Reduction in catheter-related bloodstream infections in critically ill patients through a multiple system intervention
- First Online:
- Cite this article as:
- Peredo, R., Sabatier, C., Villagrá, A. et al. Eur J Clin Microbiol Infect Dis (2010) 29: 1173. doi:10.1007/s10096-010-0971-6
- 235 Views
In this study, we aimed to determine the utility of a multiple system intervention to reduce catheter-related bloodstream infections (CR-BSI) in our intensive care unit (ICU). A prospective cohort study was undertaken in the medical and surgical ICU at a university hospital. We applied five measures: educational sessions about inserting and maintaining central venous catheters, skin cleaning with chlorhexidine, a checklist during catheter insertion, subclavian vein insertion and avoiding femoral insertion whenever possible, and removing unnecessary catheters. We determined the rate of CR-BSI per 1,000 catheter-days during the intervention (March to December 2007) and compared it with the rate during the same period in 2006 in which we applied only conventional preventive measures. CR-BSI was defined as the recovery of the same organism (same species, same antibiotic susceptibility profile) from catheter tip and blood cultures. We registered 4,289 patient-days and 3,572 catheter-days in the control period and 4,174 patient-days and 3,296 catheter-days in the intervention period. No significant differences in the number of patients with central venous catheters during the two periods were observed: catheters were used in 81.5% of patients during the control period and in 80.6% of patients during the intervention period. During the control period, 24 CR-BSI were diagnosed (6.7/1,000 catheter-days); during the intervention period, 8 CR-BSI were diagnosed (2.4/1,000 catheter-days) (relative risk 0.36; 95% confidence interval [CI] 0.16 to 0.80; p = 0.015). Nurses interrupted the procedure to correct at least one aspect when completing the checklist in 17.7% of insertions. In conclusion, a multiple system intervention applying evidence-based measures reduced the incidence of CR-BSI in our ICU.
Indwelling vascular catheters are often indispensable for the management of critically ill patients; however, they are also a leading source of bloodstream infections in these patients. More than 250,000 vascular catheter-related bloodstream infections (CR-BSI) occur annually in the USA [1–3], resulting in substantial morbidity, mortality, and costs [4–6]. Despite the publication of clinical practice guidelines  on the management and prevention of intravascular catheter-related infection, CR-BSI are common. According to the National Nosocomial Infections Surveillance (NNIS) system of the Centers for Disease Control and Prevention (CDC), the median rate of all types of CR-BSI ranges from 1.8 to 5.2 episodes per 1,000 catheter-days. In Spain, the mean rate of CR-BSI in the National Study of Nosocomial Infections Surveillance in the ICU (ENVIN-UCI) in 2006 was 5 episodes per 1,000 catheter-days . In our medical-surgical intensive care unit (ICU) in 2006, central venous catheters (CVC) were used in 83% of patients and the incidence of CR-BSI was 5.8 episodes per 1,000 catheter-days.
Recently, Pronovost et al.  implemented an evidence-based intervention in 108 ICUs to reduce CR-BSI, designating a team leader for each hospital instructed in the different interventions and responsible for disseminating this information among their colleagues. The evidence-based intervention consisted of five evidence-based procedures recommended by the CDC: hand washing, using full-barrier precautions during the insertion of CVCs, cleaning the skin with chlorhexidine, avoiding the femoral site if possible, and removing unnecessary catheters. A checklist was used to ensure adherence to infection-control practices. Three months after implementing the intervention, their median rate of CR-BSI had decreased from 2.7/1,000 catheter-days at baseline to 0/1,000 catheter-days (p < 0.002), and their mean rate had decreased from 7.7/1,000 catheter-days at baseline to 1.4/1,000 catheter-days (p < 0.002); this improvement was maintained throughout the 18-month study period.
We aimed to evaluate the effect of the evidenced-based intervention proposed by Pronovost et al. in our medical-surgical ICU.
We conducted a prospective cohort study of adult patients admitted to our 16-bed medical-surgical ICU in a 500-bed teaching hospital from March 2007 to December 2007. Our hospital does not attend burn, transplantation, or heart-surgery patients.
We applied a multiple system intervention to nonemergency CVC insertion aimed at lowering CR-BSI based on the recommendations of Pronovost et al. . The intervention consisted of five elements: (1) educational sessions about how to insert and maintain CVCs [9, 10]; (2) cleaning the skin with alcoholic chlorhexidine 0.5% [11, 12]; (3) using a checklist during catheter insertion to ensure adherence to infection-control practices; (4) using the subclavian vein as the preferred insertion site and avoiding the femoral vein if possible; and (5) daily evaluation to determine whether catheters were unnecessary and could be removed [13–18]. The study protocol was approved by the hospital’s institutional review board and informed consent was not required.
Before implementing any of the components of the study intervention, nurses and physicians were instructed during meetings from January 2007 to February 2007. The educational sessions were mainly focused on recommended procedures to control infection, such as hand washing and using full-barrier precautions during the insertion of CVCs. In addition, staff were introduced to the checklist intended to ensure adherence to infection-control practices and instructed to interrupt the procedure whenever these practices were not being followed.
The procedure for CVC maintenance was the same in all of the patients in both the control and intervention periods: guidelines for managing CVCs were posted in each box. The main measures included in these guidelines were the use of sterile gloves for inspecting and cleaning the skin at the catheter insertion site, replacement of administration sets every 72 h, and replacement of administration sets used to administer blood products and parenteral nutrition or propofol every 24 h. The attending physician decided on a daily basis whether to remove the catheter in function of whether the catheter was considered to be unnecessary or catheter-related infection was suspected.
