British Journal of Anaesthesia 98 (1): 131–5 (2007)

doi:10.1093/bja/ael288 Advance Access publication October 30, 2006

REGIONAL ANAESTHESIA
Efficacy of spray disinfection with a 2-propanol and
benzalkonium chloride containing solution before epidural
catheter insertion—a prospective, randomized, clinical trial
G. Debreceni1 *, R. Meggyesi2 and G. Mestyán3
1
Department of Anaesthesiology and Intensive Care, University of Pécs, Hungary.
2
Anaesthesiology Residency Program, Ministry of Health, Hungary. 3Department of Microbiology,
University of Pécs, Medical Center, Hungary
*Corresponding author: Department of Anaesthetics, Monklands Hospital, Lanarkshire NHS Trust,
Monkscourt Avenue, Airdrie ML6 0JS, UK. E-mail: [email protected]
Background. Skin disinfection before neuroaxial blockade procedures is usually obtained with
sterile swabs impregnated in disinfectant. Spray disinfection is also an option which is frequently
used in minor invasive procedures. The purpose of our study was to compare the efficacy of
conventional swab disinfection with spray disinfection prior to epidural catheterization.
Methods. Seventy patients who requested epidural analgesia were randomly selected. The first
group (n=35) received disinfection with swabs (SW) containing 2-propanol and benzalkonium
chloride. The other 35 patients received spray (SP) disinfection with the same solution. Three
microbiological cultures were obtained: one culture prior to skin disinfection, a second
immediately after disinfection and a third from the tip of the epidural catheter upon removal.
Results. One patient in the SW group had a positive skin culture immediately after the
disinfection with a very low number of colony forming units. The other skin culture specimens
were all sterile in both groups. The colonization rate of catheters was not statistically different
between the groups at removal.
Conclusion. In this study, spray disinfection was equally efficacious compared with the
conventional skin disinfectant technique. Our results support the routine use of this simple
and cheap alternative method of skin disinfection before epidural anaesthesia.
Br J Anaesth 2007; 98: 131–5
Keywords: anaesthetic techniques, epidural; skin, disinfection, spray
Accepted for publication: September 16, 2006

Since the development of the rules of asepsis in the late 19th with sterile swabs soaked in disinfectants. Another conven-
century, skin disinfection is obligatory before all interven- tional disinfection technique is spray disinfection. There is
tions which penetrate the intact skin. The increasing use of very little evidence about the superiority of either method at
epidural anaesthesia highlights the importance of catheter present.4–6 The aim of this prospective, randomized study
placement under aseptic circumstances. Although serious was to compare the efficacy of skin disinfection prior to
infections like bacterial meningitis and epidural abscess epidural catheter insertion using either swabs or spray
are rare following epidural techniques, these complications disinfection.
do occur and can have devastating consequences. Methods
for skin disinfection vary1 and in most of the textbooks there
are dogmatic regulations without evidence based recom- Methods
mendations. Recently published reviews2 and guidelines3 This study was approved by the Regional Research Ethics
rely heavily on data extrapolated from studies on central Committee and the Infection Control Service of the Univer-
venous catheterization. The most common practice for sity of Pécs, Medical and Health Sciences Centre, Pécs,
epidural catheter insertion is disinfecting the puncture site Hungary. Written informed consent was obtained from all

The Board of Management and Trustees of the British Journal of Anaesthesia 2006. All rights reserved. For Permissions, please e-mail: [email protected]
 Debreceni et al.

