OBJECTIVES: Sepsis, a systemic inflammatory response to infection, is a leading cause of inpatient mortality and consumes substantial resources. This study aims to examine real-world healthcare resource utilization (HCRU) among septic patients in intensive care units (ICU).

METHODS: Electronic medical records from the Medical Information Mart for Intensive Care III database between 2001 and 2012 were used. Key HCRU components examined were length of stay (LOS), number of microbiology tests, mechanical ventilation (MV) use, and vasopressor use. Subgroup analyses based on ICU care setting and different diagnosis criteria (Sepsis-1 vs Sepsis-3) were also conducted.

RESULTS: A total of 17,848 sepsis-related admissions and 17,802 septic patients were identified. Over half of the patients were >=65 years (51.7%), male (53.2%), white (72.6%), and had Medicare (63.1%) coverage. The median interquartile range (IQR) scores for systemic inflammatory response syndrome (SIRS) and Sequential Organ Failure Assessment (SOFA) were 3 (2-4) and 5 (3-8), respectively. About 92.8% of admissions were classified as Sepsis-1, while 89.9% were Sepsis-3. Greatest differences were found in the number of septic admissions in surgical ICU (SICU: 92.8% and 87.1%) and trauma surgical ICU (TSICU: 93.4% and 84.5%) using Sepsis-1 and Sepsis-3 criteria, respectively. Overall mean LOS in ICU was 6.9 (SD: 8.9) days, with highest LOS in cardiac surgery recovery unit (CSRU) (11.8 days) and TSICU (8.6 days). The mean number of microbiology tests was 30.2 (SD: 36.4). Mechanical ventilation (56.6%) and vasopressor use (43.5%), were mostly observed in the CSRU (80.7% for mechanical ventilation use, 70.7% for vasopressor use). HCRU was similar among both sepsis diagnosis criteria, but those that met Sepsis-1 only had lower HCRU.

CONCLUSIONS: The results demonstrated that sepsis have a substantial impact on HCRU, and this impact varies across ICU setting and sepsis diagnosis criteria. Future research should focus on unmet need among different subgroups.