Journal
of Hospital
Effects
Infection
(1991)
17, 217-272
of whole body disinfection patients undergoing elective
D. J. Byrne, Departments
of Surgery
A. Napier,
G. Phillips*
in
and A. Cuschieri
and *Microbiology, Ninewells Schosol, Dundee DDl 9SY
Accepted for publication
on skin flora surgery
Hospital
and Medical
11 December 1990
Summary: Bacterial skin flora were studied in two groups of patients having three showers with either a 4% chlorhexidine detergent solution (Group A, N= 57) or a placebo detergsent (Group B, N= 58). Previous reports on the efficacy of chlorhexidine in decreasing bacterial counts on the skin were confirmed and the time taken to recolonization (median 5 days; range l-10 days) was in ‘broad agreement with previous reports. However, concern regarding the colonization of the skin of the patients in the chlorhexidine group by potential pathogens during the recolonization period appears unfounded as there was no significant difference in the incidence of nonresident skin flora between the chlorhexidine (17/57; 30%) and the placebo (14/58; 24%) groups. These non-residents are generally lost from the skin before discharge in the chlorhexidine group but nine patients in the placebo group had abnormal skin flora at discharge from hospital. All those patients tested after discharge had lost the non-resident flora within 2 weeks of discharge. The results of this study indicate that recolonization of the skin after whole body disinfection does not present a clinical problem. Keywords:
Chlorhexidine;
skin disinfection;
surgery.
Introduction
Despite many studies, the efficacy of preoperative whole body disinfection (WBD) in reducing the postoperative wound infection rate is still unproven. ‘-lo Washing with1 chlorhexidine has been shown to reduce bacterial skin counts,11-15 but the association between this and a reduction in wound infection rate has never been precisely determined. Washing with unmedicated soap has been shown to actually increase the number of skin organisms shed. 11,12,16Many studies have stressed the importance of the normal skin microflora as protection against colonization by potential pathogens,‘7m23 and as WBD with chlorhexidine significantly reduces the density of the normal skin flora, we were concerned that this may lead to Correspondence
to: Derek J. Byrne
0195-6701/91/030217+06$03.00/0
% 1991 The Hospitai
217
Infection
Soclety
218
D. J. Byrne
et
al.
problems with recolonization of the patient’s skin with potential pathogens. As part of a large trial investigating preoperative WBD, the response of the skin flora to showering with either a 4% chlorhexidine detergent solution or a placebo detergent was studied particularly in relation to the skin flora in the postoperative period. Materials
and methods
Two groups of preoperative patients were studied (Table I). Group A had three showers with a 4% chlorhexidine detergent solution (‘Hibiscrub’, ICI Pharmaceuticals) and Group B had three showers with a placebo detergent. The allocation was on a random, blind basis. Approximately 50 ml of detergent was used for each shower which was lathered up using a sterile plastic sponge, a separate sponge being used for each shower. A standard set of washing instructions was given to each patient which directed them to wash from the head downwards (including the hair), rinse, repeat the wash and dry themselves thoroughly with a freshly laundered towel. Clean clothes were worn after showering. Before showering and daily thereafter until discharge from hospital blood agar skin contact plates (‘Sterilin’; 2.5 cm2) with a chlorhexidine neutralizer (0.3% azolectin and 2% polysorbate 80) were applied to both axillae and groins for approximately 2s each. One contact plate was applied to each site giving a total of four plates per patient every time. These plates were incubated aerobically at 37°C for 24 h after which colony counts were performed manually using a perspex grid divided up into cm squares. As counts of over 1500 colony forming units (cfu) per plate were deemed to be inaccurate, such counts were recorded as greater than 1500. The results were expressed as the means of the four plate counts, as the trend in colony counts, up or down, was largely maintained between the four areas sampled, though groin counts were generally higher than axillary counts. Organisms were identified by colony appearance under x 3 magnification and an intense angled light source, Gram staining and other techniques as required. Because of limited resources staphylococcal species were not further identified and this has to be kept in mind when the results are considered. Table
I. Patient
details
Group A (chlorhexidine) No. of patients Male Female Median age (range) Potentially contaminated cases Clean cases Median number of days in hospital
57 E 61 (2&79) (range)
ii 7 (S-11)
Group B (placebo)
Whole
For statistical analysis U-test was used.
