Assessing clinical heterogeneity in sepsis through treatment patterns and machine learning

Abstract

Objective To use unsupervised topic modeling to evaluate heterogeneity in sepsis treatment patterns contained within granular data of electronic health records. Materials and Methods A multicenter, retrospective cohort study of 29 253 hospitalized adult sepsis patients between 2010 and 2013 in Northern California. We applied an unsupervised machine learning method, Latent Dirichlet Allocation, to the orders, medications, and procedures recorded in the electronic health record within the first 24 hours of each patient’s hospitalization to uncover empiric treatment topics across the cohort and to develop computable clinical signatures for each patient based on proportions of these topics. We evaluated how these topics correlated with common sepsis treatment and outcome metrics including inpatient mortality, time to first antibiotic, and fluids given within 24 hours. Results Mean age was 70 ± 17 years with hospital mortality of 9.6%. We empirically identified 42 clinically recognizable treatment topics (eg, pneumonia, cellulitis, wound care, shock). Only 43.1% of hospitalizations had a single dominant topic, and a small minority (7.3%) had a single topic comprising at least 80% of their overall clinical signature. Across the entire sepsis cohort, clinical signatures were highly variable. Discussion Heterogeneity in sepsis is a major barrier to improving targeted treatments, yet existing approaches to characterizing clinical heterogeneity are narrowly defined. A machine learning approach captured substantial patient- and population-level heterogeneity in treatment during early sepsis hospitalization. Conclusion Using topic modeling based on treatment patterns may enable more precise clinical characterization in sepsis and better understanding of variability in sepsis presentation and outcomes.

The Presentation, Pace, and Profile of Infection and Sepsis Patients Hospitalized Through the Emergency Department: An Exploratory Analysis

“…First, our data highlight the often described, but poorly characterized, challenge of identifying and treating sepsis patients because of heterogeneity in clinical presentation ( 40 , 46 ). Compared with heart failure and stroke, which showed similarly predominant symptoms between patients, infectious patients presented with symptoms were diverse and nonspecific.…”

“…We identified key inpatient adverse outcomes (hospital mortality, the need for ICU admission during hospitalization) to assess their association with specific signs and symptoms ( 38 , 39 ). We also quantified the time to first antibiotic (grouped as ≤ or >3 hr after ED presentation) as a common sepsis process measure ( 3 , 40 ). We used multivariable logistic regression to estimate the association between the presence of each symptom and each outcome, adjusting for the presence or absence of all other signs and symptoms at a hospitalization level.…”

Machine learning in infection management using routine electronic health records: tools, techniques, and reporting of future technologies

“…1). Three studies used an unsupervised clustering approach for clustering sepsis patients [31,42,44]. Data from sepsis patients were also used to learn optimal treatment by reinforcement learning [19].…”

“…In contrast to supervised ML models, unsupervised approaches for cluster analysis were used by only three studies identified in our review [31,42]. All three focused on sepsis patients.…”

“…Given the high complexity of these patients, unsupervised methods can provide promising potential to better understand this population. The studies aimed at identifying treatment patterns, clinical phenotypes, and mortality-associated patient groups (methods used: latent Dirichlet allocation [42], partitioning around medoids methods [44], and composite mixture models [31] respectively) ( Fig. 1).…”