrobials in septic shock; weak for possible sepsis without shock), one RCT of 1,200 patients with qSOFA ≥2 showing mortality 21% vs 25%, one RCT of 800 patients showing no mortality difference, and a 4,000-patient cluster RCT showing no mortality difference but a 1.2-day reduction in length of stay. The EHR currently fires 12 sepsis alerts per clinician per shift; nurses report alert fatigue, and the antibiotic stewardship team worries about broad-spectrum overuse. The committee must present a recommendation to the medical executive committee in three weeks. Leadership wants a measurable implementation and evaluation plan, not just a policy update.
1. Synthesize the evidence into a recommendation. Specify the population, intervention, certainty of evidence, strength of recommendation, and how you handled inconsistency.
Read the worked answer
The evidence is consistent for septic shock and uncertain for possible sepsis without shock. The guideline’s strong recommendation for immediate antimicrobials in septic shock is supported by indirect and observational evidence and should be adopted as a mandatory bundle element. For possible sepsis without shock, the 1,200-patient RCT with qSOFA ≥2 suggests benefit, but the 800-patient RCT shows no mortality difference, and the 4,000-patient cluster RCT shows no mortality difference with a 1.2-day length-of-stay reduction. Inconsistency across populations and settings lowers certainty. I would recommend a risk-stratified pathway: for septic shock or high likelihood of infection with organ dysfunction, antibiotics within 1 hour; for possible sepsis without shock, a structured assessment with antibiotics within 3 hours if concern persists, plus serial reassessment. Certainty: moderate for shock, low for possible sepsis. Strength: strong for shock, conditional for the 3-hour pathway. This respects the evidence, stewardship concerns, and alert fatigue while preserving early treatment for the highest-risk patients.
2. Design an implementation and evaluation plan using an established framework. Include stakeholders, barriers, measures, and a timeline.
Read the worked answer
Use CFIR to map barriers: EHR alert design, nurse staffing, antibiotic stewardship, interprofessional roles, and leadership support. Stakeholders include ED physicians, hospitalists, nurses, pharmacists, infection prevention, informatics, and patients. Interventions: replace 12 interruptive alerts per shift with a single risk-stratified order set; nurse-initiated lactate and blood-culture protocol; pharmacist review at 1 hour; daily huddle feedback; and education for residents. Measures: door-to-antibiotic time, bundle compliance, mortality, length of stay, broad-spectrum days of therapy, alert override rate, and balancing measures such as missed sepsis cases and ED throughput. Use a PDSA ramp and monthly run charts or SPC charts with a 12-month baseline. Timeline: weeks 1–2 stakeholder mapping, weeks 3–4 build and test, month 2 pilot in ED, months 3–6 spread. Report to the medical executive committee with a dashboard, including unintended consequences, and pre-specify stopping rules.