body disinfection
of differences
219
and skin flora
between
groups the Mann-Whitney
Results
The overall results showed a 20-fold reduction in the skin flora after showering (P < 0.01) in Group A while there was no change in the skin flora of the patients in Group B after showering (Table II). There was no significant difference in pre-showering skin bacterial counts between the two groups. The median time to recolonization in Group A was 5 days (range l-10). During recolonization 17/57 (30%) patients in the chlorhexidine group were colonized by organisms of a type not usually found on the skin (Proteus, Pseudomonas and Klebsiella species and Escherichia coli) or by an organism present in greater numbers than usual (Acinetobacter and Bacillus species) (Table III). These organisms appeared on the skin at a median time of 2 days after showering (range l-4 days) and persisted on the skin in hospital for a median time of six days (range 5-> 10 days). As recolonization of the skin with the normal flora progressed (mainly staphylococcal species and diphtheroids) the other organisms were gradually replaced so that by discharge only one out of the 17 patients still had minor abnormalities of tlhe skin flora, the other 16 patients having lost the non-residents from their skin flora. The skin flora of the single patient
Table
II. Median
colony counts (range)
in colony forming Before
Group Group
units/plate, showering
726 (76-> 603 (31->
A B
before and after showering After
1500)* 1500)
showering
36 (l-1441)* 594 (15->1500)
*p 1 week after WBD’5,24,25 and our results broadly agree with these reports. However, detailed studies of the types of organisms found on the skin during recolonization have never been published. The results of this study have shown that: (1) bacterial species normally non-resident on the skin are found there in some patients during convalescence and (2) bacterial species normally found on the skin in relatively small numbers appear in large numbers during convalescence. Many studies have stressed the importance of the normal skin microflora as protection against colonization by potential pathogens. Suppression of the normal flora by deodorants has led to colonization by Proteus species.” A similar effect has been found with hexachlorophane in neonates” and topical and systemic antimicrobial agents have been shown to produce overgrowth and clinical infection with Gram-negative bacilli and other bacteria not normally able to establish themselves on the skin.8a2s23 Various factors may be involved in this phenomenon including nutritional competition26 and the decreased production of a wide range of antimicrobial substances including fatty acids and antibiotic-type substances.17,** Of particular surgical interest and relevance to WBD in reducing wound infection rates, Selwyn** showed that among 90 surgical patients the presence of inhibitory skin bacteria was associated with a relatively low rate of wound colonization by pathogenic bacteria. Our study has shown a similar spectrum of non-resident skin organisms regardless of whether an antiseptic or placebo detergent was used in the One small difference was that the preoperative WBD regimen. chlorhexidine group had three patients with two contaminating organisms on the skin during recovery whereas the placebo group had only one
Whole
body
disinfection
and skin
221
flora
dominant organism per patient. Another difference was that there was only one patient in the chlorhexidine group with abnormal skin flora at discharge but in the placebo group nine patients had abnormalities of their skin flora when leaving hospital. This effect did not seem to last long when the patient returned home. The question is whether these organisms can cause problems while they are on the skin. It could be argued that they may contaminate the operative wound during the postoperative period, but as resistance to infection from surface this shows almost complete is unlikely. contamination within 24-48 h of operation 27,28this possibility group and The wound infection rate was 14% (8/57) in the chlorhexidine 10% (6/58) in the placebo group. Valid comparison is not possible due to the small numbers of patients. Another problem could be dissemination into the air of these bacteria by the patients. However, as the patients who were initially contaminated after showering probably picked up the bacteria from the ward environment, ffurther spread from these patients is not likely to cause problems. However, those patients who are contaminated and are discharged home before their skin is back to normal may introduce ‘hospital’ bacteria into the h’ome with the possibility of starting infection there. It has not been possible for this study to address this problem, except to point out that the five patients followed up at home had lost their non-resident skin flora within 2 weeks of discharge from hospital. The disinfection
trial
was supported
by ICI
Pharmaceuticals,
Macclesfield,
UK.
