Reduction of catheter-associated bloodstream infections through procedures in newborn babies admitted in a university hospital intensive care unit in Brazil

1. Laboratório de Microbiologia, Instituto de Ciências Biomédicas, Universidade Federal de Uberlândia, Uberlândia, MG. 2. Departamento de Neonatologia, Faculdade de Medicina, Uberlândia, MG. Address to: MSc. Daiane Silva Resende. Laboratório de Microbiologia/ICBIM/UFU. Av. Pará 1720, Campus Umuarama, 38400-902 Uberlândia, MG, Brasil. Phone: 55 34 9204-5388 e-mail: daianeresende_bio@hotmail.com Received in 03/05/2011 Accepted in 29/07/2011 ABSTRACT


Reduction of catheter-associated bloodstream infections through procedures in newborn babies admitted in a university hospital intensive care unit in Brazil
Redução de infecções de corrente sanguínea associadas ao cateter, após procedimentos em neonatos admitidos em uma unidade de teapia intensiva de um hospital universitário no Brasil Daiane Silva Resende 1 , Jacqueline Moreira do Ó 1 , Denise von Dolinger de Brito 1 , Vânia Olivetti Steffen Abdallah 2 and Paulo Pinto Gontijo Filho 1 Hospital-acquired infection is a significant cause of morbidity and mortality in neonatal intensive care units (NICUs) 1 and leads to an increase in the length of stay and hospital costs 2 .Bloodstream infections (BSIs) are the most common healthcareassociated infection in neonates [3][4] .Very limited information is available from developing countries on infection rates per hospital-days or device-days including Brazilian units 5 .Incidence density rates per healthcare-associated infection in few studies [6][7] were higher than those observed in most studies in the United States or Europe 8 .
Risk factors associated to BSI include prematurity, low birth weight, poor skin integrity, low gestational age, use of parenteral nutrition (PN), central venous catheter (CVC), prolonged duration of NICU stay, and exposure to broad-spectrum antibiotics 3,8 .
To reduce these infections, guidelines and, recently, bundles of practices have been shown effective in preventing a large proportion of catheterassociated BSI (CA-BSI) 2,9-10 .These bundles include ongoing surveillance, healthcare workers' education, a trained team of caregivers for catheter insertion and care, and strategies designed to prevent intraluminally and extraluminally acquired BSIs [11][12] .
The aim of this study was to reduce CA-BSIs in a Brazilian NICU by means of a bundle, including few evidence-based strategies or procedures recommended by the Centers for Disease Control and Prevention (CDC) to reduce risks of CA-BSIs.

Design of the study
The study was conducted in the NICU of the University Hospital of Uberlândia.All neonates admitted between August 2008 and September

RESULTS
Resende DS et al -Reduction of catheter-associated bloodstream infections in neonates 2009 who required at least one CVC were included in this trial.An epidemiological surveillance for evaluation of nosocomial infection occurrence in the NICU was performed by a team trained by a physician according to the National Healthcare Safety Network, with daily visits to the unit to search BSI and possible risk factors and to the hospital laboratory to recover positive cultures isolated from infections in neonates hospitalized in the period of the trial.Patients were followed from their entry into the study to their discharge or death.An intervention to reduce BSI was followed up from April 2009 to May 2009.Standard definitions for healthcare-associated infections were used 13 .The ethical approval was obtained from the ethics committee of the Federal University of Uberlândia according to the Health Ministry demands.
Intervention Before implementing any of the components of the study intervention, one physician was designated as the leader.This leader was supposed to train the colleagues about intervention measures.Literature was assessed for methodological quality and applicability and based primarily on categories IA and IB, and guidelines were drafted for CVC insertion and maintenance.
The intervention focused on group sessions and feedback on pathogenesis and data of CA-BSI per 1,000 CVC days in the unit before intervention.Emphasis was placed in five evidence-based procedures (bundle) recommended by the CDC and identified as having the greatest effect on the rate of CA-BSI and the lowest barriers to implementation 9,14 .The recommended procedures are hand hygiene, using full-barrier precautions during the insertion of CVCs, cleaning the skin with chlorhexidine 0.2%, avoiding the femoral site if possible, and removing unnecessary catheters besides better knowledge about prevention of CA-BSI in neonates.At the first meeting, the pathogenesis and rates of BSI in the unit before intervention (August 2008 to March 2008) were discussed.In the second meeting, the procedures to prevent BSI were addressed.Emphasis was placed on identifying solutions to overcome difficulties to comply with the bundle.Otherwise, visual displays with A3-size color posters that emphasized care with CVC were distributed at strategic points of the unit.