Table 1 Distribution of operations between experimental groups: SW, catheter insertion site was documented at the time of
disinfection using swab group; SP, disinfection using spray group. Excluded removal. Four signs of irritation: redness, swelling, leakage
patients’ data were not in the analysis (see text). Others: 1 lysis of abdominal
adhesions, 1 retroperitoneal tumour, 1 partial gastric resection and local pain were scored as present (=1) or not (=0). An
insertion site without irritation was scored as 0, maximal
Types of operations SW SP
irritation scored 4 points. During the observation period and
Lung resection 9 6 at catheter removal the syndrome of possible epidural
Gastrectomy 2 1 infection (two major symptoms: back pain and neurological
Liver resection 6 7
Oesophagus resection 1 2
deterioration and the two minor symptoms: elevated white
Colon resection 4 6 blood cell count and fever) was noted. We did not note fever
Rectum resection 3 4 or increased white blood cell count per se if one of the major
Pancreas resection 0 2
symptoms such as back pain or neurological signs were
Choledocho-jejunostomy 2 0
Femoro-popliteal bypass 2 1 missing.
Hip arthroplasty 1 2 In the SP (spray) group (n=35) skin was disinfected with
Others 2 1
the same disinfectant as mentioned above using a multi-dose
Total 32 32
pressurized plastic container from which the disinfectant
was sprayed on an approximately 20 cm area in diameter
surrounding the puncture site three times from about 20 cm
patients. Seventy patients with American Society of Anes- and waited 3 min as recommended by the manufacturer.
thesiologists physical status II–III who underwent various Covering, local anaesthesia, insertion technique, fixing
operations (Table 1) were enrolled and randomly assigned and the local anaesthetic agents were identical as in the
via an envelope system to undergo skin decontamination SW group.
prior to epidural catheter insertion. Patients were excluded Three microbiological cultures were obtained from each
if they were under 18 yr old, had fever, diabetes, received subject: For skin cultures Envirocheck Rodac Blister for
antibiotics before the operation, were receiving steroid ther- Total Colony Count (Merck, USA) medium was used. The
apy, were immunocompromised, had HIV or pre-existing agar has a convex surface in it easing the obtaining of skin
skin infection at the planned puncture site and/or had iodine culture. The square lines on its transparent case help with the
allergy. All epidural catheter insertions were carried out counting of the colony forming units of bacteria. Using
before the initiation of general anaesthesia. sterile gloves and holding the sterile case of the medium,
The disinfectant used consisted of 2-propanol 63 g, the convex surface of the agar was pushed onto the skin for
benzalkonium chloride 0.025 g, water and dye in 100 g 3 s. The first sample was obtained from the area adjacent to
solution (Cutasept G , Bode Chemie, Hamburg, Germany). the planned puncture site just prior to skin disinfection to
In the SW (swab) group (n=35), the conventional technique determine baseline skin flora, the second was obtained from
was used for skin preparation. After hygienic hand washing the same area immediately after the 3 min drying time fol-
wearing cap, face mask and sterile gloves an approximately lowing disinfection of the skin to determine initial efficacy
20 cm area in diameter around the planned puncture site was of the disinfectant. The plates were taken to Microbiology
disinfected using sterile forceps and sterile swabs soaked in within an hour. Colonies were enumerated and identified
disinfectant. We repeated this procedure three times and by standard methods at genus level. The third sample was
then waited for 3 min as recommended by the manufacturer. the distal tip of the epidural catheter itself. To reduce the risk
After skin disinfection, the site was covered with a sterile of tip contamination by skin during removal, the area sur-
sheet. Lidocaine, 80–100 mg was used to locally anaes- rounding the insertion site of the catheter was disinfected by
thetize the puncture site. Following this, the epidural cathe- aqueous povidone-iodine solution just before removal. Once
ter was inserted with the standard loss of resistance the povidone-iodine had dried, the catheters were removed
technique at the lumbar or thoracic level. A sterile adhesive using sterile gloves and a 2–3 cm distal part of the catheter
semi-permeable polyurethane dressing (Tegaderm 3M, was cut and placed in a sterile container using sterile
USA) was placed over the area surrounding the epidural scissors. All these specimens were similarly taken to
insertion site. A bacterial filter was attached to the epidural Microbiology within an hour. The catheter tips were rolled
catheter. During surgery 0.2 ml kg 1 boluses of ropivacaine four times on the surface of blood agar plates and these
0.2% were given via the epidural catheter. After the opera- plates were incubated at 37 C for 48 h. Colonies were
tion patients received bupivacaine 0.125% solution continu- enumerated and identified by standard methods at genus
ously for pain management using syringe pumps via the level.
same filter. In each case the patient’s age, weight, height, The anaesthesiologist obtaining the skin samples could
sex, epidural site (thoracic, lumbar) and the time needed to not be blinded to the method used because the shape of the
insert the epidural catheter, hours of keeping the catheter in area after using SW or SP, were noticeably different.
place, type of single shot prophylactic antibiotics, type of In each case, the anaesthesiologist who obtained the
operation, and the signs of local irritation during catheter epidural specimens and microbiologist handling the culture
removal were recorded. Local irritation around the epidural samples were blinded to the disinfection method used.