References 1. Brunn JN. Postoperative wound infection: predisposing factors and the effects of a reduction in the dissemination of staphylococci. Acta Med &and 1970; 188: l-89. 2. Brandberg A, Holm J, Hammarsten J, Schersten T. Postoperative wound infections in vascular surgery: effect of preoperative whole body disinfection by shower-bath with chlorhexidine soap. Royal Society of Medicine: International Congress and Symposium No. 23 1979; 75-85. of wound infection: a 10 year prospective study 3. Cruse PJE, Foord R. The epidemiology of 62939 wounds. Surg Clin North Am 1980; 60: 2740. 4. Ayliffe GAJ, Noy MF, Babb JR, Davies JG, Jackson J. A comparison of preoperative bathing with chlorhexidine-detergent and non-medicated soap in the prevention of wound infection. 7 Hasp Infect 1983: 4: 237-244. Leigh DA, Strange JL; Marriner J, ‘Sedgwick J. Total body bathing with ‘Hibiscrub’ (chlorhexidine) in surgical patients: a controlled trial. J Hasp Infect 1983; 4: 229-235. Randall PE, Ganguli L, Marcuson RW. Wound infection following vasectomy. Br J Ural 1983; 55: 56+567. Wells FC, Newsom SWB, Rowland C. Wound infection in cardiothoracic surgery. Lancet 1983; 1: 1209-1210. Hayek LJ, Emerson JM, Gardner AMN. A placebo-controlled trial of the effect of two preoperative baths or showers with chlorhexidine detergent on postoperative wound infection rates. J Hasp Infect 1987; 10: 165-172. Party on Control of Hospital Infection. A comparison of the effects 9. European Working of pre-operative whole-body bathing with detergent alone and with detergent containing chlorhexidine gluconate on the frequency of wound infection after clean surgery. r Hasp Infect 1988; 11: 31&320. 10. Van Diemen AH. Prevention of surgical wound infection by whole body bathing with ‘Hibiscrub’. Second International Conference on Infection Control, Harrogate 1988.
222
D. J. Byrne
et al.
11. Davies J, Babb JR, Ayliffe GAJ. The effect on the skin flora of bathing with antiseptic solutions. Journal of Antimicrobial Chemotherapy 1977; 3: 473-481. 12. Brandberg A, Andersson I. Whole body disinfection by shower-bath with chlorhexidine soap. Royal Society of Medicine: International Congress and Symposium No. 23 1979; 65-70. 13. Seeberg S, Lindberg A, Bergman BR. Preoperative shower bath with 4% chlorhexidine detergent solution: reduction of Staphylococcus aweus in skin carriers and practical application. In: Maibach HI, Aly R, Eds. Skin Microbiology, Relevance to Clinical Infection. Berlin, West Germany: Springer-Verlag, 1981; 86-91. 14. Mitchell NJ. Whole body disinfection with chlorhexidine: is shower bathing more effective than bathing? J Hosp Infect 1984; 5: 96-99. 15. Rowland C, Newsom SWB. Effect of chlorhexidine baths on peri-operative skin flora. Second International Conference on Infection Control, Harrogate 1988. 16. Meers PD, Yeo GA. Shedding of bacteria and skin squames after handwashing. J Hyg 1978; 81: 99-105. 17. Shehadeh NH, Kligman AM. The effect of topical antibacterial agents on the bacterial flora of the axilla. J Invest Dermatol 1963; 40: 61-71. 18. Savin JA. Pseudomonas aeruginosa infections in a skin ward. Trans St. John’s Hasp Dermatol Sot 1967; 53: 75-79. 