Chlorhexidine bathing of the exposed circuits in extracorporeal membrane oxygenation: an uncontrolled before-and-after study

This was an uncontrolled, single center, before-and-after study. Patients who required ECMO for cardiogenic or respiratory failure and were older than 18 years were screened for eligibility. Exclusion criteria included a termination of ECMO support within 48 h and presence of chlorhexidine allergy. The intervention was a daily disinfection of all exposed circuits and hub including ECMO catheter insertion sites with 2% CHG/IPA during ECMO support. An additional file shows this in more detail (see Additional file 1). In order to minimize the performance bias, we trained the intervention practice for one month in August 2018. The intervention group was prospectively enrolled after obtaining consent from the patient or guardian from September 2018 to August 2019 and collected data. The control group was retrospectively collected data on patients who performed ECMO before the intervention from March 2017 to July 2018. During the pre-intervention period, standard practice was for all ECMO patients to receive daily 2% CHG disinfection at the ECMO catheter insertion site according to the infection precaution policy of the hospital [4]. A total of 244 patients were screened during the study period (Fig. 1). Of those, 52 patients who conducted ECMO within 48 h were excluded, and 192 were enrolled in this study. Except for the intervention, all patient care and ECMO management were performed identically. Basically, all patients with central lines received daily 2% CHG bathing to clean neck and chest. The dialysis line was connected to the ECMO hub when RRT was required during ECMO, and post-membrane blood gas analysis was performed when oxygenator dysfunction was suspected. Also, oxygenator was exchanged when oxygenator dysfunction or massive clot was observed. There were no differences in general practice and diagnostic approach since reaching the learning curve for multidisciplinary team setup in ECMO [9]. This project was approved by the Pusan National University Yangsan Hospital Institutional Review Board (05-2018-149) and the study was registered on the Clinical Research Information Service (KCT 0004431). In the intervention group, written informed consent for enrollment or consent to continue and use patient data was obtained from each patient or their legal surrogate. The control group was waived from consent due to retrospective data collection by the IRB.

The primary outcome was the incidence of BSI during ECMO. Consistent with Centers for Disease Control and Prevention/National Healthcare Safety Network criteria, BSI during ECMO was defined as at least one of the following criteria, measured from 48 h after insertion of ECMO catheters to 1 day after termination of ECMO: (1) a recognized bacterial/ fungal pathogen cultured from one or more blood cultures that is not related to an infection at another site and (2) a common commensal organism (e.g., coagulase-negative staphylococcus) in two or more blood cultures collected on different days or from different sites that is not related to an infection at another site and that occurs in the setting of one of the following signs/symptoms: fever (> 38.0 °C), chills, or hypotension [10, 11]. Secondary outcomes were colonization of ECMO catheters, overall mortality, sepsis-related death, ECMO complications, and microbiology of catheter colonization. Catheter colonization was defined as a positive culture of ECMO catheters at the time of removal. Growth of > 15 colony forming units (CFU) from a 5-cm segment of the catheter tip by semiquantitative (roll-plate) culture or growth of > 10² CFU from a catheter by quantitative (sonication) broth culture reflects catheter colonization [12]. Other central venous catheter-related infections were defined by the clinical practice guidelines of the Infectious Disease Society of America (IDSA) [12]. Ventilator-associated pneumonia (VAP) was diagnosed by the IDSA/ATS guidelines [13]. Sepsis-related death was defined as the immediate cause of death was sepsis, or sepsis was a factor in a chain of events leading to death.

All ECMO catheter insertions were performed using a strict aseptic technique by standardized guidelines by an experienced cardiothoracic surgeon using a Seldinger approach under ultrasound guidance [14]. The optimal insertion site for each individual patient was selected by experienced cardiothoracic surgeons. All involved personnel were trained in catheter maintenance procedure, and standardized continuous catheter care was provided according to the Extracorporeal Life Support Organization guidelines [14]. The ECMO system consisted of a polymethylpentene fiber oxygenator system (QUADROXPLS; Maquet Inc., Hirrlingen, Germany) with simplified Bioline-coated circuits (Maquet Inc., Rastatt, Germany). All patients were supported with centrifugal pumps (Maquet Inc., Hirrlingen, Germany). Our hospital had no set protocol for antimicrobial prophylaxis in patients receiving ECMO. Standardized, empirical antibiotic therapy with broad-spectrum drugs was initiated according to anti-infection guidelines for suspected and confirmed infections, when deemed necessary by the attending intensivist. During ECMO support, antibiotics were adjusted according to the clinical course of the disease and culture results. All insertion sites were regularly checked for potential dressing contamination. If clinically indicated, the perfusionists and cardiothoracic surgeon changed the dressing, and cleaned the skin site additionally. There was effective removal of urine and feces.

The catheter samples were harvested aseptically and prospectively collected and cultured at the end of ECMO support. After removal, the sections (5 cm) were cut from the intravascular region, split longitudinally, and immediately transported to the laboratory.

The oxygenator membranes were also harvested aseptically and prospectively collected and cultured. The membrane oxygenator was rinsed through the tube system with sterile physiological saline solution (0.9% w/v NaCl, twice) to remove blood. Subsequently, it was opened under sterile conditions. A hollow fiber membrane smear was taken from the opened membrane oxygenator, using a sterile swab.

