Prognostic value of serum anion gap in pulmonary embolism independent of lactate: evidence from a natural experiment of missing lactate data

Pulmonary embolism (PE) is a life-threatening condition whose prognosis remains difficult to predict. While metabolic disturbances are common in PE, the prognostic value of the serum anion gap (AG) independent of lactate—a key confounder—remains unestablished. This study leveraged the setting of incomplete lactate testing as a natural experiment to determine whether AG provides independent prognostic information in critically ill PE patients. A cohort of 2147 ICU-admitted PE patients was extracted from the MIMIC-IV database. A significant linear relationship between AG and all-cause mortality was observed. Using an AG threshold of 14 mmol/L, patients were dichotomized into high- and low-risk groups. After comprehensive adjustment for confounders, multivariate Cox regression confirmed that a high AG was independently associated with significantly increased risks of ICU (HR = 2.02, 95% CI 1.25–3.28), in-hospital (HR = 1.87, 95% CI 1.28–2.73), 30-day (HR = 1.77, 95% CI 1.28–2.45), and 90-day (HR = 1.45, 95% CI 1.10–1.91) mortality (all p < 0.05). Crucially, this association remained robust in the substantial subgroup of patients without lactate data, providing conclusive evidence that AG’s prognostic value is not merely a surrogate for lactate. This study establishes serum AG as a readily available and independent prognostic marker in PE, highlighting its particular utility in clinical scenarios where lactate is not measured.