CLINICAL NUTRITION (1985) 4: 24 3 1

Skin Bacteria and Op Site@ Chest Dressings

B. P. Waxman’, C. S. F. Easmort and H. A. F. Dudley’ 1. University of Melbourne, Department of Surgery 2. Department of Medical Microbiology, St. Mary’s Hospital Medical School. London 3. Academic Surgical Unit, St. Mary’s Hospital, London (Reprint requests to C.S.F.E.)

ABSTRACT Growth of skin bacteria on the infraclavicular region was studied in two series of male volunteers. In the first, Op Site@, a polyurethane adhesive film dressing, was compared with an occlusive polyvinyl chloride (PVC) dressing, on povidone iodine (WI) prepared skin in 10 volunteers. Bacteria were sampled, using perspex cylinders and buffered Triton X-100 detergent, at 2, 4, 7 and 14 days, cultured aerobically and anaerobically, and colonies counted at 24 and 48 h respectively. Colony counts under Op Site@’ were less than for undressed (control) skin and under PVC, at all days sampled, the difference being statistically significant at 2 and 14 days for controls and 4 days for PVC. In the second, four regimens of skin preparation, with Op Site@ dressings were compared in 12 volunteers, skin being sampled 4, 7 and 14 days. Chlorhexidine (CHD) and WI were compared with and without defatting. Defatting significantly reduced colony counts at 4 and 7 days, whereas no differences were demonstrated between CHD and WI. The combination of defatting and CHD resulted in colony counts consistently less than 10’ organisms at 4 and 7 days. Op Site@ does not potentiate the growth of skin bacteria and is preferable to an occlusive PVC adhesive dressing. Op Site@ may be left intact on the chest for up to 7 days, colony counts remaining within acceptable limits.

INTRODUCTION

Catheter-related sepsis is a common complication of total

parenteral nutrition (TPN) [1,2]. That skin commensals form the largest group of bacteria cultured from catheter tips and blood of patients on TPN is now well documented

[3,4]. Two factors that may influence the proliferation of skin bacteria around central venous catheters are the type of catheter dressing and the method of skin preparation.

Op Site@ (Smith & Nephew Limited) membrane dressing

(Op Site@), a vapour permeable, adhesive, polyurethane film, has become a popular dressing for TPN catheters because it is sterile, non-occlusive, transparent and easily

applied and changed [2,5,6]. Other types of adhesive dressings are either non-transparent or occlusive and

occlusion is known to potentiate the growth of skin bacteria [7]. Results from recent studies comparing Op Site@ either directly [8] or indirectly [9] with other forms of dressings

and nursing care, indicate there may be an increase in

episodes of catheter-related sepsis with Op Site@. However, well controlled data are not available on the effects of Op Site@ on skin bacteria.

The optimal regimen of skin preparation, for the insertion of TPN catheters and subsequent dressing and change of dressings, has neither been well defined nor adequately studied. Kaul and Jennett [lo] recommended alcoholic chlorhexidine for routine surgical skin disinfection, whereas Jarrand et al. [11,12] found povidone iodine effective in almost eliminating skin bacterial growth under TPN catheter dressings. Moreover, defatting the skin, before applying an antiseptic, is recommended [ 13,141

to reduce the growth of the lipid dependent anaerobes (diphtheroids). However, again controlled data are not available comparing the different antiseptics and the

virtues of defatting, applicable to TPN catheter dressings.

The aims of this study were therefore twofold:

(1) To define the effects of Op Site9 on skin bacteria

compared with an occlusive adhesive dressing. (2) To

determine the optimal skin preparation to use with Op Site@ by comparing chlorhexidine and povidone iodine

with and without defatting the skin, and by inference

determining the appropriate timing for dressing changes.

The study was conducted in two parts and in both male volunteers without indwelling central-venous catheters

were used, to provide controlled base line data without the

added variables of: illness; malnutrition; surgical sepsis and injury; and the presence of a catheter. The skin on the infraclavicular region was used throughout the study.

METHODS

Op Site@ Compared with occlusive (PVC) dressings

Skin preparation and dressings. The infraclavicular region of the chest of 10 male volunteers was shaved immediately before the skin was cleansed with 10% povidone iodine (Vidine; Stuart Pharmaceuticals Limited) using sterilized swabs (Propax; Smith and Nephew Limited) and allowed to dry, then gamma-irradiated 10 cm x 2 cm strips of either Op Site@ or a polyvinyl chloride (PVC) dressing with identical adhesive, were applied to the prepared area of skin

29

30 SKIN BACTERIA AND OP SITE@ CHEST DRESSINGS

in 4 pairs, 2 pairs on either side of the chest, in vertical

rows. Skin preparation and application of dressing was

performed under sterile conditions, the operator wearing

gown and gloves and the site protected with sterile drapes. Volunteers were encouraged to carry on normal daily activities, but avoiding sport and washing the chest.

