Hospitals should prioritize five infection-focused strategies to improve sepsis care, including rapid blood culture diagnostics paired with antimicrobial stewardship, timely antibiotic delivery in septic shock and optimized beta-lactam administration, according to a position paper.
The paper, published in Clinical Infectious Diseases, was developed by a 24-member multidisciplinary panel convened by the Infectious Diseases Society of American (IDSA), with representatives from eight other professional societies and the Sepsis Alliance. Panel members represented a range of specialties, including adult and pediatric infectious diseases, critical care, emergency medicine, hospital medicine, antimicrobial stewardship, infection prevention, microbiology, pharmacy, quality improvement, health systems leadership and patient advocacy.
“Sepsis remains a leading cause of serious illness and death, and sepsis guidelines and quality improvement efforts have understandably focused largely on early recognition and treatment. Those remain essential, but they are only part of the solution,” Chanu Rhee, MD, MPH, FIDSA, director of the Center for Sepsis Epidemiology and Prevention Studies at the Harvard Pilgrim Health Care Institute and chair of the panel, told this news organization. “We saw a need for practical guidance on the hospital-level systems, workflows, and practices that can strengthen the infection-related aspects of sepsis care throughout a patient's course.”
The recommendations are intended to complement, not replace, existing sepsis guidelines and quality initiatives. They focus on six domains: diagnostic testing and pathogen detection; antimicrobial management and delivery; surveillance and performance metrics; adjunctive therapy; program infrastructure and organizational support; and infection prevention.
Using a modified Delphi process, panelists proposed recommendations, reviewed supporting evidence and considered operational feasibility and possible unintended consequences. Each recommendation was independently scored on a 1-to-5 scale for potential impact on outcomes, opportunity size, feasibility, risk of unintended consequences and overall priority. A formal systematic review was not conducted.
Recommendations were classified as Tier 1, or high priority, and Tier 2, or moderate priority, based on scoring patterns and panel consensus rather than a numerical cutoff. Tier 1 strategies had the strongest overall combination of potential to improve outcomes, feasibility for hospital implementation and supportive evidence and rationale. They were recommended for broad implementation. Tier 2 strategies had greater uncertainty surrounding factors such as impact, feasibility, cost or implementation readiness.
Tier 1 recommendations
Five recommendations received Tier 1 designation.
1. Implement multiplex nucleic acid amplification testing for positive blood cultures, paired with active antimicrobial stewardship support to ensure timely interpretation, treatment optimization and follow-up.
The panel cited evidence that blood cultures identify bloodstream pathogens in 10% to 20% of patients with suspected sepsis and up to 40% of those with septic shock. Multiplex testing can identify about 90% of common bloodstream pathogens and key resistance genes within hours after cultures become positive. The panel noted that evidence is strongest for improving time to optimal therapy and antibiotic optimization, while evidence for reducing mortality is less certain.
2. Implement hospital workflows and performance monitoring to target first antibiotic administration within 1 hour of septic shock recognition.
The recommendation distinguishes septic shock from sepsis without shock when considering antibiotic timing. Multiple large observational studies cited by the panel found that each hour of antibiotic delay in patients with septic shock was associated with 7% to 14% greater odds of death, while evidence supporting strict time targets for sepsis without shock was inconsistent. The panel recommended a target of less than 1 hour after recognition of septic shock.
3. Monitor and minimize time from antibiotic order to infusion start in patients with septic shock, with a target of less than 30 minutes.
The panel emphasized that the less-than-30-minute target is a pragmatic benchmark rather than an evidence-based threshold. Studies linking administration delays with increased mortality generally involved delays of approximately 1 to 1.5 hours and were not specifically focused on septic shock. The panel cautioned against extending the aggressive infusion-time target to all antibiotic orders because it could strain resources and encourage unnecessary broad-spectrum antibiotic use.
4. Establish institutional policies that default antipseudomonal beta-lactam antibiotics to prolonged infusion after an initial loading dose when ordered for critically ill patients with sepsis.