132
 Spray disinfection before epidural anaesthesia

Statistics molytic Streptococci. The samples collected 3 min after skin

As we did not have data on colonization rate in our disinfection and after 48 h of incubation were sterile except
institution, sample sizes were based on previously published for one specimen in the SW group with very low CFU 1 of
results.7 According to previous data the smallest difference coagulase negative Staphylococcus and 1 of Bacillus
between the two groups in the epidural catheter colonization species.
rate which we did not want to be overlooked was around Six tips in the SW group and 2 tips in the SP group of the
30%. For the study to have 80% power at a significance level epidural catheters were colonized, this difference was not
of 0.05, the estimated sample size was 30–35 patients per statistically significant (x2-test, P=0.257). No catheters
group. Data were presented as mean and SD. For statistical showed multiple colonization. The colonizing bacteria
analysis we used Student’s t-test for parametric data and specimens were similar to normal skin flora (Table 3).
x2-test for non parametric data. All analyses were performed None of the patients had any signs of epidural infection
with the Statistical Package for the Social Sciences (SPSS at the time of catheter removal.
version 11.5 for Windows, Chicago, IL, 2002).

Results Discussion
Out of the seventy patients enrolled, three from each group Effective skin disinfection is thought to be an adequate
had to be excluded due to accidental removal of the epidural measure for prevention of infection as a consequence of
catheter by ward nurses. The SW and SP groups were simi- procedures that disrupt the skin barrier. Various kinds of
lar with respect to height, weight, male–female ratio and skin disinfection have been used and their efficacy has been
age. There were no significant differences in thoracic versus tested for various medical procedures.4–13 The most
lumbar epidural proportions, time of successful catheter commonly used solutions are povidone–iodine, polyalco-
placement, time of holding the catheter, rate of irritation holic preparations and chlorhexidine solutions or their
at the puncture site at the time of removal (Table 2). All the mixtures.4–10 For interventions such as central venous
patients received third generation cephalosporins as a single catheterization, epidural and spinal puncture most text-
shot for antibiotic prophylaxis with metronidazole added for books, reviews and guidelines suggest disinfecting the
abdominal interventions. No patients received long term skin with swabs three or four times. This method is not
antibiotic treatment. evidence based regarding epidural catheterization. At the
In both groups cultures from normal, non-disinfected same time for minor procedures (peripheral venous cannu-
skin showed a normal bacterial skin flora. The most lation, minor skin operations) spray disinfection with
common isolates were on each patient’s skin, coagulase multiple-use pressurized containers can be used. We could
negative Staphylococci. Other species isolated included not find any evidence in the literature suggesting the supe-
Corynebacteria, Micrococci, Bacillus species, and a hae- riority of swab over spray disinfection. Although, the
efficacy of different solutions have been reported in many
articles,4 9 12 only a few studies have investigated different
Table 2 Demographics and clinical characteristics of the two groups: SW, methods with the same solution.4 5
disinfection with swab group; SP, disinfection with spray group. Three patients
in each group had inadvertent catheter removal before bacterial sampling and
their data were excluded from the analysis. Rate of irritation: see text. For
statistical analysis Student’s t-test and *x2 test were used. P<0.05 considered Table 3 Number of positive microbiological cultures at the tips of the epidural
significant catheters after removal. For detailed microbiological methods and statistics see
text. SW, disinfection using swab group, number of patients=32; SP, disinfection
SW n=32 SP n=32 Statistics using spray group, number of patients=32; CFU, count of colony formation
units; Coag. neg., coagulase negative staphylococci
Age (yr) 58.4 (10) 58.0 (12) NS
mean (SD) Coag. neg. Corynebactericae
Sex (male/female ratio) 15/17 16/16 NS* staphylococci
Site of the epidural 21/11 21/11 NS*
(thoracic/lumbar ratio) Number of positive 5 1
Weight of the patients (kg) 80.28 (14.4) 74.1 (14.2) NS cultures of the epidural
mean (SD) catheter tips in SW group
Height of the patients (cm) 169.4 (10.2) 166.5 (9.3) NS Number of colony Patient 1:2 Patient 1:118
mean (SD) formation units in SW Patient 2:15
Time to succesful 13:31 (5:46) 14:13 (8:04) NS group in different patients Patient 3:31
insertion (min) Patient 4:11
mean (SD) Patient 5:7
Epidural catheters 87.9 (31) 80.7 (27) NS Number of positive 2 –
were kept in place (h) cultures of the epidural
mean (SD) catheter tips in SP group
Rate of irritation at time 1.16 (1.0) 1.03 (1.0) NS Number of colony formation Patient 1:7 –
of removal units in SP group in Patient 2:2
mean (SD) different patients

