If compared with the rates of developed countries, the baseline rate of CLAB found in this study (22.7 per 1000 CL-days) was more than ten-fold higher than the US 1.1 CLAB rate per 1000 CL-days determined by the CDC/NSHN ; and more than ten-fold higher than the 1.4 CLAB rate determined by KISS .
In comparison with global CLAB rates from developing countries, our CLAB baseline rate was considerably higher than the fourth international INICC reports published in 2012 (6.8 CLABs per 1000 CL-days) . Likewise, within the scope of other studies addressing the burden of CLABs in Turkey, our CLAB rate of our study was higher than the rate found in other two studies conducted in Turkey showing 17.6 CLABs per 1000 CL days , and 11.8 CLABs per 1000 CL days .
In studies performed by INICC member hospitals, it was shown that the implementation of a multidimensional approach for CLAB--which includes a bundle of interventions, education, outcome and process surveillance, feedback of CLAB rates, and performance feedback--resulted in significant reductions in rates of CLAB in Argentina (46.63 vs. 11.10 CLABs per 1000 CL-days) ; in Mexico (46.3 vs. 19.5 CLABs per 1000 CL-days) ; in adult ICUs (14.5 vs. 9.7 CLABs per 1000 CL-days) ; and in pediatric ICUs (10.7 vs. 5.2 CLABs per 1000 CL-days) .
The INICC multidimensional approach for CLAB included the following elements. First, the implementation of an infection prevention bundle based on the guidelines published by the SHEA and IDSA , which provide evidence-based recommendations and cost-effective infection control measures, which can be feasibly adapted to the ICU setting in developing countries. Second, education of HCWs about infection preventive measures. Third, CLAB outcome surveillance by applying the definitions for CLAB developed by the U.S. CDC/NHSN [21, 22]. Fourth, CLAB process surveillance to monitor compliance with easily measurable infection control measures, including HH performance. Fifth, feedback of CLAB rates. Sixth, performance feedback of process surveillance, particularly, by reviewing and discussing charts results at monthly infection control meetings.
In our study, patients’ characteristics, such as age, gender, and underlying diseases showed similar patient intrinsic risk in both study periods. But ASIS score, CL use, and CL duration were higher during the intervention period, meaning that the patient intrinsic risks were higher in the intervention period. During the implementation of the INICC multidimensional approach, we found an improvement in process surveillance rates, with HH compliance improved by 52%, compliance with date on administration set improved by 17%, compliance with placed sterile dressing improved by 20%, and compliance with correct condition of dressing was high during both periods. During the study period, the high CLAB rate at baseline was reduced from 22.7 to 12.00 per 1000 CL-days, showing a 39% CLAB rate reduction and evidencing the effectiveness of the applied multidimensional approach.
Our study can be compared with an earlier bundle study , and a number of important differences between them can be mentioned. First, this previously published bundle included five elements. In contrast, we included eleven. Second, compliance was not measured for any of the bundle components, whereas we checked compliance of 5 bundle components. Third, characteristics of patients during baseline and intervention periods were not collected nor analyzed so as to check and compare such individual features, whereas we did and could find that our patients were statistically similar during both periods. Fourth, the follow-up period was 18 months, whereas we included a 36-month follow-up period. Fifth, intervention included only a bundle and a check list, whereas our study included the above-mentioned 6 simultaneous interventions. Finally, microorganisms responsible for CLAB were not provided, whereas in our study we included the CLAB microorganism profile for both baseline and intervention periods. The most important differences were measurements of the population’s features and compliance with bundle elements, which allowed us to analyze the real impact of our intervention by excluding confounders associated with patients’ characteristics and infection control practices.
Regarding the microorganisms profile, we identified a predominance Staphylococcus aureus, coagulase-negative staphylococci spp. and Acinetobacter spp. during both periods, which is similar to the findings of other studies conducted in limited-resource countries [7–10].
This study has several limitations. First, our findings cannot be generalized to all ICU patients from Turkey. However, this study proved that a multidimensional approach is fundamental to understand and fight against the adverse effects of CLAB in the ICU setting of Turkey. Second, the setting of three-month baseline period may be short and might have overestimated the effect of the intervention; however, during baseline period the sample size was good enough, and the confidence intervals for the baseline rate were narrow. Finally, because we did not count on the necessary resources, we were not able to differentiate between early and late onset infections; we could not quantify in detail all the interventions included in our multidimensional approach, such as education; and we could not quantify compliance with some of the components of our bundle. Therefore, we could not evaluate the components’ individual implications or other contextual factors related to the ICU or hospitals individually. Nevertheless, our main goal was to reduce the high baseline CLAB rates found in our ICU, and although our interventions were inexpensive, the individual evaluation would have required more allocation of time, contributing to unnecessary harm for ICU patients. Fortunately, as from January 2012, we have been able to collect all these process surveillance data.