Utility of Pneumonia Severity Index in hospitalized patients with pneumonia in intensive respiratory care units

INTRODUCTION: The aim of this study was to evaluate the prognostic value of Pneumonia Severity Index (PSI) for prediction of 30-day mortality in patients admitted to intensive care unit (ICU) for community-acquired pneumonia (CAP). In patients with CAP, comorbidities, complications, and physical, laboratory, radiological and microbiological findings were evaluated relative to their prognosis. PATIENT AND METHODS: In the study, 197 patients with CAP, hospitalised at ICU of Department of Respiratory Medicine, University Hospital Olomouc between 2008 and 2012, were enrolled. Risk factors according to PSI were assessed in all patients. RESULTS: In the studied cohort of patients with CAP, mean values of PSI were 115.4 ± SD 30.4 points. Overall, 29 patients (14.7 %) deceased. When comparing deceased and survived patients, statistically significant differences were found in PSI (mean ± SD: 137.4 ± 26.1 vs 111.7 ± 29.6 points, p < 0.0001), age (mean ± SD: 76.3 ± 12.9 vs 65.5 ± 14.7 years, p < 0.0001), incidence of heart diseases (86.2 % vs 67.3 %, p = 0.04) and ischaemic heart disease (58.6 % vs 38.7 %, p = 0.04). Assessment of physical and laboratory findings showed that deceased patients had significantly increased incidence of tachycardia above 90/min (51.7 % vs 27.4 %, p = 0.01), tachypnoe above 30/min (37.9 % vs 13.7 %, p = 0.001) and acidosis with pH < 7.35 (27.6 % vs 8.9 %, p = 0.004) comparing to survived patients. No significant correlation between PSI and the length of hospitalisation in survived patients was observed. In patients with Staphylococcus sp. and Klebsiella pneumoniae infection, longer hospitalisation period was observed. Comparison of other parameters such as comorbidities, physical and laboratory findings, and pathogens showed no significant differences when comparing deceased to survived patients. CONCLUSION: Our study showed that PSI represents an important predictor of 30-day mortality in patients with CAP at ICU, but does not correlate neither with the length of hospitalisation nor with particular pathogens. Independent negative prognostic factors in CAP were age, incidence of heart diseases (most importantly ischaemic heart disease), tachycardia, tachypnoe and acidosis. Staphylococcus sp. and Klebsiella pneumoniae infection led to longer hospitalisation period. All these factors point out the need for increased care in CAP patients.