Impact of a modified Broviac maintenance care bundle on bloodstream infections in paediatric cancer patients

Background: During intensive chemotherapy, bloodstream infection (BSI) represents an important complication in paediatric cancer patients. Most patients carry a long-term central venous access device (CVAD). Improved maintenance care of these vascular catheters may decrease the risk of BSI. Methods: Intervention study (adapted CVAD prevention protocol) with two observation periods (P1: 09-2009 until 05-2011; P2: 09-2011 Oncology and Haematology, until 05-2013); prospective surveillance of all laboratory confirmed BSIs. In P2, ready to use sterile NaCl 0.9% syringes were used for CVAD flushing and octenidine/isopropanol for the disinfection of catheter hubs and 3-way stopcocks. Results: During P1, 84 patients were included versus 81 patients during P2. There were no significant differences between the two patient populations in terms of median age, gender, underlying malignancy or disease status (first illness or relapse). Nearly all CVADs were Broviac catheters. The median duration from implantation to removal of the CVAD was 192 days (Inter-quartile-range (IQR); 110-288 days) in P1 and 191 days (IQR; 103-270 days) in P2. 28 BSI were diagnosed in 22 patients in P1 (26% of all patients experienced at least one BSI) and 15 BSI in 12 patients in P2 (15% of all patients). The corresponding results for incidence density (ID) were 0.44 (CI95 0.29-0.62) for P1 vs. 0.34 (0.19-0.53) BSI per 100 inpatient days for P2 and for incidence rate (IR) 7.76 (5.16-10.86) in P1 vs. 4.75 (2.66-7.43) BSI per 1,000 inpatient CVAD utilization days. In P1, 9 BSI were caused by CoNS vs. only 2 in P2 (IR 2.49; CI95 0.17-4.17 vs. 0.63; CI95 0.08-1.72). In P1 two BSI (7%) lead to early removal of the device. During P2 one CVAD was prematurely removed due to a Broviac-related BSI (6.7%). Conclusion: The preventive protocol investigated in this study led to a reduction of BSI in paediatric cancer patients. This result was clinically relevant but - due to insufficient power in a single centre observation - the difference did not reach statistical significance. The most pronounced trend in BSI reduction was observed for CoNS infections. Thus, improving maintenance care of the CVAD may result in lower CVAD-linked infection rates. The higher acquisition cost of the ready to use NaCl 0.9% flushing syringes and octenidine/propanol hub disinfection were probably balanced by cost savings in the intervention period.