Early Sepsis Control Tied to Lower 90-Day Mortality

Early rapid source control of community-acquired sepsis infections led to a lower risk of death at 90 days, a retrospective cohort study found.

Among over 4,900 hospitalized patients with sepsis, adjusted multivariable logistic regression analysis showed early source control -- within 6 hours -- was associated with a 29% lower odds of 90-day mortality compared with late source control, defined as 6 to 36 hours (adjusted OR 0.71, 95% CI 0.63-0.80), reported Katherine Reitz, MD, MSc, of the Presbyterian Hospital in Pittsburgh, and colleagues, in JAMA Surgery.

The relationship was even stronger for those who underwent gastrointestinal (GI) interventions (aOR 0.56 95% CI 0.43-0.80) or soft tissue interventions (aOR 0.72 95% CI 0.55-0.95), compared with orthopedic and cranial ones (aOR 1.33, 95% CI 0.96-1.83, P<0.001 for interaction).

Interestingly, each hour of source-control delay was linked with a greater risk of mortality at 90 days. The adjusted absolute risk difference of mortality at 90 days was 0.5% lower at 3 hours, but increased by 1% at 12 hours, followed by increases of 3% at 24 hours and 5% at 36 hours.

"In my own interpretation of this study in the context of prior studies, I will advocate for early anti-infectives and source control within 6 hours in all patients with community-acquired infections requiring source control when the source of the infection is gastrointestinal, abdominal, or soft tissue in origin," wrote Pamela Lipsett, MD, MHPE, of Johns Hopkins Hospital in Baltimore, in an accompanying editorial.

"We should prospectively collect quality outcomes about how well we are able to follow these protocols and to what extent we can improve the mortality from these diseases," Lipsett added.

To improve patient outcomes with life-threatening sepsis, rapid source control is recommended but the process required to do so is not well understood, Reitz's group noted. One in five patients do not survive hospitalization with sepsis. Sepsis source control by abscess drainage, debridement of infected or necrotic tissue, microbial contamination control, or removal of infected devices is a key step in early treatment, they pointed out.

"Among the stated goals for focused improvement efforts are early recognition of those with septic shock, early administration of effective antibiotics, and early source control in those who require it," Lipsett said.

Reitz and colleagues examined integrated healthcare system data on 4,962 hospitalized patients from 14 hospitals from January 2013 to December 2017. Adults with community-acquired sepsis, as evidenced by sepsis-3, who underwent source control procedures were included. Excluded were patients admitted for trauma, those who had concurrent solid organ transplants, or those who did not undergo source control procedures, among others.

Follow-up extended through January 2019. The analysis adjusted for sepsis severity, patient factors, physiologic stress of procedures, and resource availability.

Mean patient age was 62 and 52% were men. Most patients were white (85%) with a mean sequential organ failure assessment (SOFA) score of 3.8 (out of 6). Mean BMI was 30.9 and mean Elixhauser comorbidity index was 4 (out of 31). Median time to receive antibiotics was 1.5 hours, while the median time to source control was 15.4 hours.

Sepsis source control was often required for GI and abdominal regions (28%), followed by soft tissue (21%). Notably, 45% were admitted to the ICU and 30% required mechanical ventilation.

Mean length of hospital stay was 11.8 days. About 6% of participants died in the hospital; mortality was 14% at 90 days and 24% at 1 year.

Overall, 27% underwent source control within 6 hours. Early source control was linked to a lower risk of mortality in-hospital and at 1 year.

"Early source control was associated with the greatest risk reduction among middle-aged patients (i.e., 35-54 years), not older adults," which was also true for more men than women, the researchers noted.

Study limitations included residual confounding, since source control-specific pathology was not assessed. Also, quality and completeness of post-intervention source control were unknown.

Comment