Random pairs of strips were removed at 2, 4, 7 and 14 days,

using a Latin square method, and the skin underneath the dressings sampled.

Bacteriological analyses

Sampling. The method was developed from that

described by Williamson and Kligman [ 151. Perspex

cylinders, 3 cm long and with an internal area of 1 cm

square, were washed in 70% ethanol, dried and applied to

the skin. One millilitre of 0.1% Triton X-100 detergent (p- tert-octyl phenoxy-polyethoxyethanol) in 0.075M

phosphate buffer at pH 7.9 was pipetted into the cylinder,

the skin and detergent rubbed with a sterile blunted teflon

policeman for 2 min. This was repeated and the two 1 ml aliquots pooled in sterile universal glass containers. Duplicate samples were taken from under dressings and

adjacent undressed skin (within 2 cm of the dressing), the

latter being used as control. No skin preparation was performed at the time of skin sampling. A preliminary

study on growth of skin bacteria from skin immediately

after skin preparation with povidone iodine showed no

growth, therefore data at time zero was not collected.

Culture and bacterial identification

Tenfold dilutions were then performed with 0.05% Triton X-100 in 0.037M phosphate buffer at pH 7.9 and 0.1 ml

aliquots spread for surface viable counts in aerobic or anaerobic culture. The aerobic medium was Columbia

blood agar (Lab M) plates being incubated at 37°C and

read at 24 h. The anaerobic medium was Clostridial agar (Oxoid) plates being incubated in an anaerobic jar at 37’C and read at 48 h. Colonies were identified and counted by

one of us (CSFE), the duplicates averaged and expressed as logarithm to the base 10 bacteria counts per square centimetre. Colonies were then identified by Gram stain,

Gram positive cocci tested for coagulase and catalase and

Gram positive bacilli for catalase.

Statistical Methods

The average colony counts for pairs of dressings and controls for each volunteer were compared using the Wilcoxon two-tailed matched pairs signed ranks test. Confidence limits were set at the 95% level.

Skin preparation for Op Site? Chlorhexidine and povidone

iodine compared with and without defatting

Skin Preparation and dressings. The infraclavicular region of the chest of 12 male volunteers, different from the

previous study, was shaved and divided by marking into three sagittal segments. Each segment was divided into quadrants, making adjacent vertical and horizontal pairs of quadrants (Fig. 1). The skin of vertically adjacent pairs of

quadrants was either defatted with Arklone (trichloro- trifluro-ethane) or not defatted. Then, the skin of horizontally adjacent pairs of quadrants were cleansed with

Not &fatted defatted

24 Chlorhexidine

Fi. 1 One of three sagittal segments of chest skin, showing division into quadrants, and allocation of vertical quadrants to defatting or not and horizontal quadrants to chlorhexidine or povidone iodine. Vertically applied strips of Op Site@ are also shown.

either 10% alcoholic povidone iodine (Vidine@, Stuart

Pharmaceuticals Limited) or 0.5% Chlorhexidine in

industrial methylated spirits, and allowed to dry for 2 min. All quadrants were then washed with 70% ethanol (to

obtain universal adhesion of Op Site@), allowed to dry and

sterile 5 cm X 2 cm strips of Op Site@ applied one to each

quadrant (Fig. 1). All agents were applied using a sterile technique with sterilized swabs (Propax, Smith & Nephe J Limited). Allocation of segments and quadrants was

pseudo-randomised using a Latin square method.

The four Op Site@ dressings were removed from each segment at 4, 7 and 14 days and the skin beneath each dressing sampled with adjacent skin as control.

Analyses and statistical methods were as outlined above.

RESULTS

The data are presented in Figures 2-5 each data point representing the average of duplicate samples and the horizontal lines the median for each set of data. Lines between corresponding pairs of data points have been omitted to improve clarity.

Op Site@ Compared with Occlusive (PVC) dressings

Aerobic staphylococci epidermidis and the anaerobic diphtheroids, being the major resident bacterial flora of the skin, made up the majority of the cultures, but no pathogens were cultured from either under Op Site@, under PVC or controls.

x AA

5

Log 4

Bacteria

Counts 3

cm -2 2

A . .

- .

A - -n-

a A . .

