Background: Identification of risks for development of ventilator-associated pneumonia (VAP), which might be identified early after injury, would allow for prognostic estimates and targeting of high-risk cohorts for clinical trials of preventive strategies. This study was performed to develop an equation that can be applied to estimate the probability of pneumonia based on parameters collected in the early postinjury interval.
Methods: Over a 28-month period, patient admissions were reviewed for mechanism and severity of injury, patterns of injury, shock, and need for emergent intubation. Early deaths (<48 hours) were excluded. VAP diagnosis required > or = 10(5) colony-forming units/mL organisms in the bronchoalveolar lavage effluent. Multiple logistic regression analysis was used to develop the prediction equation and estimate odds ratios. The equation was then tested on consecutive patients admitted over a 2-month period.
Results: We reviewed 9,721 admissions (77% blunt, 23% penetrating). VAP incidence was 5.6%. Overall mortality was 2% (21% for patients with VAP vs. 1% for no VAP; p < 0.0001). Multiple logistic regression analysis for all patients produced the following equation: f(x) = -3.08 - 1.56 (MOI) - 0.12 (GCS) + 1.37 (SCI) + 0.30 (chest AIS) + 1.87 (lap) + 0.67 (tx) + 0.05 (ISS) + 0.66 (int), where MOI is mechanism of injury (penetrating = 1, blunt = 0), GCS is Glasgow Coma Scale score, SCI is spinal cord injury (yes = 1, no = 0), lap is emergent laparotomy (yes = 1, no = 0), ISS is Injury Severity Score, tx is units of blood transfused in the resuscitation room, and int is intubation in either the field or the resuscitation room (yes = 1, no = 0). The probability of VAP was calculated as follows: P(VAP) = e(f)(x)/1 + e(f)(x). This formula was concordant in 95% and discordant in 5%.
Conclusion: It is possible to accurately predict risk for VAP in trauma patients based on data available early after injury. This calculation could be useful for counseling families relative to prognosis and research protocols, and addressing hospitalization issues with third-party payors.