Early Detection of Ventilator-Associated Pneumonia in a Neurosurgical Patient: Do Biomarkers Help Us?

Abstract

Background: Ventilator-associated pneumonia (VAP) remains a significant complication in patients undergoing mechanical ventilation. It is particularly prevalent in neurosurgical patients, contributing to high morbidity and healthcare costs. Early diagnosis and timely treatment are crucial for preventing the progression of VAP. However, diagnosing VAP remains challenging because no diagnostic tool or biomarker can reliably confirm the condition. Nevertheless, biomarkers remain the most frequently used surrogates for VAP. Therefore, this study aimed to evaluate the predictive value of interleukin-6 (IL-6), procalcitonin (PCT), and C-reactive protein (CRP) and look at relevant clinical markers for the early detection of VAP. Additionally, the impact of the use of IL-6 or PCT/CRP on clinical decision-making in the diagnosis and management of VAP in neurosurgical patients was explored.

Methods: In this retrospective single-center study, we screened 1,817 neurosurgical intensive care patients between January 2020 and December 2023. In the first step, the ability of IL-6, PCT, and CRP to predict VAP was tested. We distinguished microbiologically confirmed VAP (mcVAP), suspected VAP (suspVAP ), and non-VAP. In the second step, two cohorts were compared: patients monitored with IL-6 alone and those monitored with PCT and CRP. Here, we compared the relevance of these markers for treatment decisions, antibiotic usage, and clinical parameters.

Results: A total of 240 neurosurgical patients fulfilled the eligibility criteria: 155 in the IL-6 group and 85 in the PCT/CRP group. The IL-6 level on the day of mcVAP and suspVAP was 106 pg/mL (IQR: 58.3-259.0) and 112 pg/mL (IQR: 58.3-259), respectively, whereas it was 33.55 ng/L (IQR: 19.6-59.2) in non-VAP patients (p<0.001; η² = 0.14, AUC 0.82 and 0.81, respectively). PCT also showed significant differences, although with a small effect size (η² = 0.008, p = 0.010). For CRP, no significant differences were observed (p = 0.317). In the IL-6 group, the start of treatment did not differ from that in the PCT/CRP group. The duration of antibiotic administration was slightly shorter in the IL-6 group than in the control group, although the differences were not statistically significant (6.25 (±3.73) and 5.9 (±2.23) days versus 7.37 (±4.95) and 7.67 (±2.65) days; p = 0.339 and p = 0.214, respectively).

Conclusion: While PCT has diagnostic value, IL-6 has superior predictive value for VAP, which is reflected in its high effect size and shorter duration of antibiotic treatment, although the difference was not significant. This study suggests that incorporating IL-6 as a routine biomarker may improve the early recognition of VAP, potentially optimizing treatment strategies in the neurosurgical ICU setting. However, the major limitation lies in the study's retrospective design, which limits its generalizability.

Keywords: interleukin (il)-6; neurosurgical intensive care unit; procalcitonin (pct); serum biomarkers; ventilator-associated pneumonia (vap).

Figure 1. Exclusion tree

The exclusion tree outlines the selection process of 1,817 neurosurgical patients. The final cohort of 240 patients was divided into the CRP + PCT group (n = 85) and the IL-6 group (n = 155).

Figure 2. Interval plots on day 0 by the VAP status

Interval plots of (A) IL-6 levels (pg/mL) on day 0 for non-VAP, suspVAP, and mcVAP; (B) CRP levels (mg/L) on day 0 across the three groups; (C) PCT levels (ng/mL) on day 0; (D) WBC levels (×10³/µL) on day 0; and (E) PaO₂/FiO₂ ratio on day 0 illustrating the differences in oxygenation status. IL-6: interleukin-6, PCT: procalcitonin, CRP: C-reactive protein, WBC: white blood cell, VAP: ventilator-associated pneumonia

Figure 3. IL-6 in the repetitive assessment

The graphs illustrate the median IL-6 levels over time (days -4 to +4) in three patient groups: non-VAP, mcVAP, and suspVAP. Median values are plotted for each day, highlighting trends and differences between groups. IL-6: interleukin-6, VAP: ventilator-associated pneumonia

Figure 4. PCT in the repetitive assessment

The graph illustrates the median PCT over time (days -4 to +4) in three patient groups: non-VAP, mcVAP, and suspVAP. Median values are plotted for each day, highlighting trends and differences between groups. PCT: procalcitonin, VAP: ventilator-associated pneumonia

Figure 5. CRP in the repetitive assessment

The graph illustrates the median the CRP level over time (days -4 to +4) in three patient groups: non-VAP, mcVAP, and suspVAP. Median values are plotted for each day, highlighting trends and differences between groups. CRP: C-reactive protein, VAP: ventilator-associated pneumonia

Figure 6. White blood cells in the repetitive assessment

The graph illustrates the median WBC levels over time (days -4 to +4) in three patient groups: non-VAP, mcVAP, and suspVAP. Median values are plotted for each day, highlighting trends and differences between groups. WBC: white blood cell, VAP: ventilator-associated pneumonia

Figure 7. P/F levels in the repetitive assessment

The graph illustrates the P/F levels over time (days -4 to +4) in three patient groups: non-VAP, mcVAP, and suspVAP. Median values are plotted for each day, highlighting trends and differences between groups. VAP: ventilator-associated pneumonia

Figure 8. ROC curves for IL-6, PCT, and CRP

The combined ROC curve was used to compare the diagnostic performance of A) IL-6 on day -1, B) IL-6 on day 0, C) PCT on day 0, and D) CRP on day 1 based on their significance levels for differentiating between suspVAP and non-VAP patients and between mcVAP and non-VAP patients. The sensitivity, specificity, area under the curve (AUC), and odds ratios (ORs) with 95% confidence intervals are displayed for each comparison. The dashed diagonal line represents the line of no discrimination (AUC=0.5). IL-6: interleukin-6, PCT: procalcitonin, CRP: C-reactive protein, ROC: receiver operating characteristic, WBC: white blood cell, VAP: ventilator-associated pneumonia