Hyperphosphatemia as a Predictor of Clinical Outcomes in Lactic Acidosis Due to Sepsis

Background: Sepsis occurs in over 1.7 million patients in the U.S. annually. Lactic acidosis often occurs in critically-ill patients with sepsis. Hyperphosphatemia is associated with worse prognosis in conditions that cause tissue ischemia and thus may reflect the extent of ischemic tissue injury. Recent studies found associations between hyperphosphatemia and increased mortality in critically-ill patients with spesis or septic shock. The aim of this study was to determine whether hyperphosphatemia determines clinical outcomes in this patient population better than lactate levels. Methods: This was a retrospective cohort study of adult ICU patients with sepsis at a large integrated healthcare system. A multivariate logistic regression model with its respective ROC curve was constructed to explore the relationship between mortality (in-hospital, 30-day, 90-day), phosphate and lactate. Results: 526 patients were included with an overall 38.4% mortality. For lactate and phosphorus, the OR for predicting in-hospital mortality was 1.17 and 1.26 respectively. The ROC curve demonstrates that the Area Under the Curve (AUC) for lactate was not significantly higher than that for phosphorus. The AUC for combined lactate and phosphorus was not significantly higher than the AUC for lactate alone. This study is the first to examine whether hyperphosphatemia determines clinical outcomes better than lactate levels. As expected, lactate predicted in-hospital mortality with an OR of 1.17, though it did not predict 30-day or 90-day mortality. Hyperphosphatemia also predicted in-hospital mortality with a higher OR of 1.26 compared to lactate. Additionally, unadjusted phosphorus predicted 30-day and 90-day mortality, though it was not statistically significant. The ROC curve's AUC demonstrated that lactate and phosphorus performed similarly in predicting in-hospital mortality, though combining phosphorus and lacate did not perform better than lactate alone. Conclusion: In ICU patients with sepsis and lactic acidosis, phosphorus predicts in-hospital mortality as well as lactate does. Phosphorus showed a trend toward predicting 30-day and 90-day mortality better than lactate, though it was not statistically significant. Further work needs to be done to determine if the addition of phosphorus to existing prediction models can improve their accuracy. Funding: Commercial Support - Regional Research Committee of Kaiser Permanente. Grant No. KP-RRC-20210504