19. Forfar JO, Gould JC, MacCabe AF. Effect of hexachlorophane on incidence of staphylococcal and gram-negative infection in the newborn. Lancet 1968; 2: 177-178. 20. Aly R, Maibach HI, Strauss WG, Shinefield HR. Effect of a systemic antibiotic on nasal bacterial ecology in man. Appl Microbial 1970; 20:240-244. 21. Marples RR, Kligman AM. Ecological effects of oral antibiotics on the microflora of human skin. Arch Dermatol 1971; 103: 148-153. 22. Selwyn S. Natural antibiosis among skin bacteria as a primary defence against infection. Br J Dermatol 1975; 93: 487-493. 23. Pollock A. Surgical Infections. London: Edward Arnold 1987. 24. Lindberg A. Staphylococcus aureus i hudfloran. En method for semikvantitativ bedomning-effekt av tvattning med 4-proc klorhexidin. Lakartidningen 1977; 74: 1185-1186. 25. Taggart DP, Paul B, Pennycook A, Poon F, Mackenzie I. A comparison of preoperative bathing with soap and chlorhexidine on reduction and times to recolonization of skin flora. 1st International Conference of the Hospital Infection Society, London 1987. 26. Wickman K. Studies of bacterial interference in experimentally produced burns in guinea pigs. Acta Pathol, Microbial Immunol Stand (B) 1970; 78: 1 S-28. 27. Schauerhamer RA, Edlich RF, Panek P, Thul J, Prusak M, Wangensteen OH. Studies in the management of the contaminated wound. VII. Susceptibility of surgical wounds to postoperative surface contamination. AmJ Surg 1971; 122: 74-77. model for assessing risk of 28. Bibby BA, Collins BJ, Ayliffe GAJ. A mathematical postoperative wound infection. J Hosp Infect 1986; 8: 31-38.
of Hospital
Effects
Infection
(1991)
17, 217-272
of whole body disinfection patients undergoing elective
D. J. Byrne, Departments
of Surgery
A. Napier,
G. Phillips*
in
and A. Cuschieri
and *Microbiology, Ninewells Schosol, Dundee DDl 9SY
Accepted for publication
on skin flora surgery
Hospital
and Medical
11 December 1990
Summary: Bacterial skin flora were studied in two groups of patients having three showers with either a 4% chlorhexidine detergent solution (Group A, N= 57) or a placebo detergsent (Group B, N= 58). Previous reports on the efficacy of chlorhexidine in decreasing bacterial counts on the skin were confirmed and the time taken to recolonization (median 5 days; range l-10 days) was in ‘broad agreement with previous reports. However, concern regarding the colonization of the skin of the patients in the chlorhexidine group by potential pathogens during the recolonization period appears unfounded as there was no significant difference in the incidence of nonresident skin flora between the chlorhexidine (17/57; 30%) and the placebo (14/58; 24%) groups. These non-residents are generally lost from the skin before discharge in the chlorhexidine group but nine patients in the placebo group had abnormal skin flora at discharge from hospital. All those patients tested after discharge had lost the non-resident flora within 2 weeks of discharge. The results of this study indicate that recolonization of the skin after whole body disinfection does not present a clinical problem. Keywords:
Chlorhexidine;
skin disinfection;
surgery.