Post-intervention
Epidemiological surveillance was also performed in the postintervention period ( June 2009 to September 2009).Rates of BSI on this period were calculated per 1,000 CVC days and then presented to healthcare workers as a feedback on the epidemiological indicators of CA-BSI in the unit.Differences in proportions of BSI between the two periods were compared by conducting a statistical analysis.

Microbiological techniques
Hemocultures: blood cultures were collected based on clinical criteria, including apnea, bradycardia, temperature instability, feed intolerance, increased oxygen requirement, fever, and lethargy.To avoid contamination of blood cultures, a specialized nurse drew blood after meticulous skin cleaning.Blood was processed using the BACTEC 9240 (Becton Dickinson Diagnostic Instrument Systems, Sparks, MD, USA) method.Positive cultures were further subcultured in blood agar plates.

Microorganism identification
The identification of samples was performed through traditional phenotype tests 15 .

Statistical analysis
Two periods were compared to evaluate the impact of the intervention: the period before the start of intervention (Group 1) and the period after initiation of the intervention (Group 2).Categorical variables were compared using the Chi-square test or the Mann-Whitney U test when appropriate.Univariate analysis of risk factors for BSI was performed by applying the Chi-square test.Statistical significance level was set at p≤0.05.The variables with p≤0.05 in the univariate analyses were included in the multivariate logistic regression model to identify independent risk factors for BSI.Software BioStat version 5.0 (Brazil) was used for multiple logistic regression, and GraphPad Prism (Version 5.0, San Diego, USA) was used for other analyses.
A total of 251 neonates submitted to CVC use were included in this study.One hundred forty-four neonates were admitted before the intervention (Group 1), and 107 neonates were admitted after the implementation of the five evidence-based procedures (bundle) in the NICU (Group 2).No significant differences were observed between the patients' characteristics in both groups (Table 1).

