Preventable Mortality Among Previously Healthy Patients With Sepsis
Recommended Citation
Hechtman RK, Munroe E, McLaughlin E, Horowitz J, Heath M, Creutz E, Posa P, Jayaprakash N, Blamoun J, Angus DC, Flanders S, Prescott HC. Preventable Mortality Among Previously Healthy Patients With Sepsis. Am J Respir Crit Care Med 2024; 209:A1101.
Document Type
Conference Proceeding
Publication Date
6-1-2024
Publication Title
Am J Respir Crit Care Med
Abstract
Rationale: Sepsis is the leading cause of in-hospital mortality. However, many patients with sepsis are elderly or have significant comorbidities, so it is unclear how many deaths may be preventable with optimal care. By contrast, deaths among patients who were healthy prior to hospitalization may reflect preventable deaths. We completed a multi-hospital study of previously healthy patients who died following sepsis to assess for evidence of preventability. Methods: Multicenter, retrospective cohort study of adults hospitalized with community-onset sepsis at 10 hospitals in the Michigan Hospital Medicine Safety Consortium (HMS-Sepsis), a Collaborative Quality Initiative within Blue Cross Blue Shield of Michigan. Data on sepsis hospitalizations were abstracted into the HMSSepsis registry by professional abstractors. We identified a cohort of previously healthy patients by serially excluding frail, multimorbid patients while retaining ≥10% of hospitalizations, then identified patients with 90-day mortality. We asked attending physicians at hospital systems with ≥5 previously healthy deaths to conduct chart reviews. They completed a 10-question survey evaluating patients' baseline health, quality of sepsis management, and potential preventability of death on a 6-point Likert scale from “definitely not preventable” and “slight evidence for preventability” to “definitely preventable”. Results: Of 10,380 patients in the HMS-Sepsis registry, 1,075 were classified as previously healthy. Of these, 109 (10.1%) died within 90 days, including 85 (7.9%) at hospitals with ≥5 previously healthy deaths. Physicians completed reviews for 36/85 (42.3%). Among these, mean age was 55.8, 58.7% were female, and 76.1% died during sepsis hospitalization. Patients were predominantly ASA Class I (N=11, 30.5%) or Class II (N=20, 55.5%) and 77.8% had an estimated life expectancy of >5 years if they had been cured of sepsis. 6 deaths (16.7%) were felt to be “possibly preventable”, 10 deaths (27.8%) had “slight evidence for preventability”, and 20 deaths (55.5%) were classified as “definitely not preventable”. Of deaths graded not preventable or having only slight evidence for preventability, the most common reason was severity of illness at presentation (N=22, 73.3%). Of the 6 deaths graded possibly preventable, the most common reasons were development of iatrogenic/nosocomial complications and surrogate decision maker elected to discontinue life sustaining measures. Conclusion: In this statewide cohort of sepsis hospitalizations, we identified previously healthy patients who died following sepsis. On physician review, nearly half of deaths were judged to have evidence of potential preventability. Most “not preventable” deaths were attributed to severity of illness at presentation, which suggests opportunity for community-level intervention.
Volume
209
First Page
A1101