Introduction
Despite many studies, the efficacy of preoperative whole body disinfection (WBD) in reducing the postoperative wound infection rate is still unproven. ‘-lo Washing with1 chlorhexidine has been shown to reduce bacterial skin counts,11-15 but the association between this and a reduction in wound infection rate has never been precisely determined. Washing with unmedicated soap has been shown to actually increase the number of skin organisms shed. 11,12,16Many studies have stressed the importance of the normal skin microflora as protection against colonization by potential pathogens,‘7m23 and as WBD with chlorhexidine significantly reduces the density of the normal skin flora, we were concerned that this may lead to Correspondence
to: Derek J. Byrne
0195-6701/91/030217+06$03.00/0
% 1991 The Hospitai
217
Infection
Soclety
218
D. J. Byrne
et
al.
problems with recolonization of the patient’s skin with potential pathogens. As part of a large trial investigating preoperative WBD, the response of the skin flora to showering with either a 4% chlorhexidine detergent solution or a placebo detergent was studied particularly in relation to the skin flora in the postoperative period. Materials
and methods
Two groups of preoperative patients were studied (Table I). Group A had three showers with a 4% chlorhexidine detergent solution (‘Hibiscrub’, ICI Pharmaceuticals) and Group B had three showers with a placebo detergent. The allocation was on a random, blind basis. Approximately 50 ml of detergent was used for each shower which was lathered up using a sterile plastic sponge, a separate sponge being used for each shower. A standard set of washing instructions was given to each patient which directed them to wash from the head downwards (including the hair), rinse, repeat the wash and dry themselves thoroughly with a freshly laundered towel. Clean clothes were worn after showering. Before showering and daily thereafter until discharge from hospital blood agar skin contact plates (‘Sterilin’; 2.5 cm2) with a chlorhexidine neutralizer (0.3% azolectin and 2% polysorbate 80) were applied to both axillae and groins for approximately 2s each. One contact plate was applied to each site giving a total of four plates per patient every time. These plates were incubated aerobically at 37°C for 24 h after which colony counts were performed manually using a perspex grid divided up into cm squares. As counts of over 1500 colony forming units (cfu) per plate were deemed to be inaccurate, such counts were recorded as greater than 1500. The results were expressed as the means of the four plate counts, as the trend in colony counts, up or down, was largely maintained between the four areas sampled, though groin counts were generally higher than axillary counts. Organisms were identified by colony appearance under x 3 magnification and an intense angled light source, Gram staining and other techniques as required. Because of limited resources staphylococcal species were not further identified and this has to be kept in mind when the results are considered. Table
I. Patient
details
Group A (chlorhexidine) No. of patients Male Female Median age (range) Potentially contaminated cases Clean cases Median number of days in hospital
57 E 61 (2&79) (range)
ii 7 (S-11)
Group B (placebo)
Whole
For statistical analysis U-test was used.
body disinfection
of differences
219
and skin flora
between
groups the Mann-Whitney
Results
The overall results showed a 20-fold reduction in the skin flora after showering (P < 0.01) in Group A while there was no change in the skin flora of the patients in Group B after showering (Table II). There was no significant difference in pre-showering skin bacterial counts between the two groups. The median time to recolonization in Group A was 5 days (range l-10). During recolonization 17/57 (30%) patients in the chlorhexidine group were colonized by organisms of a type not usually found on the skin (Proteus, Pseudomonas and Klebsiella species and Escherichia coli) or by an organism present in greater numbers than usual (Acinetobacter and Bacillus species) (Table III). These organisms appeared on the skin at a median time of 2 days after showering (range l-4 days) and persisted on the skin in hospital for a median time of six days (range 5-> 10 days). As recolonization of the skin with the normal flora progressed (mainly staphylococcal species and diphtheroids) the other organisms were gradually replaced so that by discharge only one out of the 17 patients still had minor abnormalities of tlhe skin flora, the other 16 patients having lost the non-residents from their skin flora. The skin flora of the single patient
Table
II. Median
colony counts (range)
in colony forming Before
Group Group
units/plate, showering
726 (76-> 603 (31->
A B
before and after showering After
1500)* 1500)
showering
36 (l-1441)* 594 (15->1500)
