Predictive value of the prothrombin time-to-albumin ratio for 28- and 90-day mortality in lung cancer patients admitted to the intensive care unit: A retrospective cohort study

This retrospective observational study assesses the prognostic utility of the prothrombin time/albumin ratio (PAR) for 28- and 90-day mortality in patients with lung cancer. A total of 959 patients with lung cancer were included from the Medical Information Mart for Intensive Care IV v2.2 database (USA). The time origin was defined as intensive care unit (ICU) admission, and patients were censored at 28 and 90 days post-admission for the respective outcome analyses. The main variables were the PAR index, 28-day mortality, and 90-day mortality. Relevant clinical data were extracted, and Cox regression analysis identified PAR as an independent predictor of both 28- and 90-day mortality. Receiver operating characteristic curve analysis compared the predictive performance of PAR (AUC: 0.694 [95% confidence interval: 0.6606-0.7267] for 28-day mortality; AUC: 0.703 [95% confidence interval: 0.6689-0.7366] for 90-day mortality) against prothrombin time, albumin, and established ICU prognostic scores (Logistic Organ Dysfunction System, Oxford Acute Severity of Illness Score, Simplified Acute Physiology Score II, and Sequential Organ Failure Assessment). PAR demonstrated moderate predictive accuracy for both 28- and 90-day mortality. Kaplan-Meier survival analysis further confirmed a significantly higher mortality rate in the high-PAR cohort compared with the low-PAR group. PAR demonstrates moderate predictive value for 28- and 90-day mortality, with performance comparable to several established ICU scoring systems. Its simplicity and ready availability may offer practical advantages for early risk stratification and provide a more accessible prognostic tool for physicians.