DISCUSSION
Figure 1 shows that incidence of BSI decreased significantly after the intervention measures from Group 1 to Group 2, 32% to 19.6% (p=0.04).The incidence rate of CA-BSI per 1,000 CVC days also decreased significantly between the groups from 24.1 per 1,000 CVC days to 14.9 per 1,000 CVC days, and that of CA-BSI per 1,000 neonate days decreased from 21.9 per 1,000 neonate days to 12.5 per 1,000 neonate days; however, the average central line length was the same (13.2) in both groups (Table 1).
Analyses were repeated for a subgroup of low-birth-weight infants (<1,500g).A total of 113 low-birth-weight infants were included, 67 in Group 1 and 46 in Group 2. There were 25 episodes of cultureproven sepsis in Group 1 and 7 episodes in Group 2. The rate of BSI decreased significantly from Group 1 to Group 2 (37.3% to 15.2%, respectively, p<0.01).Central line use was similar in both groups.The incidence density rate of CA-BSI per 1,000 CVC days decreased between groups (26.2 per 1,000 CVC days to 10.2 per 1,000 CVC days), and CA-BSI per 1,000 neonate days decreased from 23.6 per 1,000 neonate days to 9.3 per 1,000 neonate days (Table 2).In this report, we affirm the effectiveness of evidence-based prevention bundles in reducing the rate of CA-BSI in our NICU patients.To our knowledge, this is the first study to evaluate the impact of an intervention aimed to decrease CA-BSI among neonates in a Brazilian NICU.
Neonatal BSI is a commonly encountered hospital infection among neonates, mainly premature infants, particularly those who require a CVC 1 .Unfortunately, the use of these devices is associated with a considerable risk of infection 16 .Of note in both adult and neonatal ICUs, in developing countries, such as Brazil, deviceassociated infection rates are reported to be several-fold higher than those detected by The National Nosocomial Infection Surveillance in the United States, 16.1 for five developing countries against 6.6 for the US 17 .For example, CA-BSIs per 1,000 CVC days occur at rates of 3.1-6.4 in NICUS according to The National Healthcare Safety Network and 22.7 in seven units located in the Brazilian cities 3 .Our preintervention rate of CA-BSI was similar (24.1) and much higher than those from developed countries 17 .
CA-BSIs are largely preventable, and many studies have been published documenting the effectiveness of interventions; reports of several studies have revealed a reduction in CA-BSI rates.Most available data from these interventions are related to adult units 14,18 and some in pediatrics units 10,19 .In settings with limited resources, in infants, few data have been published 2,20 .
In our NICU, we used simple and inexpensive intervention to increase compliance with evidence-based infection control practices and decrease CA-BSI rates.After implementation of the bundle, CA-BSI rate, CA-BSI per 1,000 CVC days, and CA-BSI per 1,000 neonate days had decreased from 32% to 19.6%, 24.1 to 14.9, and 21.9 to 12.5, respectively.The rate of BSI also decreased significantly from Group 1 to Group 2. In low-birth-weight neonates, the rates decreased from 37.3% and 15.2%.We observed a reduction of CA-BSI per 1,000 CVC days and CA-BSI per 1,000 neonate days, observed in association with the new practices, from 26.2 to 10.2 and 23.6 to 9.3, respectively.We have shown that specific education and recommended procedures of CVC insertion and care and feedback to healthcare workers (HCWs) can result in significant reduction on rates of CA-BSI.The overall compliance to the catheter care policy improved significantly in the post-intervention period.Leadership came from within the NICU, which, we think, increased our ability to achieve cooperation from the healthcare workers and to create a change in culture regarding CA-BSI prevention and might explain the success of our intervention.
The authors declare that there is no conflict of interest.

CONFLICT OF INTEREST REFERENCES
Coagulase-negative Staphylococcus is the most common cause of late-onset BSI in our NICU, accounting for 45.8%, similar to the observations in developed countries, where CNS is the most important microorganism causing CA-BSI 8,12 .Other studies have shown previously that CNSes were the most frequent agents of BSI as in our NICU 7 .In contrast, Gram-negative bacilli, mainly Klebsiella spp., Escherichia coli, and Enterobacter spp., were most frequent in developing countries [5][6] .
In this report, the multivariate analysis identified just two independent risk factors for BSI: exposure to ≥3 antibiotics and length of stay ≥8 days.Risk factors found in previous studies include these, as well as parenteral nutrition and very low birth weight in infants 21 .Parenteral nutrition and mechanic ventilation may be stronger predictors of BSI than CVC use only 22 but were not used in a minority of our infants (24.3) and was significant only in the univariate analysis.
There are some limitations in our study.First, we did not collect data of factors that could influence CA-BSI rates such as the line type that was placed and the healthcare worker who placed the line or its duration of use.We are not aware of any market changes, and the procedures have been shown not to change substantially over time; it is likely that the groups remained similar.The postintervention period was only four months, so it was unlikely that there were significant changes.
Second, we did not collect data on preintervention process to measure compliance, and this lack of data decreased our ability to measure the full impact of the bundles, although it is unlikely that the precollaborative performances were markedly different from those in the initial months of the project.
In conclusion, education can significantly improve infection control practices.We implement a simple intervention to reduce these infections in a university NICU.Coinciding with the intervention, the median rate decreased from 24.1 to 14.9 per 1,000 CVC days.We think that it is feasible to sustain this intervention over a long term.The challenge is to continue this reduction, considering the rate of CA-BSI in a university hospital, with more pronounced mobility of students, chiefs, and professionals in the unit.

FIGURE 1 -
FIGURE 1 -Graph on rates of catheter-associated bloodstream infection in the neonatal intensive care unit in the preintervention and postintervention period.