*p 1 week after WBD’5,24,25 and our results broadly agree with these reports. However, detailed studies of the types of organisms found on the skin during recolonization have never been published. The results of this study have shown that: (1) bacterial species normally non-resident on the skin are found there in some patients during convalescence and (2) bacterial species normally found on the skin in relatively small numbers appear in large numbers during convalescence. Many studies have stressed the importance of the normal skin microflora as protection against colonization by potential pathogens. Suppression of the normal flora by deodorants has led to colonization by Proteus species.” A similar effect has been found with hexachlorophane in neonates” and topical and systemic antimicrobial agents have been shown to produce overgrowth and clinical infection with Gram-negative bacilli and other bacteria not normally able to establish themselves on the skin.8a2s23 Various factors may be involved in this phenomenon including nutritional competition26 and the decreased production of a wide range of antimicrobial substances including fatty acids and antibiotic-type substances.17,** Of particular surgical interest and relevance to WBD in reducing wound infection rates, Selwyn** showed that among 90 surgical patients the presence of inhibitory skin bacteria was associated with a relatively low rate of wound colonization by pathogenic bacteria. Our study has shown a similar spectrum of non-resident skin organisms regardless of whether an antiseptic or placebo detergent was used in the One small difference was that the preoperative WBD regimen. chlorhexidine group had three patients with two contaminating organisms on the skin during recovery whereas the placebo group had only one
Whole
body
disinfection
and skin
221
flora
dominant organism per patient. Another difference was that there was only one patient in the chlorhexidine group with abnormal skin flora at discharge but in the placebo group nine patients had abnormalities of their skin flora when leaving hospital. This effect did not seem to last long when the patient returned home. The question is whether these organisms can cause problems while they are on the skin. It could be argued that they may contaminate the operative wound during the postoperative period, but as resistance to infection from surface this shows almost complete is unlikely. contamination within 24-48 h of operation 27,28this possibility group and The wound infection rate was 14% (8/57) in the chlorhexidine 10% (6/58) in the placebo group. Valid comparison is not possible due to the small numbers of patients. Another problem could be dissemination into the air of these bacteria by the patients. However, as the patients who were initially contaminated after showering probably picked up the bacteria from the ward environment, ffurther spread from these patients is not likely to cause problems. However, those patients who are contaminated and are discharged home before their skin is back to normal may introduce ‘hospital’ bacteria into the h’ome with the possibility of starting infection there. It has not been possible for this study to address this problem, except to point out that the five patients followed up at home had lost their non-resident skin flora within 2 weeks of discharge from hospital. The disinfection
trial
was supported
by ICI
Pharmaceuticals,
Macclesfield,
UK.
References 1. Brunn JN. Postoperative wound infection: predisposing factors and the effects of a reduction in the dissemination of staphylococci. Acta Med &and 1970; 188: l-89. 2. Brandberg A, Holm J, Hammarsten J, Schersten T. Postoperative wound infections in vascular surgery: effect of preoperative whole body disinfection by shower-bath with chlorhexidine soap. Royal Society of Medicine: International Congress and Symposium No. 23 1979; 75-85. of wound infection: a 10 year prospective study 3. Cruse PJE, Foord R. The epidemiology of 62939 wounds. Surg Clin North Am 1980; 60: 2740. 4. Ayliffe GAJ, Noy MF, Babb JR, Davies JG, Jackson J. A comparison of preoperative bathing with chlorhexidine-detergent and non-medicated soap in the prevention of wound infection. 7 Hasp Infect 1983: 4: 237-244. Leigh DA, Strange JL; Marriner J, ‘Sedgwick J. Total body bathing with ‘Hibiscrub’ (chlorhexidine) in surgical patients: a controlled trial. J Hasp Infect 1983; 4: 229-235. Randall PE, Ganguli L, Marcuson RW. Wound infection following vasectomy. Br J Ural 1983; 55: 56+567. Wells FC, Newsom SWB, Rowland C. Wound infection in cardiothoracic surgery. Lancet 1983; 1: 1209-1210. Hayek LJ, Emerson JM, Gardner AMN. A placebo-controlled trial of the effect of two preoperative baths or showers with chlorhexidine detergent on postoperative wound infection rates. J Hasp Infect 1987; 10: 165-172. Party on Control of Hospital Infection. A comparison of the effects 9. European Working of pre-operative whole-body bathing with detergent alone and with detergent containing chlorhexidine gluconate on the frequency of wound infection after clean surgery. r Hasp Infect 1988; 11: 31&320. 10. Van Diemen AH. Prevention of surgical wound infection by whole body bathing with ‘Hibiscrub’. Second International Conference on Infection Control, Harrogate 1988.