In the more than 7,200-patient BLING III trial cited in the paper, 90-day mortality was 24.9% among ICU patients receiving continuous infusion vs 26.8% among those receiving intermittent infusion, although the difference was not statistically significant in the primary analysis. A concurrent meta-analysis incorporating BLING III and 16 other randomized controlled trials found a 99.1% probability that prolonged infusion was associated with lower 90-day mortality, with an estimated number needed to treat of approximately 26 to prevent one death. The panel noted that evidence supporting prolonged infusion is strongest among critically ill patients and that direct comparisons of extended vs continuous infusion strategies are lacking. They added that clinical benefits of prolonged beta-lactam infusion have not been clearly demonstrated in children and suggested considering it for patients at higher risk of suboptimal pharmacokinetic or pharmacodynamic target attainment.
5. Implement clinical decision support and education to prioritize beta-lactam administration before vancomycin when both are ordered for suspected sepsis.
The panel cited the relatively low prevalence of methicillin-resistant Staphylococcus aureus (MRSA) and the potential for vancomycin administration to delay beta-lactam therapy. MRSA was identified in fewer than 4% of patients treated for suspected community-onset sepsis and about 12% of those with culture-confirmed community-onset sepsis. Because vancomycin requires a longer infusion time, its administration can delay beta-lactam delivery. Two retrospective studies associated beta-lactam-first administration with lower mortality, although the panel noted the evidence was observational and subject to residual confounding. The recommendation is intended to guide default workflow rather than override clinical judgment when vancomycin-first treatment or targeted anti-MRSA therapy is appropriate.
Tier 2 recommendations
Tier 2 recommendations include clinical decision support to improve blood culture ordering; timely administration of second antibiotic doses with frequent dosing intervals in patients with suspected sepsis; protocols for patients reporting beta-lactam allergies; and electronic health record-based tools that estimate infection site-specific risk for multidrug-resistant organisms. The panel also recommended adopting the CDC adult sepsis event definition to monitor sepsis incidence, outcomes and care processes; monitoring inadequate and unnecessarily broad empiric antibiotic therapy; measuring de-escalation of anti-MRSA and antipseudomonal therapy; and monitoring time to definitive source control in patients requiring procedural intervention and tracking transfer timeliness when source control is unavailable on-site.
Additional recommendations and emerging strategies
The panel also addressed adjunctive therapies and emerging strategies. Among the Tier 2 recommendations were protocols to promote corticosteroid use in appropriately selected patients with severe community-acquired pneumonia and in patients with refractory septic shock.
The panel also recommended multidisciplinary sepsis governance that incorporates infectious diseases, antimicrobial stewardship and clinical microbiology expertise; antimicrobial stewardship participation in ICU rounds; and support for pharmacokinetic and pharmacodynamic approaches to antimicrobial dosing for critically ill patients.
The primary infection prevention recommendation was to ensure high adherence to the SHEA/IDSA Compendium for health care-associated infection prevention. Additional recommendations included universal chlorhexidine gluconate bathing with universal or targeted MRSA nasal decolonization for all ICU patients and non-ICU patients with invasive devices, along with structured programs for daily toothbrushing for all hospitalized patients.
Several approaches were classified as emerging and promising rather than recommended for routine implementation because evidence continues to evolve or real-world implementation experience remains limited. These included automated sepsis prediction tools often using machine learning or artificial intelligence; host immune response-based diagnostics; culture-independent rapid diagnostics for bloodstream infection; multiplex nucleic acid amplification testing for lower respiratory tract samples; and procalcitonin-guided antibiotic duration.
The recommendations are intended for both adult and pediatric settings, although the panel noted that much of the supporting evidence derives from adult populations and care settings. Pediatric-specific considerations are included where relevant, and some recommendations may require age-specific adaptation. Neonatal sepsis was outside the position paper's scope.
The panel acknowledged that small, rural and safety-net hospitals may face barriers to implementing recommendations that require advanced diagnostics, informatics support or real-time performance monitoring. They proposed incremental adoption based on local capacity and pointed to regional collaboration, shared laboratory services and tele-infectious diseases, tele-stewardship and tele-critical care as potential approaches for resource-constrained hospitals.
“Improving sepsis outcomes requires coordinated efforts across many different disciplines,” said Dr. Rhee. “Our hope is that the paper provides hospitals with a practical framework for bringing these efforts together into stronger sepsis programs that consistently deliver high-quality care and improve patient outcomes.”
Several panelists reported relationships with government agencies, research funders, professional societies, publishers, and health-care organizations. These included research funding, consulting or advisory roles, honoraria and royalties, editorial positions, and committee or task-force leadership roles. Full disclosures can are available in the published paper.