Risk factors for slowly resolving pneumonia in the intensive care unit

Abstract

Slowly resolving pneumonia (SRP) poses early challenges for identification and medical expense for clinicians in intensive care units (ICUs); to date, the literature has been very limited in this regard. This was a retrospective and cohort-based study in the ICU of a university-affiliated hospital in Shanghai. Medical records of pneumonia patients in the ICU between April 2008 and February 2011were reviewed retrospectively to evaluate the risk factors for SRP. In all, 106 pneumonia patients in the ICU were identified as immune-competent with a diagnosis of bacterial pneumonia. There were 62 (58.49%) patients who showed SRP and their radiographic infiltrations were completely resolved between 5 weeks and 8 weeks. Multivariate logistic regression analysis demonstrated that initial treatment with an inappropriate antibiotic, multilobar infiltration, and a high CURB-65 score were independent risk factors for SRP, with odds ratio (OR) values of 8.338 [95% confidence interval (CI) 2.117-32.848], 11.184 (95% CI 2.526-49.514), and 2.329 (95% CI 1.172-4.626), respectively. The length of the ICU stay in the SRP group was twice as long as that of the normally resolving pneumonia (NRP) group (62.27±73.73 vs. 32.25±23, p=0.002). The 28-day and 60-day mortality rates in the SRP group were 17.74% and 25.81%, respectively. In addition, the 60-day mortality rate was significantly higher in the SRP group than the NRP group (25.81% vs. 6.82%, respectively; p=0.012). Moreover, SRP was an independent risk factor for 60-day mortality (OR 5.687, 95% CI 1.334-24.240). Treatment with an inappropriate antibiotic, multilobar infiltration, and a high CURB-65 score were independent risk factors for SRP.