Efficacy of two antiseptic regimens on skin colonization of insertion sites for two different catheter types: a randomized, clinical trial

A randomized, prospective, controlled trial was conducted to compare the skin recolonization at central venous catheter (CVC) and epidural catheter (EC) insertion sites over 48 h in patients receiving a skin disinfection with either a BAC- or an OCT-containing alcohol-based disinfectant.

From July to December 2014, 216 patients in an academic teaching hospital in Cologne, Germany, were enrolled in the trial. The majority of patients received a catheter prior to surgery, all others (<2 %) were patients in the Intensive Care Unit (ICU).

The Ethical Review Board of the North Rhine Medical Association approved the trial in July 2014 (application-no.: 2014222).

Adult patients likely to receive a CVC or EC or both according to pre-operative standards were asked for their informed consent. Pre-trial exclusion applied for minors, emergency patients and patients with nosocomial infections. Patients whose swabs could not be obtained according to protocol were excluded. This applied for catheters being removed during the follow-up period as well as for patients whose sterile dressing did not cover the insertion site sufficiently after 48 h. Applying house standards according to National Hospital Hygiene Standards [12], all catheter insertions were carried out by experienced anesthesiologists, following a strict protocol including the use of surgical masks and caps, sterile gowns, gloves and catheter kits. The sterile dressing was applied right after the insertion, using a transparent foil to facilitate inspection of the insertion site. According to the study protocol, nurses or doctors changed the sterile cover every 48 h after disinfecting the catheter and its surrounding skin with a propranolol spray and a sterile pad.

The disinfection regimen was changed at weekly intervals, thereby randomly assigning patients to the trial or control group, respectively. The patients, the microbiological laboratory and the statistician were blinded to the assignment. The medical staff could not be blinded due to the colorant solely in the BAC agent and the use of transparent dressings.

After having given their informed consent, patients received skin disinfection with either one of the following commercially available and legally registered antiseptics:

All patients enrolled were treated and followed-up between July and December 2014 in the Department of Anaesthesiology at St. Vinzenz-Hospital, Cologne, Germany. Patients received catheters prior to surgery in the hospital’s departments of abdominal, thoracic, gynecological and traumatological surgery or as part of the intensive care or pain treatment in the ICU or the department of pain therapy respectively. The environment for catheter insertions was either the operating theater or the ICU.

Patients who were scheduled for catheters repeatedly during the same or various hospitalizations were allowed to be enrolled twice (respectively, three times) if at least 4 weeks had passed since the previous catheterization.

All adult patients scheduled to receive a CVC or EC who gave their written consent were included.

To reduce the loss of patients due to protocol deviations, the fixation especially for CVC was improved by applying strain-relief to the connected lines. In addition, all sterile covers were additionally fixated with adhesive tape on the rim. For EC patients, the sterile dressings’ size was increased and its rims supported by adhesive tape.

All swabs were obtained by trained members of the trial team under standardized conditions using a sterile drape with a circular hole with a 7 cm diameter. The thereby marked skin surrounding the catheter insertion site was swiped with a sterile flocked swab (eSwab Liquid Amies, Copan, Brescia, Italy) with a movement ten times vertically and ten times horizontally. The tip was then immersed in the swabs' prefilled 1.0 mL sterile 0.01 M phosphate buffered saline (PBS) for transport to the microbiological laboratory.

To determine the bacterial burden, samples were cultured quantitatively (native, 1:10 and 1:100 dilution) on trypticase-soy-agar plates with sheep-blood (bioMérieux, Nürtingen, Germany). After vortex mixture, cultures were incubated for 36–48 h at 35 ± 1 °C. All tests were run in duplicate. The plates were inspected by a technician and the supervising microbiologist at 18–24 and 36–48 h. Results were expressed as CFU per swab.

Main outcome parameter was the difference in CFUs/swab (colony forming units per swab) of swab 3 (i.e. 48 h after insertion of CVC or EC) in comparison to swab 2 (i.e. immediately after disinfection and insertion). Secondary objective was the comparison of swab 3 and swab 1 (i.e. before disinfection and insertion). Patients were divided into cohorts primarily by skin disinfection regimen and secondly by type of catheter, so that four groups could be compared (CVC with OCT, CVC with BAC, EC with OCT, EC with BAC).

Patients’ comorbidity factors such as smoking, diabetes and permanent corticoid therapy were monitored. None of the enrolled patients suffered from chronic skin conditions. The amount of white blood cells and C-reactive protein were also taken into account.

Side effects, such as skin irritation (burning, itching, swelling, redness) were also closely monitored.

Statistical analysis based on the CFUs counted by the laboratory personnel was statistically conducted by the University Hospital of Cologne’s Institute for Medical Statistics, Informatics and Epidemiology (Institut für medizinsche Statistik, Informatik und Epidemiologie, IMSIE).

The hypothesis was that octenidine as an additive to alcoholic skin disinfectants has a longer lasting antimicribiological effect on catheter-surrounding skin than benzalkonium chloride.

The analysis was conducted using the Statistical Package for the Social Sciences, version 22 (SPSS, IBM, Armonk, NY, USA). CFUs counted were logarithmized due to their great span. The disinfection regimen was compared at two time points: immediately after catheter insertion (swab 2) and 48 h after catheter insertion (swab 3). Swab 1 prior to skin disinfection was conducted as a baseline, as patients with different backgrounds were expected to start off with extremely different CFU amounts on their untreated skin. The hypothesis was tested by a Wilcoxon test with a two-sided alpha = 5 %.

Patients in the trial and control group, even when being catheterized for different reasons, showed similar characteristics regarding comorbidity, sex, age and physical constitution.

41 patients received both a CVC and an EC in one session, eight entered the trial twice, and one patient was enrolled three times in total. At least four weeks had to elapse after first catheterization to re-enter again.

110 patients qualified for skin colonization analysis. There was no difference in the immediate effect (i.e. representing swab 2) between the trial and control group. The difference in CFUs from swab 2 to swab 3 showed a significantly lower value of recolonization in the trial group. In the subgroups, additionally divided by type of catheter, the values for CVCs were generally higher than for ECs in swabs 2 and 3 respectively, the level of recolonization again being much lower in the patients treated with OCT. Figure 1 shows the logarithmized amount of CFUs in all three swabs. The direct comparison of trial and control group shows the highly significant difference in recolonization from swab 2 to swab 3, as could be expected. The difference of mean values in patients treated with BAC was 1.97 (95 % CI: 1.45; 2.50), whereas the trial group treated with OCT only showed a delta of 0.72 (95 % CI: 0.42; 1.02; p = 0:0005).

Side effects such as redness, swelling or pathological serum parameters were mild, rare and occurred in all groups. Few patients in all groups developed slight redness (n = 7, trial: n = 3; control: n = 4) or swelling (n = 3; trial: n = 2; control: n = 1) around the catheter insertion site (see Fig. 2; Table 1).

None of the patients developed signs of CRBSI or any other systemic side effects, which was expected due to the short period of follow-up.