Preventable proportion of intubation-associated pneumonia: Role of adherence to a care bundle

Objective: The aim of the present study was to estimate the preventable proportion of Intubation-Associated Pneumonia (IAP) in the Intensive Care Units (ICUs) participating in the Italian Nosocomial Infections Surveillance in ICUs (SPIN-UTI) network, taking into account differences in intrinsic patients’ risk factors, and additionally considering the compliance with the European bundle for IAP prevention. Methods: A prospective patient-based survey was conducted and all patients staying in ICU for more than 2 days were enrolled in the surveillance. Compliance with the bundle was assessed using a questionnaire for each intubated patient. A twofold analysis by the parametric g-formula was used to compute the number of infections to be expected if the infection incidence in all ICUs could be reduced to that one of the top-tenth-percentile-ranked ICUs and to that one of the ICU with the highest compliance to all five bundle components. Results: A total of 1,840 patients and of 17 ICUs were included in the first analysis showing a preventable proportion of 44% of IAP. In a second analysis on a subset of data, considering compliance with the European bundle, a preventable proportion of 40% of IAP was shown. A significant negative trend of IAP incidences was observed with increasing number of bundle components performed (p<0.001) and a strong negative correlation between these two factors was shown (r = -0.882; p = 0.048). Conclusions: The g-formula controlled for time-varying factors is a valuable approach for estimating the preventable proportion of IAP and the impact of interventions, based entirely on an observed population in a real-world setting. However, both the study design that cannot definitively prove a causative relationship between bundle compliance and IAP risk, and the small number of patients included in the care bundle compliance analysis, may represent limits of the study and further and larger studies should be conducted.