133
 Debreceni et al.

We have chosen an alcohol based disinfectant solution Epidural abscess is a serious but fortunately rare com-
since many studies revealed that alcohol based disinfectants plication of epidural catheterization.10 The suggested
are more powerful against microorganisms than aqueous mechanism of development of the epidural abscess is
povidone–iodine. In addition, alcohol based solutions sufficiently controversial. Invasion by skin bacteria through
have lower rate of allergy and fast drying.14 Birnbach a needle track, contaminated syringes, contaminated local
and colleagues13 found that 40% of the insides of multiple- anaesthetics and haematogenous spread from a distant site
use povidone–iodine bottles were contaminated. In contrast, of infection were all accused but the exact mechanism is not
Robins and colleagues4 demonstrated in their study that understood.10 In our study we examined the colonization
multiple-use chlorhexidine in alcohol solution’s bottles rate of the catheters although colonization according to
were not contaminated. There are no available data on Simpson and colleagues8 is not a good predictor of epidural
contamination of 2-propanol and benzalkonium chloride space infection and can be the result of contamination of the
containing multidose containers therefore we chose this catheter tip upon removal.8 However, it is hard to find a
solution for disinfection. However, we did not take micro- perfect method to examine the connection between skin
biological samples from the empty vials. Robins and preparation and epidural infection if it exists at all. So we
colleagues4 in their study compared the effect of chlorhexi- chose the commonly used colonization rate as an endpoint
dine in alcohol spray from a multidose vial with an aqueous knowing that its casuality with the infection is questionable.
single dose chlorhexidine solution from a sachet, on skin In conclusion, under these circumstances spray disinfec-
preparation before obstetric epidural anaesthesia. They tion is at least effective as swab disinfection. Further studies
found that spraying technique was as effective as scrubbing are needed to establish any superiority.
the skin and was significantly cheaper. In their study they
did not examine the epidural catheters at removal and the
average duration of catheterization was about only 2.2 h
Acknowledgements
The authors would like to thank: Lajos Bogár, MD, PhD; Zsolt Molnár, MD,
while in our study this period was more than 80 h and PhD DEAA; Subhamay Ghosh, MD, Department of Anaesthesiology and
our patient pool was based on abdominal and thoracic Intensive Care, University of Pécs, Hungary; and Dr Rory Mackenzie,
surgery. Department of Anaesthetics, Monklands Hospital, Lanarkshire NHS
Trust, Scotland for their reviewing of the manuscript and to the staff of
Birnbach and colleagues7 and Sato and colleagues10 in the Department of Anaesthesiology, University of Pécs, Hungary for their
their studies found that skin disinfection on the back was not help and generous support. Soli Deo Gloria.
capable of decreasing the number of bacteria below the level
of detection. In our study there was only one patient out of
64 with detectable bacterial growth after skin disinfection.
References
In this patient a very low number of colony forming units
1 Visser L. Epidural anaesthesia. Update in Anaesthesia 2001; 11: 1–4
were found and this patient’s epidural catheter tip was sterile 2 Grewal S, Hocking G, Wildsmith JAW. Epidural abscesses.
at the time of removal. In our study the incubation period Br J Anaesth 2006; 96: 292–302