1

0 . *

** I *c DAY 2 DA*Y 4 DAY 7 DAY 14

(n-10) (n-lo) (n-lo) (n-7)

Fig. 2 Colony counts of bacteria for Op Site@ compared with occlusive PVC dressings: aerobic culture. (Raw data and medians.)

CHEST DRESSINGS: Aerobic culture op Ske ._

CI.INICAI. NL’TRITION 3 1

Wllcoxon

p -CO.06 *

p <0.02 xx

Occlusive A

Control .

For aerobic organisms (Fig. 2) growth under Op Site@ was in general less than for PVC and controls for all days,

being statistically significant when compared with PVC at

day 4 (p < 0.05) and 14 (p < 0.02) and with controls at day 2

(p<O.Ol). For anaerobic organisms (Fig. 3) growth under Op Site@

was again less than for PVC and controls for all days, being

statistically significant when compared with PVC at day 4

(p<O.Ol) and with control at day 2. Growth under PVC was not different from controls

neither for aerobic nor anaerobic culture, and by 14 days,

though not statistically significant, growth under PVC had

increased. The small numbers of dressings available at 14 days were because of loosening of dressings in some volunteers, making the sampling of bacteria invalid.

Colony counts under Op Site@ were in the majority of

volunteers less than 100 per square centimetre at day 2 and

day 4, being greater than that at 7 and 14 days.

Skin preparation for Op Site 7 Chlorhexidine and povidone

iodine compared with and without defatting

As with the first part of the study no pathogens were

CHEST DRESSINGS: Anaerobic culture 013 Site

Occlusive A

Control .

Wilcoxon

p <0.02 *

p <O.Ol **

. . >5 _L

5

Log 4

Bacteria

Counts 3

cm -2 2

1

0

.

.

.

. z5

. I

* .

L A .

a fi

2

4 : t

‘?

: 8

?

DA*Y 2 DAY 7 DAY 14

(n-10) (n-lo) (n-7)

Fig. 3 Colony counfs of bacteria for Op Site” compared with occlusive PVC dressings; anaerobic culture. (Raw data and medians.)

. **

DAY 4

0-w lo)

.

: .

.

L. : ;

-t

: I

A .

4

A . .

- l

. -o-

., . B .

32 SKIN BACTERIA AND OP SITE* CHEST DRESSINGS

CHEST DRESSfNGS(Op Site): Aerobic culture (n-12)

CHD CHD+defatthg i Control . PVI 0 PVl+defatting .

6

4

Log Bacteria 3 Counts

cm -2 2

1

0

DAY 4

Fii. 4 Comparison of skin preparation regimens for Op Site@. Colt PVI = Povidone Iodine. (Raw data and medians.)

cultured from any site. When Chlorhexidine and povidine

iodine prepared skin were compared with defatted or not, no statistically significant differences were found.

Defatting the skin did confer benefit in reducing aerobic

organisms for Chlorhexidine prepared skin compared with

controls at day 4 (~~0.01) and day 7 (PC 0.02) and for povidone iodine prepared skin when compared without defatting at 7 days (p<O.O5) and when compared with

control at 14 days (p<O.Ol) (Fig. 4).

5

4

LOQ Bacteria 3 Count8

cm -2 2

1

0

Wilcoxon

P < 0.05 + P < 0.02 - P<O.Ol -

my counts of bacteria: aerobic culture. CHD = Chlorhexidine;

The benefits of defatting were more consistent for centimetre.

CHEST DRESSINGS(Op Site): Anaerobic culture (n-12)

CHD a CHDidefatthg . Control . PVI PVI+defatting t

0

Y ”

il- ma

DAY 4

8 .

9 l

0 l - 0 0 t

de

DAY 14

reducing growth of anaerobic organisms, the majority of colony counts being less than lo3 per square centimetre for

days 4 and 7, the difference being statistically significant

for Chlorhexidine prepared (p<O.O5) and for povidone

iodine at 7 (p < 0.01) and 14 days (p < 0.05) (Fig. 5). No regimen of skin preparation completely eliminated

the growth of skin bacteria under Op Site@ but a regimen

combining Chlorhexidine and defatting resulted in nearly all colony counts being less than 10’ organisms per square

WlkOXO#l

P < 0.05 *

P< O.Olw+

. r”

DAY 14

Fig. 5 Comparison of skin preparation regimens for Op site@‘. Colony counts of bacteria: anaerobic culture. CHD = Chlorhexidine. PVI = Povidone Iodine. (Raw data and medians.)

CLINICAL NVTRITIOK 33

organisms on the skin are not rapidly destroyed by the fatty

acids.

Despite the theoretical disadvantages, defatting the skin

has been demonstrated in this study to significantly reduce

the growth of skin bacteria under Op Sitea dressings.

