June 30, 2020

Most Sepsis Hospital Deaths Unpreventable Due to Patient Complexity

Despite the prevalent view that most sepsis-based deaths are preventable with better care, the authors of a study published in JAMA Network Open report that it is unlikely that a large proportion of these deaths can be prevented.

While the authors agree that optimal treatment of sepsis is definitely important to avoid preventable deaths, they suggest that, rather than focusing primarily on improving hospital-based sepsis treatment, the effects of sepsis may be better minimized by improving awareness about vulnerable patient populations and managing serious comorbidities that often occur alongside death by sepsis.

40% Were Hospice-Eligible But Weren't Referred

One of the study’s key findings: 40.3% of patients who died from sepsis exhibited hospice-qualifying conditions at the time of admission but were not referred to hospice for care.

“Further innovations in the prevention and care of underlying conditions may be necessary before a major reduction in sepsis-associated deaths can be achieved.” —Chanu Rhee, MD, MPH Harvard Medical School

Many deaths in the US involve sepsis (possibly more than 250,000 each year), according to the authors, and yet sepsis is often underreported among hospital discharge codes and death certificates. As a result, the prevalence, causes and best approaches to care can be difficult to study.

“Sepsis disproportionately affects patients who are elderly, have severe comorbidities, and have impaired functional status,” the authors note. “As a result, some of these patients may receive optimal, guideline-compliant care yet still die due to overwhelming sepsis or from their underlying disease.”

Researchers performed a retrospective analysis of adult patients admitted to six US hospitals between 2014 and 2015 who died or were enrolled in hospice and not readmitted. Clinicians reviewed the medical records of 568 patients who died of sepsis during the study period.

These reviewers rated the preventability of death from 1 to 6 on a Likert scale (1, definitely preventable; 2, moderately likely to be preventable; 3, potentially preventable under ideal circumstances; 4, unlikely to be preventable even though some circumstances may not have been optimal; 5, moderately likely not to be preventable; and 6, definitely not preventable due to fatal illness and/or goals of care precluding aggressive treatment). Clinicians also identified any suboptimal aspects of care.


  • Of 568 patients in the analysis, mean age was 70.5
  • 395 (69.5%) died in the hospital, and 173 (30.5%) were discharged to hospice
  • Among sepsis-associated deaths, 121 of 300 patients (40.3%) were eligible for hospice at the time of admission to the hospital

Prevalence of Sepsis and Serious Comorbidities

Sepsis was present in 300 hospitalizations (52.8%; 95% confidence interval [CI], 48.6%–57.0%) that resulted in death. Researchers found 121 of the 300 patients who died with sepsis (40.3%; 95% CI, 34.7%–46.1%) to have “end-stage comorbidities” as defined by hospice criteria. The most common comorbidities were solid cancer (20.0%), refractory hematologic cancer (5.3%), severe dementia (5.0%), severe stroke (4.0%) and severe chronic lung disease (5.3%).

Sepsis was considered the immediate cause of death in 198 cases (34.9%; 95% CI, 30.9%– 38.9%). Of an additional 102 patients (18.0%; 95% CI, 14.9%– 21.4%) for whom sepsis was present but resolved before death, reviewers still considered sepsis to contribute to the death of 44 (43.1%; 95% CI, 33.4%–53.3%).

The most common underlying causes of death in sepsis patients were solid cancer (21.0%), chronic heart disease (15.3%), hematologic cancer (10.3%), dementia (9.7%) and chronic lung disease (9.0%), the authors write.

Preventability of Sepsis Deaths

Among patient deaths involving sepsis:

  • 264 (88.0%; 95% CI, 83.8%–91.5%) were judged to be definitely or most likely unpreventable (4 to 6 on a Likert scale)
  • Only 36 (12.0%; 95% CI, 8.6%–16.2%) were judged possibly preventable (1 to 3 on a Likert scale); of these, only 11 (3.7%) were definitely or moderately likely preventable
  • 232 (77.3%) had “no identifiable suboptimal aspects of care”
  • 68 (22.7%) had aspects of suboptimal care. Of these, 32 (47.1%) were judged definitely, moderately or possibly preventable

“Patients who died with sepsis tended to be older adults with multiple comorbidities and recent hospitalizations, and underlying causes of death were mostly associated with severe chronic comorbidities,” the authors say.

Consistent with previous studies, these results suggest a large proportion (over 50%) of adult hospital deaths and discharges to hospice likely involve sepsis, despite the fact that a much smaller proportion (approximately 6%) of death certificates indicate the presence of sepsis.

Hospice Guidelines for Patients with Sepsis and Concomitant End-Stage Disease >

“[A]lthough the burden of sepsis-associated mortality is high, our study indicates that most of these deaths may not be preventable through better hospital-based care,” the authors write, noting that clinician reviewers found most cases of sepsis-associated death to be unpreventable due to “incurable underlying diseases” or severe illness that persisted despite appropriate treatment.

Although the study did identify some cases of suboptimal care among patients dying of sepsis, the authors stress that these cases were infrequent, occurring in less than one-quarter of cases, and that “death was still thought to be unpreventable in more than half of those patients.”

The findings suggest that only about 1 in 8 sepsis-associated deaths may be preventable, the authors state, with only 1 in 25 being judged definitely or moderately preventable. “Our findings do not diminish the importance of trying to prevent as many sepsis-associated deaths as possible,” they write, “but rather underscore that most fatalities occur in medically complex patients with severe comorbid conditions.”

Source: “Prevalence, Underlying Causes, and Preventability of Sepsis-Associated Mortality in US Acute Care Hospitals,” JAMA Network Open; February 15, 2019. Rhee C et al; Department of Medicine, Brigham and Women’s Hospital; and Department of Population Medicine, Harvard Medical School, Harvard Pilgrim Health Care Institute, both in Boston, Massachusetts.