A definitive diagnosis of CR-BSI required that the same microorganism be found in at least one percutaneous blood culture and in a culture of the catheter tip . The catheter was considered to be positive when a semiquantitative (roll-plate) culture of a 5-cm catheter tip yielded >15 colony-forming units (cfu) of a microorganism.
Routine replacement of CVCs to prevent CR-BSI was not permitted [20, 21], and antibiotic or antiseptic impregnated catheters were not used during the study period. Moreover, no other infection-reducing practices were implemented during the study.
We determined the rate of CR-BSI during the application of the evidence-based intervention from March to December of 2007, and we compared it with the rate of CR-BSI obtained in the same period (from March to December) of 2006, in which we applied conventional preventive measures during CVC insertion, such as using full-barrier precautions during insertion and cleaning the skin with chlorhexidine
Demographic data, patient-days, catheter-days, catheter insertion site, APACHE II, mean catheter duration, and the number of CR-BSI were collected prospectively during the intervention period (from March to December 2007) and were compared with the data collected in the same period (from March to December) in 2006.
Descriptive statistics included frequencies and percentages for categorical variables and means and standard deviations for continuous variables. Associations between categorical variables were assessed with the Chi-square test or Fisher’s exact test. Student’s t-test was used to compare groups on continuous variables. Relative risk ratios were used to compare rates of CR-BSI. All statistical tests were two-tailed and significance was set at 0.05. The statistical analysis was carried out using SPSS software, version 11.5.
We registered 4,289 patient-days and 3,572 catheter-days during the control period and 4,174 patient-days and 3,296 catheter-days during the intervention period (p-value: NS). The mean age of the patients was 62 ± 17 years in the control period and 62 ± 16 years in the intervention period (p-value: NS), and the mean APACHE II score was 14 ± 8 in both periods (p-value: NS).
Location and mean catheter duration
Location of catheter
Mean duration of catheter (days)
7.5 ± 4.5
7.1 ± 4.1
In 38 of 214 catheters used during the intervention period (17.7%), nurses interrupted the procedure to correct at least one aspect mentioned in the checklist during the insertion of the catheter. The following aspects were corrected: failure to use a large sterile drape n = 12 (31%); failure to use a mask for all personnel entering the box n = 11 (29%); failure to maintain a sterile field n = 8 (21%); failure to use a hat, mask, and sterile gown n = 5 (13%); failure to use sterile gloves n = 1 (3%); and inadequate cleaning of the insertion site n = 1 (3%).
We have applied the evidenced-based intervention for the prevention of CR-BSI in our ICU since December 2007 and the benefits of this intervention have persisted. During 2008, the rate of CR-BSI was 3.4 episodes per 1,000 catheter-days and during the first 6 months of 2009, it was 1.7 episodes per 1,000 catheter-days, representing an overall rate of 2.7 episodes per 1,000 catheter-days during 18 months follow-up.
We aimed to determine whether following the recommendations of Pronovost et al.  would reduce the incidence of CR-BSI in our medical-surgical ICU. Our results confirm that applying an inexpensive multiple evidence-based system intervention can significantly reduce the incidence of CR-BSI. No expensive technology or additional staff were required for the intervention, and we emphasize that these measures can be easily applied in most ICUs. Furthermore, CR-BSI rates remained lower 18 months after the intervention was initiated.
We also evaluated the relative impact of the individual components of this multifaceted intervention. Among the five recommendations used, the most significant change between the study and control periods resulted from the implementation of a checklist to ensure adherence to infection-control practices during CVC insertion. In 17.7% of cases, nurses interrupted the procedure to correct at least one point mentioned in the checklist during CVC insertion; failure to maintain a sterile field, failure to use a mask for all personnel entering the box, and failure to use a large sterile drape were the most frequently corrected aspects. On the other hand, we also observed a trend toward greater use of the subclavian vein, a trend toward shorter duration of catheter use during the intervention period, and a small tendency to reduce the use of jugular vein, and this is an important point to improve. In addition, educational sessions led by a team of ICU nurses about how to insert and maintain CVCs were repeated twice a year after the implementation of the preventive measures as a complementary measure.
This study has several limitations. First, it was conducted in a single hospital with particular clinical and microbiological characteristics. It would be necessary to conduct a multicenter study to generalize our findings. However, whereas our incidence before the implementation of prevention measures was similar to the mean rate of CR-BSI reported in the National Study of Nosocomial Infections Surveillance in the ICU in Spain (ENVIN-UCI) in 2007 and 2008 (4.6 and 4.9 episodes per 1,000 catheter-days, respectively), it has been much lower since the implementation of these measures. Second, the design of the study did not allow the randomization of patients to the intervention and control groups; instead, we compared the CR-BSI rate during the intervention period with that of a baseline period. However, patients and CVC use in both periods were similar, no additional infection-reducing practices were implemented during the study period compared with the baseline period, and the same months of the year were compared to avoid possible effects from a seasonal trend. These factors support a strong association between the implementation of the intervention and the lower rates of CR-BSI in our ICU.
In summary, our study confirms that the implementation of a simple and inexpensive evidence-based intervention is useful in the sustained reduction of CR-BSI in a medical-surgical ICU.
Conflicts of interest
All authors report no conflicts of interest relevant to this article except David Suarez, who was supported by a research grant from Instituto de Salud Carlos III, Ministerio de Sanidad y Consumo, Spain, and the Departament de Salut, Generalitat de Catalunya (FIS ECA07/041).