Incidence of nosocomial respiratory tract infection in pediatric intensive care unit of University Hospital Center (UHC) of Tirana

Abstract

Background: Chemotherapy-induced neutropenia is a risk for serious infection.Septic shock still causes high mortality among cancer patients. We aimed at identifying predictive factors of mortality in neutropenic patients (ANC < 0.5 x 109/l) who developed septic shock.Methods: All cases of septic shock defined according to the ACCP/SCCM criteria observed in our department (Hematology, Aziza Othmana University Hospital) between 2005 and 2009 were included in this study. All clinical, biological and microbiological data were collected at the onset of septic shock and during outcome. Initial score on the Sequential Organ Failure Assessment (SOFA score) was performed for each patient. Statistical analysis was performed using Pearson test.Results: Thirty three septic shock were observed in patients with hematological malignancies:Acute leukemia (31),lymphoma (2).Median age was 27 years (range, 3-67).58% of septic shock were observed during chemotherapy induction phase .Median time for occurrence of septic shock was 15 days from onset of neutropenia (range, 2-32). Microbiological documentation was obtained in 23 cases (69.6%).The bacteria involved were:Klebsiella (9),Pseudomonas (7), Stenotrophomonas (4), E.coli (1), Octhrobacter anthropi (1) and Acinetobacter (1). 9 (39.1%) isolates were resistant to broad spectrum antibiotics. 20 patients (60.6%) developed ARDS during the evolution. Initial SOFA score was >11 (i.e. a risk of mortality of 95%) in only 6 patients (18.1%).Only 8 patients (24.2%) were admitted in ICU with a median time of 1.8 day (range, 1-3 days).Day 30 mortality was 85%.By univariate analysis predictors of mortality were: Disease status (p = 0.009), neutropenia lasting more than 15 days (p = 0.012),fever for more than 3 days in patient on antibiotherapy (p = 0.009), hemoglobin level <50 g/l (p = 0.038), isolate resistant to piperacillin/tazobactam (p = 0.025), presence of clinical symptoms from more than 1 site (p = 0.008), patients not on imipenem antibiotherapy at the onset of septic shock (p = 0.019), occurrence of ARDS during evolution of septic shock (p = 0.003) and non-admission in ICU (p = 0.043). No independent predictor found in multivariate analysis.Conclusion: This study revealed that several factors play a significant role in mortality during septic shock. Despite low proportion of patients with baseline high SOFA score, the mortality rate in our study was very high highlighting the need for appropriate management and early admission in ICU to improve outcome.