Predictive values of serum amyloid A, toll-like receptor 4, and sTREM-1 for ventilator-associated pneumonia in elderly patients undergoing abdominal surgery with tracheal intubation and general anesthesia.

OBJECTIVES

This study aimed to conduct a retrospective study to identify inflammatory biomarkers for predicting ventilator-associated pneumonia in elderly patients.

METHODS

Our retrospective study included 265 elderly patients (age ≥60 years) undergoing abdominal surgery with tracheal intubation and general anesthesia, with 93 experiencing varying degrees of ventilator-associated pneumonia during hospitalization, and 172 without. Serum concentrations of serum amyloid A (SAA), toll-like receptor 4 (TLR4), and soluble myeloid triggering receptor 1 (sTREM-1) were measured at 24 h post-operation using enzyme-linked immunosorbent assay. Comparisons of SAA, TLR4, and sTREM-1 and other risk factors at 24 h post-operation between elderly patients with and without ventilator-associated pneumonia were performed.

RESULTS

The study revealed a 35.1% incidence of postoperative ventilator-associated pneumonia among elderly patients. Upregulations of SAA, TLR4, and sTREM-1 were observed in patients with ventilator-associated pneumonia. Chronic obstructive pulmonary disease, smoking, and tracheal intubation were identified as independent risk factors. The joint prediction model was demonstrated with superior predictive accuracy (area under the curve = 0.89) compared to individual biomarkers. Correlations with procalcitonin further supported the predictive potential of SAA, TLR4, and sTREM-1 in an inflammatory response.

CONCLUSIONS

SAA, TLR4, and sTREM-1, particularly when combined, serve as valuable prognostic indicators for postoperative ventilator-associated pneumonia in elderly patients undergoing abdominal surgery with tracheal intubation and general anesthesia. The joint prediction model offered a promising tool for early risk assessment.