RATIONALE Sepsis is the leading cause of mortality among medical patients in the Philippine General Hospital (PGH). A previous study illustrated variations in sepsis management. The Department of Medicine developed a sepsis pathway based on the Surviving Sepsis Campaign bundles to standardize care and improve outcomes. We determined the coverage and compliance with the pathway, the barriers to compliance and sepsis-related mortality. METHODS This was a single-center mixed methods study on the pilot implementation of the sepsis pathway (April 8 to July 7, 2019) in the medical service areas, i.e. emergency department (ED), medical wards and medical intensive care unit (MICU), of a tertiary level teaching hospital. We tracked all medicine charity admissions with infections to determine coverage. Compliance and patient outcomes were assessed through chart reviews. Focus group discussions and interviews were done to identify barriers to implementing the sepsis bundle. RESULTS Among 296 admissions with infections (49% female, mean age 51.4 years), there were 422 patient-days eligible for pathway coverage but only 199 patient-days (47.16%) were covered. The ED had the highest coverage rate. Overall mortality rate among the admissions was at 39.2%. Among septic patients who were covered, 40% died. Missed cases were associated with increased odds of in-hospital death (adjusted odds ratio [aOR]: 1.42, 95% CI: 1.13 to 1.88) on multivariate analysis. Compliance with recommended diagnostics was high except for lactate and bilirubin. Blood cultures were sent 98% of the time. Only 35% of patients received antibiotics by one hour after activation. Fluids recorded over 6 hours were inadequate (mean 4.77 mL/kg, standard deviation: 2.82 mL/kg). Of 73 patients with hypotension needing fluid resuscitation, only 12 had blood pressure documented 30 minutes post-activation. Stakeholders identified inadequate human and physical resources, hospital policy changes and pathway form construction as barriers to compliance. Fellows, nurses, and students reported lack of orientation on their roles.

RATIONALE Sepsis remains a leading cause of mortality and hospitalization costs. Individual randomized controlled trials (RCTs) on macrolides with their posited immunomodulatory effects showed no clear effect on mortality among septic patients. This study aimed to determine the effectiveness of adding macrolides to standard care in reducing mortality, shortening stay in the intensive care unit (ICU) and hospital, and duration of mechanical ventilation among adult patients with sepsis. METHODS We searched electronic databases (PubMed, CINAHL Plus with Full Text, CENTRAL, ClinicalTrials.gov, ISRCTN Registry and Herdin) until March 2018 for RCTs comparing macrolides with standard care or placebo in adults with sepsis or septic shock. We also examined Philippine journals and reference lists and attempted to contact authors for unpublished data. Two authors independently extracted data and appraised study characteristics, quality, and outcomes. Statistical analysis was done using RevMan 5.3. RESULTS We found three completed randomized trials with 847 participants. Most patients had Gram negative infections and received carbapenem-based antibiotics. Pooled analysis of the moderate to high quality studies showed no significant effect on all-cause mortality at one month (RR 1.00, 95% CI: 0.77, 1.30) and no heterogeneity (I 2 =0). Planned subgroup analyses by macrolide type were likewise inconclusive: azithromycin versus placebo (RR 1.70, 95% CI: 0.31, 9.28) and clarithromycin versus placebo (RR 0.98, 95% CI: 0.75, 1.29). Azithromycin did not reduce long-term all-cause mortality at 180 days nor 360 days. No significant difference in ICU and hospital length of stay, and duration of mechanical ventilation was observed. CONCLUSIONS Few high-quality RCTs on macrolide therapy for sepsis are currently available. Add-on macrolides to standard management of septic adult patients did not significantly reduce mortality, shorten length of stay, nor hasten liberation from mechanical ventilation.