Construction of a risk prediction model for postoperative pneumonia based on the prognostic nutritional index and analysis of related factors in patients with intracerebral hemorrhage.

Introduction: Postoperative pneumonia (POP) is a common complication following hematoma extraction in patients with cerebral hemorrhage, contributing to poor prognosis. Prognostic nutritional index (PNI), a composite index combining serum albumin (a marker of nutritional status) and lymphocyte count (a marker of immune function), reflects both nutritional reserve and immune competence. Impaired nutritional status and immune dysfunction are key drivers of postoperative infections, making PNI a theoretically plausible indicator for predicting POP. This study aimed to explore the relationship between POP and nutritional indices (with a focus on PNI) after hematoma clearance and to develop a predictive model for POP.

Methods: A retrospective analysis was conducted on 325 patients who underwent hematoma removal, including 133 patients diagnosed with POP. The PNI was calculated using the formula: PNI = 5 × lymphocyte count (×10⁹/L) + serum albumin (g/L). Univariate and multivariate logistic regression analyses were performed to identify independent risk factors for POP. The performance of the predictive model was evaluated using the area under the receiver operating characteristic curve (AUC), internal validation, and visualization via a Nomogram.

Results: Significant POP risk factors: low PNI (p < 0.001, OR = 0.84, 95%CI 0.77-0.90), hypoproteinemia (p = 0.008, OR = 2.91), low admission GCS (p = 0.009, OR = 2.92), tracheotomy (p = 0.002, OR = 3.92), and obstructive lung diseases (p = 0.014, OR = 4.22). The model (incorporating these factors) had an AUC of 0.87, passed validation, and was visualized as a Nomogram. This is the first identification of PNI as a POP risk factor in this population.

Conclusion: The predictive model, which integrates PNI and four other clinical factors, demonstrates favorable discriminative ability in identifying patients at high risk of POP following hematoma extraction for cerebral hemorrhage. By quantifying the risk of POP preoperatively, this model can assist clinicians in stratifying patients, prioritizing targeted preventive interventions (such as nutritional optimization or respiratory care) for high-risk individuals, and thereby contributing to the reduction of postoperative complications.