BACKGROUND: Subphenotypes of infected patients have been reported in Europe and North America, but few studies have investigated populations in Southeast Asia. We sought to identify and differentiate subphenotypes of patients hospitalized with suspected infection in rural Thailand using biological markers implicated in the dysregulated host response. METHODS: In a cohort of prospectively enrolled patients hospitalized with suspected infection in northeastern Thailand, we measured 15 circulating biomarkers from a random selection of 585 subjects and developed latent profile models to identify subphenotypes. Patient characteristics were compared after subphenotype assignment, and a parsimonious model was developed to identify patient subphenotypes. FINDINGS: We identified and assigned 585 patients to three subphenotypes termed latent biological profile (LBP)-1 (52%), LBP-2 (39%) and LBP-3 (9%). Patients assigned to LBP-3 had a higher risk of 28-day mortality compared to those in LBP-1 and LBP-2 (adjusted relative risk 1.8, 95% confidence interval [CI] 1.1-2.9, P = 0.02). Patient clinical characteristics and biomarker concentrations also differed by subphenotype assignment. A parsimonious three-biomarker model identified subphenotypes in an internal validation cohort (LBP-1: area under the receiver operating curve [AUC] 0.96, 95% CI: 0.94-0.98
LBP-2: AUC 0.77, 95% CI 0.71-0.83
LBP-3: AUC 0.99, 95% CI 0.98-1.00). INTERPRETATION: We identified three biological subphenotypes of patients with suspected infection in rural Thailand, where the burden of infection is high but understudied. Patient subphenotype assignment was characterized by distinct clinical outcomes and biological profiles which could inform contextualized future study design. FUNDING: The US National Institutes of Health, the Wellcome Trust, and the Firland Foundation.