The catheter segments and oxygenator swab were submitted to the local microbiology laboratories for routine culture and antibiogram. Blood culture samples from patients at the time of catheter removal underwent standard culture methods. Bacterial growth from oxygenator swab cultures was assessed using a semi-quantitative method [15].

All statistical analyses were performed using R software, version 3.6.2 (http://​www.​R-project.​org) or SPSS version 25 (IBM corp., Armonk, NY). Continuous variables were described as means ± standard deviation and were compared between groups using the t test. Categorical variables were described as numbers (%) and were compared using the χ² test. Two-tailed p values < 0.05 were considered to reflect statistical significance. To determine whether there was a difference in BSI and colonization, we calculated the incidence difference (per 1000 ECMO catheter days) between 2 groups and reported the associated 95% confidence interval (CI). Univariate logistic regression analysis was performed to evaluate the risk factors of BSI during ECMO. Significant factors (p < 0.05) were included in a multiple logistic regression analysis to determine the risk factors of BSI during ECMO. The backward stepwise (likelihood ratio) method was used for multivariate analysis, with entry and removal p values set at 0.05. We graphically compared occurrences of BSI between 2 groups using the Kaplan-Meier method and log-rank test. Also, the cox regression analysis was performed to evaluate the predictors of overall mortality.

A total of 1740 ECMO catheter days in 192 patients were studied. The mean age was 58 ± 14.4 years and the male ratio was 64.1%. The intervention and control group each had 96 eligible patients enrolled during the study period. There was no significant difference of baseline characteristics between the two groups, except Charlson comorbidity index (2 vs. 2.5, p = 0.043) and ratio of RRT during ECMO (53.1% vs. 34.4%, p = 0.009, Table 1).

The intervention of disinfection with 2% CHG/IPA of exposed circuits and hub significantly decreased the incidence of BSI during ECMO from 11.7/1000 ECMO catheter days to 2.3/1000 ECMO catheter days (difference 9.4, 95% CI 1.5–17.3, p = 0.019) in the study period (Table 2). As well, the incidence of ECMO catheter colonization was significantly higher in the control group (10.5/1000 ECMO catheter days vs. 2.3/1000 ECMO catheter days (difference 8.3, 95% CI 0.7–15.8, p = 0.032). The positive rate of oxygenator culture was higher in the control group (14% vs. 1.1%, p = 0.002). The incidence of VAP was higher in the control group (25.7% vs. 10.2%, p = 0.016). Other central venous catheter-related BSI was higher in the control group (21.1% vs. 7.9%, p = 0.023).

There were no significant differences of hemorrhage and thrombosis between two groups (Table 3). Although there were no significant differences of weaning rates from ECMO, ventilator duration, ICU stay, and hospital stay, the ICU mortality was significantly higher in the control group (38.5% vs. 21.9%, p = 0.012). Also, hospital mortality was significantly higher in the control group (41.7% vs. 24%, p = 0.009) and sepsis-related death was significantly higher in the control group (17.7% vs. 6.3%, p = 0.014).

In the univariate regression analysis, long-term ECMO support of ≥ 2 weeks (odds ratio (OR) 3.88, 95% confidence interval (CI) 1.15–13.09, p = 0.029), 2% CHG/IPA intervention (OR 0.18, 95% CI 0.04–0.86, p = 0.031), RRT during ECMO (OR 4.20, 95% CI 1.10–16.04, p = 0.036), and oxygenator exchange (OR 4.00, 95% CI 1.10–14.49, p = 0.035) were significantly associated with BSI (Table 4). In the multivariate analysis, 2% CHG/IPA intervention (OR 0.16, 95% CI 0.03–0.76, p = 0.021) and oxygenator exchange (OR 5.13, 95% CI 1.32–20.00, p = 0.019) were significantly associated with BSI. The Kaplan-Meier curves showed that the intervention was significantly related to lower BSI rate (χ² = 5.70, p = 0.017, Fig. 2).

In the univariate cox regression analysis, factors, which were significantly associated with overall mortality, included age ≥ 65 years (OR 1.74, 95% CI 1.05–2.88, p = 0.031), APACHE II (OR 1.04, 95% CI 1.01–1.08, p = 0.027), RRT during ECMO (OR 2.95, 95% CI 1.73–5.03, p < 0.001), and 2% CHG/IPA intervention (OR 0.49, 95% CI 0.29–0.82, p = 0.006). In the multivariate cox regression analysis, age ≥ 65 years (OR 1.98, 95% CI 1.20–3.28, p = 0.008), RRT during ECMO (OR 2.89, 95% CI 1.69–4.96, p < 0.001), and 2% CHG/IPA intervention (OR 0.52, 95% CI 0.31–0.87, p = 0.013) were significantly associated with overall mortality.

The causative microorganisms cultured in 11 patients with ECMO catheter colonization are shown in Additional file 2. The main microorganisms identified in the control group were gram-positive bacteria such as S. epidermidis (3.1%), Enterococcus faecalis (1.0%), and other staphylococcus species (1.0%). Acinetobacter baumannii (3.1%) and Candida species (2.1%) were also found. One patient had both Enterococcus faecalis and Candida tropicalis infection. However, only Candida species, i.e., Candida glabrata (1.0%) and Candida albicans (1.0%), were found in the intervention group.