Epidemiological Predictors of Severe Leptospirosis in Adults: A Multicenter Retrospective Cohort Study in Northeastern Peninsular Malaysia

Leptospirosis is among the most widespread zoonotic infections globally and is particularly prevalent in tropical settings, where it remains endemic. In Malaysia, the disease burden has increased markedly over recent decades. Severe leptospirosis can result in multi-organ dysfunction and death; however, early identification of patients at risk of a severe clinical course remains a significant challenge. This study aimed to determine independent clinical and laboratory predictors of severe leptospirosis among adults admitted to hospitals in northeastern Malaysia. A multicenter retrospective cohort study was conducted at two tertiary referral hospitals in Kelantan, northeastern Peninsular Malaysia, over a seven-year period. Adults with confirmed leptospirosis were categorised as severe or non-severe according to established clinical criteria. Group comparisons were performed using appropriate statistical tests, followed by multivariable logistic regression with internal bootstrap validation to identify independent predictors of severe disease. Of 525 patients included in the final analysis (mean age 38.1 ± 16.8 years; 65.5% male), 303 (57.7%) met criteria for severe leptospirosis and overall, in-hospital mortality was 6.5%. Common presenting symptoms (>30% of patients) included fever, myalgia, nausea, vomiting, and arthralgia. Independent predictors of severe disease identified on multivariable analysis were; age > 40 years, delayed hospitalisation (presentation after four days of symptom onset), T-wave changes on electrocardiography, conjunctival suffusion, hyponatraemia, prolonged prothrombin time, and elevated alanine aminotransferase (all p < 0.05). The predictive model demonstrated good discriminatory ability (AUC 0.794, 95% CI: 0.747–0.841). Severe leptospirosis is associated with distinct clinical and laboratory features at presentation. Early recognition of key risk factors—particularly older age, delayed presentation, and electrocardiographic abnormalities—may enable clinicians to stratify risk, initiate timely targeted interventions, and potentially reduce morbidity and mortality in this endemic setting.