was 48 h while it was 72 h in the study of Birnbach and 3 Association of Anaesthetists of Great Britain and Ireland. Infect
colleagues.7 We only touched the skin with the surface of Control Anaesthesia, 23, 32, 33. Published by The Association of
the medium, Birnbach and colleagues5 used swabs for skin Anaesthetists of Great Britain and Ireland, November 2002.
specimens while Sato and colleagues10 excised the skin to 4 Robins K, Wilson R, Watkins EJ, Columb MO, Lyons G.
Chlorhexidine spray versus single sachets for skin preparation
obtain bacterial specimens from the deeper layers of the
before regional nerve blockade for elective caesarean section:
skin. This may explain at least in part the difference between an effectiveness, time and cost study. Int J Obstet Anesth 2005;
our findings. 14: 189–92
We also assessed the signs of irritation at the insertion site 5 Moen MD, Noone MB, Kirson I. Povidone–iodine spray
before removal of the catheter and rated it. The scores were technique versus traditional srub-paint technique for preopera-
very low (1 in average out of maximum 4). We did not tive abdominal wall preparation. Am J Obstet Gynecol 2002; 187:
perform skin sampling before the epidural catheter removal 1434–6
6 Gundermann KO, Christiansen B, Holler C. New methods for
because we presumed that after 80 h of catheter holding we
determining pre-operative and postoperative skin disinfection.
might have found normal skin flora. J Hosp Infect. 1985; 6 (Suppl. A): 51–7
In our study there was no difference between groups 7 Birnbach DJ, Meadows W, Stein DJ, Murray O, Thys DM,
with regard to demographics, the time of catheter insertion Sordillo EM. Comparison of povidone iodine and DuraPrep an
procedure and the time of using the catheters. Skin flora iodophor-in-isopropyl alcohol solution, for skin disinfection prior
was similar in each group. The microbiological results of to epidural catheter insertion in parturients. Anesthesiology 2003;
tips of the catheters showed that there was a tendency of 98: 164–9
8 Simpson RS, Macintyre PE, Shaw D, Norton A, McCann JR,
higher colonization rate in the SW group but this dif-
Tham EJ. Epidural catheter tip cultures: results of a 4-year
ference was not statistically significant. It should be audit and implications for clinical practice. Reg Anesth Pain Med
noted that among the persons with colonized catheters 2000; 25: 360–7
none of them showed any sign of epidural catheter related 9 Parienti JJ, du Cheyron D, Ramakers M, et al. Alcoholic
infection. povidone-iodine to prevent central venous catheter colonization:

134
 Spray disinfection before epidural anaesthesia

a randomized unit-crossover study. Crit Care Med 2004; 32: versus chlorhexidine ethanol. Dermatology 2002; 204 (Suppl. 1):
887–8 42–6
10 Sato S, Sakuragi T, Dan K. Human skin flora as a potential source 13 Birnbach DJ, Stein DJ, Murray O, Thys DM, Sordillo EM.
of epidural abscess. Anesthesiology 1996; 85: 1276–82 Povidone iodine and skin disinfection before initia-
11 Yentur EA, Luleci N, Topcu I, Degerli K, Surucuoglu S. Is skin tion of epidural anesthesia. Anesthesiology 1998; 88:
disinfection with 10 % povidone iodine sufficient to prevent epidu- 668–72
ral needle and catheter contamination? Reg Anesth Pain Med 2003; 14 Maki DG, Ringer M, Alvarado CJ. Prospective randomised trial of
28: 389–93 povidone-iodine, alcohol and chlorhexidine for prevention of
12 Kasuda H, Fukuda H, Togashi H, et al. Skin disinfection before infection associated with central venous and arterial catheters.
epidural catheterization: comparative study of povidone-iodine Lancet 1991; 338: 339–43

135