Biomarkers associated with post-operative pneumonia: a systematic review and meta-analysis

BACKGROUND: Post-operative pneumonia is a commonly occurring surgical complication associated with poor patient outcomes. This study aimed to synthesise pre-operative and post-operative blood-based biomarkers associated with post-operative pneumonia. METHODS: Electronic databases were searched up to April 14th, 2026. Primary studies investigating blood-based biomarkers in adults hospitalised after surgery were included. Meta-analysis was performed using the random effects model to compare pooled data for pneumonia and no-pneumonia groups using standardised mean difference. Risk of bias was assessed using the ROBINS-E tool. RESULTS: Thirty-seven studies (n = 15,842 patients) were included, with an overall pneumonia rate of 17.8% [15.8-20.0%]. One hundred and fifteen biomarkers were identified. Meta-analysis identified that patients who developed post-operative pneumonia had significantly lower platelet-to-neutrophil ratio (p < 0.0001), red blood cell count (p = 0.001), haemoglobin (p = 0.002), albumin (p = 0.005) and lymphocyte count (p = 0.008) and significantly higher monocyte-to-lymphocyte ratio (p < 0.0001), systemic immune inflammation index (p < 0.001), systemic inflammatory response index (p < 0.0001) and blood urea nitrogen (p = 0.0001) at pre-operative baseline, compared to those without pneumonia. Patients who developed post-operative pneumonia were associated with significantly higher procalcitonin (PCT) at post-operative day one (p = 0.03), two (p = 0.0001), three (p = 0.0004) and six (p < 0.0001); higher C-Reactive Protein (CRP) at day two (p = 0.02), four (p < 0.0001) and five (p < 0.0001); higher interleukin (IL)-6 at day three (p = 0.001) and four (p = 0.0002); and higher white blood cell (WBC) count at day three (p = 0.01) and day four (p = 0.001) post-operatively, compared to those without pneumonia. In addition, pre-operative CRP was associated with mortality within the patients who later developed post-operative pneumonia (r = 0.70, p = 0.0009). CONCLUSIONS: The results identify blood-based biomarker signals associated with post-operative pneumonia. However, interpretation is limited by heterogeneous pneumonia definitions and inconsistent reporting of diagnosis timing across studies. The certainty of evidence is therefore limited, and thus these biomarkers cannot currently be used for prediction or diagnosis in clinical practice. Further high‑quality, prospective studies are required to establish clinically meaningful thresholds. PROTOCOL REGISTRATION: PROSPERO: CRD42024570654.