222
D. J. Byrne
et al.
11. Davies J, Babb JR, Ayliffe GAJ. The effect on the skin flora of bathing with antiseptic solutions. Journal of Antimicrobial Chemotherapy 1977; 3: 473-481. 12. Brandberg A, Andersson I. Whole body disinfection by shower-bath with chlorhexidine soap. Royal Society of Medicine: International Congress and Symposium No. 23 1979; 65-70. 13. Seeberg S, Lindberg A, Bergman BR. Preoperative shower bath with 4% chlorhexidine detergent solution: reduction of Staphylococcus aweus in skin carriers and practical application. In: Maibach HI, Aly R, Eds. Skin Microbiology, Relevance to Clinical Infection. Berlin, West Germany: Springer-Verlag, 1981; 86-91. 14. Mitchell NJ. Whole body disinfection with chlorhexidine: is shower bathing more effective than bathing? J Hosp Infect 1984; 5: 96-99. 15. Rowland C, Newsom SWB. Effect of chlorhexidine baths on peri-operative skin flora. Second International Conference on Infection Control, Harrogate 1988. 16. Meers PD, Yeo GA. Shedding of bacteria and skin squames after handwashing. J Hyg 1978; 81: 99-105. 17. Shehadeh NH, Kligman AM. The effect of topical antibacterial agents on the bacterial flora of the axilla. J Invest Dermatol 1963; 40: 61-71. 18. Savin JA. Pseudomonas aeruginosa infections in a skin ward. Trans St. John’s Hasp Dermatol Sot 1967; 53: 75-79. 19. Forfar JO, Gould JC, MacCabe AF. Effect of hexachlorophane on incidence of staphylococcal and gram-negative infection in the newborn. Lancet 1968; 2: 177-178. 20. Aly R, Maibach HI, Strauss WG, Shinefield HR. Effect of a systemic antibiotic on nasal bacterial ecology in man. Appl Microbial 1970; 20:240-244. 21. Marples RR, Kligman AM. Ecological effects of oral antibiotics on the microflora of human skin. Arch Dermatol 1971; 103: 148-153. 22. Selwyn S. Natural antibiosis among skin bacteria as a primary defence against infection. Br J Dermatol 1975; 93: 487-493. 23. Pollock A. Surgical Infections. London: Edward Arnold 1987. 24. Lindberg A. Staphylococcus aureus i hudfloran. En method for semikvantitativ bedomning-effekt av tvattning med 4-proc klorhexidin. Lakartidningen 1977; 74: 1185-1186. 25. Taggart DP, Paul B, Pennycook A, Poon F, Mackenzie I. A comparison of preoperative bathing with soap and chlorhexidine on reduction and times to recolonization of skin flora. 1st International Conference of the Hospital Infection Society, London 1987. 26. Wickman K. Studies of bacterial interference in experimentally produced burns in guinea pigs. Acta Pathol, Microbial Immunol Stand (B) 1970; 78: 1 S-28. 27. Schauerhamer RA, Edlich RF, Panek P, Thul J, Prusak M, Wangensteen OH. Studies in the management of the contaminated wound. VII. Susceptibility of surgical wounds to postoperative surface contamination. AmJ Surg 1971; 122: 74-77. model for assessing risk of 28. Bibby BA, Collins BJ, Ayliffe GAJ. A mathematical postoperative wound infection. J Hosp Infect 1986; 8: 31-38.
