Variable Extraction

We used the Structured Query Language with Navicat Premium (version 15.0.21) to extract the data. The primary exposure of interest was whether a patient underwent CVP measurement. The patients who had CVP measurement within 24 h of admission to the ICU were categorized as the CVP group, with the remaining patients forming the no CVP group. The time to initial CVP measurement and the initial level of CVP were also recorded.

Baseline characteristics within the first 24 h after ICU admission were collected, including age, sex, body mass index (BMI), ethnicity, admission type, admission period, first care unit, risk factors of ARDS, the severity of illness and organ dysfunction as measured by acute physiology score III (APS III), oxford acute severity of illness score (OASIS) and logistic organ dysfunction system (LODS). Vital signs included temperature, heart rate, respiratory rate and mean arterial pressure (MAP), were also extracted. The PaO₂/FiO₂ and positive terminal expiratory pressure (PEEP) were collected at the time of diagnosis. The laboratory variables, including white blood cell, hemoglobin, platelet, bicarbonate, blood urea nitrogen, creatinine, lactate, glucose, sodium and potassium were collected within 24 h of ICU admission. If a variable was recorded more than once in the first 24 h, we used the value at the first measurement. In our study, all missing values were imputed using a miss forest imputation, and the details about missing values can be found in Supplementary Table 1.

All comorbidities were identified by the International Classification of Diseases (ICD)-9 or ICD-10 codes in the MIMIC-IV database, including hypertension, coronary atherosclerosis, diabetes, chronic obstructive pulmonary disease (COPD) and tumor. The Charlson comorbidity score was used for comorbidity assessment.