.jpg "Stephanie Taylor, MD, (she/her/hers) photo")
Stephanie Taylor, MD, (she/her/hers)
Wake Forest School of Medicine
Charlotte, North Carolina
Disclosure information not submitted.
Bethany Bray, PhD
Associate Professor
University of Illinois Chicago, United States
Disclosure information not submitted.
Shih-Hsiung Chou, PhD
Senior Data Scientist
Atrium Health, United States
Disclosure information not submitted.
Title: Clinical Subtypes of Sepsis Survivors Predict Readmission and Mortality After Hospital Discharge
Introduction: Sepsis survivors experience high rates of post-discharge mortality and rehospitalization. Given the heterogeneity of the condition, identifying subtypes within this population with different risks of adverse outcomes may optimize post-sepsis care. We aimed to classify individuals into subtypes using patient, illness, and discharge factors and assess the association of subtypes with 30-day hospital readmission and mortality.
Methods: We conducted a retrospective study among 20,745 patients admitted to 12 hospitals with a clinical definition of sepsis from 2014 to 2017. We selected indicator variables based on availability in the electronic health record prior to discharge and clinical knowledge of known or theoretical association with post-sepsis adverse outcomes: 4 hospital course factors, 3 functional status factors, 3 non-infection related complexity factors, 2 infection-related complexity factors and 2 post discharge healthcare access factors. We used latent class analysis to classify sepsis survivors into subtypes, which were evaluated against 30-day readmission and mortality rates. A secondary analysis evaluated subtypes against ambulatory-care-sensitive-condition readmission rate.
Results: Latent class analysis identified 5 distinct subtypes as the optimal solution. Clinical subtype was associated with 30-day readmission (overall p< 0.01), with the Chronically Ill, Severe Illness subtype demonstrating highest risk (35%) and the Low Functional Needs, Uncomplicated Illness subtype demonstrating the lowest risk (9%). 47% of readmissions in the High Functional Needs, Uncomplicated Illness subtype were for ambulatory-care-sensitive conditions, whereas only 17% of readmissions in the Previously Healthy, Severe Illness subtype were for ambulatory-care-sensitive conditions. Subtype was significantly associated with 30-day mortality (p < 0.01); highest in the Chronically Ill, Severe Illness subtype (8%) and lowest in the Low Functional Needs, Uncomplicated Illness subtype (0.1%).
Conclusions: Sepsis survivors can be classified into subtypes with differential 30-day readmission and variability in ambulatory care sensitive readmissions. Pre-discharge classification may allow an individualized approach to post-sepsis care that improves health outcomes and value by targeting appropriate care to patients.