![[Visit Client Website]](images/banner.gif)
Bacterial bloodstream infections (BBSIs) contribute significantly to transplant related mortality (TRM) post allogeneic hematopoietic cell transplant (alloHCT). In 2006, a new central-line associated bloodstream infection (CLABSI) protocol was established at our institution. We report the incidence rate and sensitivities of both Gram negative rod (GNR) and Gram positive cocci (GPC) BBSIs during Pre-CLABSI era (2004-06), CLABSI era (2006-10), and post-CLABSI era (2010-13).
Methods
A retrospective chart review between 2004-2013 was conducted. 100 person-month bacterial infection incidences were calculated and compared by Poisson regression analysis. Patients did not receive prophylactic anti-bacterial antibiotics and piperacillin/tazobactam was started at the onset of fever.
Results
Between 2004-13, 302 BBSIs were identified in 190 patients (mean age 9.97 years). Malignant 111 (58.4%), Non-malignant 79 (41.6%); donor source: Marrow 71 (37.4%), Peripheral Blood Stem Cell 59 (31.1%), Cord blood 60 (31.5%). Conditioning regimens: myeloablative= 86 (45.3%), reduced toxicity= 57 (30%), reduced intensity= 47 (24.7%); incidence of neutrophil engraftment by day +30 was 82%, and incidence of grade II-IV aGVHD was 44.2%,
For the different time periods of immune reconstitution, the rate per 100 person months of bacterial infections decreased as shown in table 1. Of the BBSIs for each time period, the proportion of GPC BBSIs decreased (2004-06:56%, 2006-10:53.1%, 2010-13:39.2%, respectively), 2004-06 vs. 2010-13, p=0.048.
GNR antibiotic resistance increased to levofloxacin (4.4% from 2004-06 to 17.2% from 2010-2013), decreased to cefepime (13.3% from 2004-06 to 7.7% from 2010-13), and remained stable to piperacillin-tazobactam (7.7% in 2004-06 to 9.1% from 2010-13). GPC antibiotic resistance to vancomycin increased from 3.4% in 2004-06 to 10.5% in 2010-13.
Lastly, in our multivariate analysis RTC regimen was a significant risk factor for the development of piperacillin-tazobactam resistance in GNR (OR 4.421, 95% CI 1.331-14.69, p=0.015). TRM was significantly higher in BBSI patient between 2004-06 (27.9%) vs. 2010-13 (9.3%) (p=0.03).
Discussion
Our analysis showed a decrease in BBSI rates after CLABSI protocol was implemented at our institution and a significant decline in BBSI in the post-CLABSI era. The proportion of GPC infections decreased over this time period. However, further investigation on factors associated with decrease in BBSI in these different time periods is ongoing. The rise in vancomycin resistance (likely due to its widespread use as empiric antibiotic choice for alloHCT patients with fever), warrants further investigation.
|
2004-06 |
2006-10 |
2010-13 |
p (2004-06 vs 2010-13) |
Day 0 to +30 |
42.1% |
33.4% |
13.8% |
0.016 |
Day +31 to +100 |
61.9% |
30.5% |
12.7% |
<0.001 |
Day +101 to +180 |
62.4% |
45.7% |
19.8% |
0.001 |
Day +181 to +365 |
26.3% |
19.9% |
5.0% |
<0.001 |