Abstract Background and Aims Leptospirosis is a zoonotic disease with manifestations ranging from mild to severe. Acute kidney injury (AKI) in leptospirosis is usually a marker of poor prognosis and there remains a paucity of scoring systems utilized to predict renal replacement therapy (RRT) and mortality. The Model for End-Stage Liver Disease-Sodium (MELD-Na) was originally developed to predict survival in patients with liver cirrhosis. The components of the MELD-Na score evaluate renal and hepatic dysfunction, which are frequent complications in leptospirosis. This study aims to evaluate the prognostic significance of the admission MELD-Na score in predicting renal replacement therapy and mortality among patients with leptospirosis-associated AKI. Method A retrospective, single-center, cross-sectional study was conducted at a tertiary hospital in Manila, Philippines. The study enrolled patients from January 2018 to September 2023 and collected demographic and clinical data, along with laboratory values taken on admission. The MELD-Na score was developed using these values and its distribution was visualized by histogram. Receiver operating characteristic curve analysis was used to determine the score's predictive ability and the optimum cut-off was determined by the cut-point with the highest Youden index. Results The study included a total of 298 patients. The median age of the participants was 39, with 88.59% being males, and 77.52% having no known comorbidities. Of the study population, 74.16% underwent RRT, while the in-hospital mortality rate was 12.08%. Univariate analysis revealed that oliguria, serum creatinine, arterial pH, serum bicarbonate, and administration of methylprednisolone pulse therapy were among the predictive factors for RRT, in line with clinical practice guidelines. Other parameters found to be predictive of RRT were serum sodium and bilirubin. An admitting MELD-Na score of >26 predicted RRT with 91.0% sensitivity [86.4, 94.4] and 87.7% specificity [76.3, 94.9], yielding an area under the curve (AUC) of 0.953 (95% CI). Conversely, an admitting MELD-Na score >31 was found to be 77.8% sensitive [60.8, 89.9] and 64.9% specific [58.8, 70.7] for predicting mortality, with an AUC of 0.748 (95% CI). Conclusion The MELD-Na is a good predictor of renal replacement therapy and in-hospital mortality among patients with leptospirosis-associated AKI. This scoring system holds the potential to serve as a tool for risk stratification and prognostication in moderate to severe cases of the disease. A validation cohort study is recommended for formulating a prediction scoring system based on the individual components of the MELD-Na score, such as bilirubin and sodium.
Read full abstract