Two recent clinical trials have shown an increase in the

episodes of catheter related sepsL for Op Site’@ dressings when compared with standard gauze and tape dressings

[8,9], however, both studies were poorly controlled. Powell et al. [8] compared two groups of patients with

TPN catheters. One group had second daily dressings of gauze and tape, on skin prepared with acetone and povidone iodine but with povidone iodine ointment around the catheter and a connection set between catheter and

giving set. The second group had Op Site’% dressings

changes every 7 days on similarly prepared skin, but no povidone iodine around the catheter and no connection set.

Bacterial analyses were performed on routine laboratory

catheter tip and blood cultures rather than more specific

methods on cultures from the skin and subcutaneous catheter segments [ 191. Increased sepsis rates were cited for

Op Site@ but did not reach statistical significance, but even

so comparison of these two poorly controlled groups make

any statistical analysis void and the conclusion that Op Site’s cannot be recommended as a dressing system for TPN systems invalid.

The other study by Keothane [9] was designed to test the hypothesis that a nutrition nurse reduced the incidence of

catheter related sepsis, and to investigate the virtues of tunnelling the catheter. Indirectly by comparing four

different groups of patients with and without tunnelling

and with and without a nutrition nurse, different regimens of dressings were also compared. In one group Op Siteg;

dressings were changed daily whereas in the other three

daily gauze and tape dressings were used. In the former group the dressings were performed by routine ward staff

whereas in the second group, those with gauze and tape dressings, the three day dressings were performed by the

nutrition nurse. The rates of sepsis were greater with Op

Site@ dressings, though again comparison cannot be valid as both skin preparation regimens and nursing staff performing these dressings were not well controlled.

Jarrand er al. [ll] have addressed the question of frequency of dressing changes and concluded that daily

changes of gauze and tape dressings with PVI cleansed skin and catheter-cutaneous junction PVI ointment, will

eliminate the growth of skin bacteria but that this regimen is expensive. We have shown that with a meticulous skin preparation of defatting, Chlorhexidine or povidone iodine and alcohol drying, Op Sitem dressings left intact for up to 7

days will restrict the proliferation of skin bacteria to a level below that considered to be infective.

This base line data on volunteers without indwelling

DISCUSSION

The major findings in the first part of this study are that

Op Site@ does not potentiate the growth of skin bacteria

and that up to 4 days after application, there are

consistently fewer organisms cultured beneath Op Site@ than either an occlusive PVC dressing or no dressing at all.

Bjornson et al. [3] have shown a statistically significant

correlation between recovering greater than 10’ per square centimetre from the skin and colonisation of the

subcutaneous segment of a central venous catheter. The

majority of colony counts beneath Op Site@ were less than 10’ organisms, the medians for both 2 and 4 days being less

than 10’ colony counts for both aerobic and anaerobic culture, whereas similar results were not achieved with

PVC dressing. Differences between the dressings were not detectable at 7 days and by 14 days adhesiveness had

deteriorated, the numbers were too small, and the only meaningful conclusion to be made is that two weeks is too

long to leave Op Site@ dressings on the skin. No pathogens

were cultured under Op Site@, though S. epidermidis may be pathogenic in sick surgical patients with an indwelling

catheter, it could not be considered pathogenic in the

framework of this study. In testing the effects of live different adhesive tapes on

skin bacteria. Marples and Kligman [16] found two forces that effected bacterial growth; hydration which was directly

proportional to the permeability (occlusiveness) of the tape and the presence of bactericidal agents in the adhesive.

Impermeable (occlusive) tapes greatly increased skin hydration with an explosive increase in the bacterial

population. The same workers have further extended this hydration effect of occlusive dressings to potentiate experimental staphylococcal infection [ 171. As identical

adhesives were used in the study the advantage of Op Site@

over PVC is that the former is semi-permeable and therefore more likely to keep the underlying skin relatively drier than PVC.

Though demonstrating that Op Site@ does not potentiate

the growth of skin bacteria, povidone iodine skin preparation alone was sufficient to keep colony counts

consistently below 10’ organisms per square centimetre. However, results from the second part of the study show

that, with a combination of defatting the skin and

Chlorhexidine or povidone iodine cleansing, all but one of the 48 average colony counts were less than 10’ organisms

per square centimetre after 4 and 7 days. The principle of defatting may none the less seem paradoxical. Though of benefit in reducing the growth of anaerobes especially the lipid dependent diphtheriods, defatting may remove certain fatty acids that have an anti-microbial role and therefore lead theoretically to colonisation [18]. However, the coagulase negative staphylococci, the predominant aerobic

34 SKIN BACTERIA AND OP SITE” CHEST DRESSINGS

catheters shows the relative safety of Op Site@ and should

provide the background for a well controlled study of patients with tunnelled TPN catheters using defatting and Chlorhexidine or povidone iodine skin preparation and

weekly Op Site@ dressing changes, with the supervision of a

nutrition nurse.

