Abstract
BACKGROUND: Pulmonary hypertension (PH) is a life-threatening disease. However, acidosis could be used to predict the prognosis of critically ill patients. Consequently, this study was to identify the link between acidosis and in-hospital death of PH patients based on the Medical Information Mart for Intensive Care-IV (MIMIC-IV) database. METHODS: Eligible subjects from the MIMIC-IV database were selected for this analysis (2008-2019), after which differences in variables between the survival statuses of PH patients were evaluated. Subsequently, employing three weighted multiple logistic regression models to investigate the link between acidosis and PH. Further, risk stratification analysis were applied to explore the relationships between acidosis as well as other covariates and PH. RESULTS: Total 2,530 PH patients (247 dead and 2,283 live or 157 acidosis and 2,373 non-acidosis) were included in the analysis. Next, the result indicated highly significant differences between the dead and live groups in factors such as acidosis and sepsis (p < 0.0001). It also showed highly significant differences between the acidosis and non-acidosis groups in factors such as creatinine and sepsis (p < 0.0001). Subsequently, a consistent significant association was found between acidosis and PH, there into, Model 1 displayed an odds ratio (OR) of 5.53 (95% confidence interval (CI): 3.83-7.92, p = 2.71 × 10(-20)), Model 2 showed an OR of 5.56 (95% CI: 3.83-8.00, p = 6.33 × 10(-20)), Model 3 reported an OR of 2.19 (95% CI: 1.36-3.51, p = 1 × 10(-3)), indicating that the impact of acidosis on PH was not significantly affected by other covariates. Notably, risk stratification further revealed acidosis as a risk factor for PH was stable across populations (OR > 1, p < 0.05). CONCLUSION: This study identified acidosis was a risk factor for PH, highlighting the importance of monitoring in PH patients at risk for acidosis.