We report a significant reduction in the CLABSI rate in a tertiary Australian ICU from a peak of 1.53 to 0.23 after the implementation of a multifaceted care bundle. Overall, the CLABSI rate per 1000-line days decreased from 1.1 in the preintervention period to 0.32 in the postintervention period. There were seven fewer cases of CLABSIs in the postintervention period. We estimated a total reduction in ICU length of stay of 16.8 days, hospital length of stay of 52.5 days, and resultant cost savings of approximately $336,576 (95% CI $190,624- $482,881) based on Markov model to assess the cost effectiveness of the intervention (22, 23).
We noted a significantly greater proportion of patients on invasive mechanical ventilation and significantly greater number of CVAD line days in the postintervention cohort; both are additional risk factors for CLABSIs. In addition, neutropenia and immunosuppression have been identified as independent risk factors for the development of CLABSIs. (24, 25). Despite the significantly greater proportion of patients in the postintervention cohort who had immunosuppression, hepatic disease, or leukaemia and were receiving immunosuppressive medications, there was a significant reduction in the CLABSI rate. This finding demonstrates the clinical effectiveness of the multifaceted intervention strategy and suggests that the observed benefits are causally associated with the intervention (21–25).
We noted that a significant proportion of the lines in the postintervention cohort were inserted in the ICU. A plausible explanation for this observation is complex; it may be related to more patients being transferred to the ICU with fewer CVAD visits or attributed to the intervention strategy of replacing the CVCs inserted outside Canberra Hospital within 48 hours of admission. There was a significant reduction in the use of CVCs with a marked increase in the use of PICC lines in the postintervention cohort. This could be related to the intervention strategy of minimising unnecessary CVC use and the trend toward the use of the PICC line as an alternative central venous access associated with a decreased incidence of CLABSIs. In their systematic review, Xiong Z et al. identified 13 studies that used a decrease in unnecessary CVC use as an intervention strategy and found marked reductions in the CLABSI rate (26).
Arterial puncture during the procedure is a potential complication, with reported rates ranging from 4.2–9.3% for small needle sizes (18G or smaller). (27, 28, 29) Unrecognised arterial puncture may result in the inadvertent placement of a large-bore catheter (> 7 Fr) into an artery, which has been reported to occur in 0.1–1.0% of attempted CVC placements (30, 31). This serious complication can lead to haemorrhage, haematoma formation, pseudoaneurysm formation, arterial dissection, distal embolization, stroke, and death (12, 13, 14, 30).
A review of our three patients revealed that use of real-time ultrasound and demonstrations of the guidewire in a collapsible vessel was insufficient in identifying and preventing arterial dilatation and cannulation. Hence, we recommended and implemented additional steps of involving two-point verification of venous placement of the cannula and guidewire before dilatation, as part of the policy. Venous pressure measurement using column manometry or pressure transduction techniques are feasible and more reliable in identifying arterial puncture and preventing arterial dilation and cannulation. We report no incidences of any arterial cannulation in post intervention period.
In the literature, multiple studies have used educational interventions for CLABSI reduction (32, 33, 34, 35). Some studies focused on education in the bundle alone (33, 34, 35), in combination with e-learning modules (36) and lectures (37, 38, 39). Some studies focused on check lists (40), reminders (40) and posters (41). Fewer studies have used only ICU-level strategies as interventions (42, 43). Our study has the advantage of demonstrating multidimensional interventional strategies involving multidisciplinary teams from four different specialties. Unlike other studies, we implemented the interventions institutionally rather than at the departmental level. We targeted multiple audiences, such as senior and junior medical practitioners and nursing staff from various subspecialties.
Most of the intervention studies have demonstrated the immediate efficacy of the intervention with a reduction in CLABSI incidence (44). However, fewer studies have shown sustained efficacy beyond one year of intervention (45). The strengths of our study include the use of a large patient cohort, multidimensional approaches, and multidisciplinary collaboration, which resulted in significant cultural and practice changes across the organisation and the sustainability of the intervention effect beyond one year. Other strengths of our study include the availability of considerably important patient cohort population characteristics and the application of the current standard CLABSI definition throughout the entire study period.
This study is limited by its single-centre, retrospective, observational design, limited external validity and inability to demonstrate robust direct causality. These limitations are comparable to those of previous similar studies. Additionally, the before-after comparison approach may need to adjust for unmeasured temporal trends that could affect CLABSI incidence. Finally, although the data around most components of bundle compliance are recorded, it is difficult to measure the compliance of all the bundle elements. We lack certainty regarding the most effective component of CVAD bundle in minimizing adverse events. CVC proceduralists may have accumulated experience in placing CVCs over study period. However, every year, about 28 to 36% of Junior Medical Officers (JMOs) commencing their rotations did not have any prior CVC insertion experience (Supplementary table 4).