REFERENCES

[l] Ryan J A, Abel R M, Abbott W M et al 1974 Catheter complications in total parenteral nutrition. A prospective study of 200 consecutive patients. New England Journal of Medicine 290: 757-761

[2] Grungerg R N 1983 Sepsis and central venous catheter systems. In: Peters J L fed) A Manual of Central Venous Catheterization and Parenteral Nutrition. Wright, P.S.G., Bristol, pp 162-171

[3] Bjornson H S, Colley R, Bonner R H et al 1982 Association between microorganism growth at catheter insertion site and colonization of the catheter in patients receiving total parenteral nutrition. Surgery 92: 720-727

[4] Snydman D R, Pober B R, Murray S A et al 1982 Predictive value of surveillance skin cultures in total- parenteral-nutrition-related infection. The Lancet II: 1385-1388

(51 Mitchell A, Draper C, Lee D R et al 1981 A simple system of parenteral nutrition. Annals of The Royal

College of Surgeons of England. 63: 173- I76 [6] Sawyer L T, Peters J L 1983 Nursing care. In: Peters J L

(ed) A Manual of Central Venous Catheterization and Parenteral Nutrition. Wright, P.S.G., Bristol, pp 180-201

[7] Marples R R 1965 The effect of hydration on the bacterial flora of the skin. In: Maibach H I, Hildick-Smith G (eds) Skin Bacteria and their role in infection, McGraw-Hill, New York

(81 Powell C, Regan C, Fabri P J, Ruberg R L 1982 Evaluation of Op Site” catheter dressings for parenteral nutrition: a prospective, randomised study. Journal of Parenteral and Enteral Nutrition. 6: 43-46

[9] Keothane P P, Attrill H, Northover J et al 1983 Effect of catheter tunnelling and a nutrition nurse on catheter sepsis during parenteral nutrition. A controlled trial. The Lancet II: 1388-1390

ACKNOWLEDGEMENTS

We wish to thank Dr P. N. Adams of Smith & Nephew (U.K.) Ltd. for advice, financial support and arranging the preparation of Op Site” and PVC strips and supply of Propax swabs and Arklone. We are grateful to Mrs Hazel Mapperson for typing the manuscript.

[lo] Kaul A F, Jewett J F 1981 Agents and techniques for disinfection of the skin. Surgery, Gynaecology and Obstetrics. 152: 677-685

[l l] Jarrand M M, Freeman J B 1977 The effects of antibiotic ointments and antiseptics on the skin flora beneath subclavian catheter dressings during intravenous hyperalimentation. Journal of Surgical Research 22: 521-526

[12] Jarrand M M, Olson C, Freeman J B 1980 Daily dressing change effects on skin flora beneath subclavian catheter dressings during total parenteral nutrition. Journal of Parenteral and Enteral Nutrition 4: 391-391

[13] Blackett R L, Bakran A, Bradley J A et al 1978 A prospective study of subclavian vein catheters used exclusively for the purpose of intravenous feeding. British Journal of Surgery 65: 393-395

[14] Padberg F T, Ruggiero J, Blackburn G L, Bristrian B R 1981 Central venous catheterization for parenteral nutrition. Annals of Surgery 193: 264-270

[15] Williamson P, Kligman A M 1965 A new method for quantitative investigation of cutaneous bacteria. Journal of Investigative Dermatology 45: 498-503

[16] Marples R R, Kligman A M 1969 Growth of bacteria under adhesive tapes. Archives Dermatology 99: 107-l 10

(171 Marples R R, Kligman A M 1975 Experimental staphylococcal infections of the skin in man. In: Jeljaszewicz J (ed) Staphylococci and Staphylococcal diseases. Gustav Fischer Verlag, Stuttgart, pp 755-760

[l8] Ricketts C R, Squire J R, Topley E 1951 Human skin lipids with particular reference to the self-sterilizing power of the skin. Clinical Sciences: 10: 89-l 11

[19] Maki D G, Weise C E, Sarafin H W 1977 A semi- quantitative culture method for identifying intravenous- catheter-related infection. New England Journal of Medicine 296: 1305- 1309

Submission date 21 June 84. Accepted after revision 15 Nov. 84.