- Email: [email protected]
Investigation of nosocomial prosthetic valve endocarditis due to antibiotic-resistant Staphylococcus epidermidis
Journal
of Hospital
Infection
Investigation endocarditis
(1991)
19, 107-l
14
of nosocomial prosthetic due to antibiotic-resistant
Staphylococcus R. Menzies, Microbiology
epidermidis
D. MacCulloch
Department,
and
Green Lane Hospital, 3, New Zealand
Accepted
valve
for publication
B. Cornere Greenlane
26 j%ly
West,
Auckland
1991
Summary: A reservoir of antibiotic-resistant Staphylococcus epidermidis strains in our cardiac surgery unit appeared to be the source of organisms responsible for three cases of early prosthetic valve endocarditis. Staphylococcus epidermidis isolates recovered from the skin of 13 patients before and after surgery were compared. All were typed by plasmid profile, antimicrobial susceptibility and slime production. The three strains from early prosthetic valve endocarditis resembled the antibiotic-resistant nosocomial strains recovered from the skin of eight patients following surgery and the environment of the operating theatres. These strains expressed resistance to oxacillin, gentamicin, kanamycin and tobramycin and most produced slime, whereas those isolated from the skin of patients at the time of admission were predominantly susceptible to antibiotics and few produced slime. Keywords: Nosocomial cus epidermidis; slime
infection; production.
prosthetic
valve
endocarditis;
Stuphylococ-
Introduction Staphylococcus epidermidis is an important cause of prosthetic heart valve infection.lm3 It is ubiquitous on human skin which is the apparent source of early postoperative infections. l-3 Some evidence indicates that strains of S. epidermidis associated with early prosthetic valve endocarditis may be acquired during hospitalization and that these strains are different from those colonizing patients’ skin at the time of admission to hospital.3B4 Strains present at the time of admission are usually susceptible to antibiotics whereas those cultured from hospital patients following surgery are frequently multiply-resistant.3s4 We investigated three cases of prosthetic valve endocarditis due to multiply antibiotic-resistant S. epidermidis. Each infection occurred within 2 months of surgery at intervals of c. 6 months. Because antimicrobial Correspondence West, Auckland 0195-670t/91/100107+08
to: Rosalie Menzies, 3, New Zealand.
Microbiology
Department,
SOS.OO,O
Green
Lane
Hospital,
0 1991 The Hospital
107
Greenlane
Infection
Society
,
108
R. Menzies
et al.
susceptibility and plasmid profiles indicated that each isolate was a different strain, and a different surgical team was responsible for each heart valve replacement, it seemed unlikely that a common source of infection such as a carrier was involved. Our aim was to determine whether the source of these strains was the patients’ own flora or a reservoir of organisms in the cardiac surgery unit. The method of investigation was a survey of S. epidermidis strains cultured from skin swabs taken from 13 patients on admission to hospital and 7 days after cardiac by-pass surgery. At the same time, the operating theatre environment was sampled for the presence of antibiotic-resistant S. epidermidis during surgical procedures. We felt that a wider investigation could not be justified in the absence of an outbreak of infection. Materials
and
methods
Patients for cardiac surgery were admitted to the general ward of the cardiac surgery unit no earlier than the day before surgery. After surgery they were transferred to an Intensive Care Unit for at least 24 h before transfer to a special care room in the general ward. At discharge they were resident in the area of the general ward which acted as a reception area for incoming patients. An average of 910 cardiac by-pass operations are carried out each year and 220 of these are heart valve replacements, with prosthetic valves. Since 1977 patients for cardiac by-pass surgery have received antibiotic prophylaxis with cephradine (E. R. Squibb & Sons), 1 g iv. on anaesthetic induction, 1 g i.v. at the start of by-pass surgery and 1 g i.v. on return to the Intensive Care Unit. This was followed by 500 mg q 6 h p.o. or i.v. for 4 days. The night before surgery patients were shaved and had a total body wash with 4% chlorhexidine gluconate. Preoperatively the body wash was repeated and 2.5% tincture of iodine was applied to the incision site. In theatre the iodine tincture application was repeated before surgery, and following surgery the wound was covered with a sterile dry dressing. Between 1987 and 1989 three patients developed early prosthetic valve endocarditis. All patients recovered; two after replacement of the infected prosthetic material and antibiotic treatment, and the third after treatment with antibiotics only. Their details are summarized in Table I.
Patient skin survey The anterior nares and skin over the sternum of 13 patients admitted to hospital between 18 January 1989 and 29 May 1989 were swabbed on admission to hospital and 7 days after cardiac by-pass surgery; heart valve replacement (six patients), coronary artery grafting (six patients) and repair of aneurysm (one patient). The first patient available each week was swabbed. All patients were from the elective surgery waiting list, 18 to 65 years of age and had not received antibiotics or been in hospital within the preceding 4 weeks. The nose was swabbed with a dry swab and the skin over
P.V.E.
due
the sternum with a swab moistened Culture swab transport system).
109
to S. epidermidis
with
Amies
transport
medium
(Difco
Operating theatres During the patient survey the two operating theatres used for cardiac surgery were sampled for the presence of antibiotic-resistant S. epidermidis strains during surgical procedures. Twenty agar plates, Columbia agar (Gibco) with 5% sheep blood, were placed to detect staphylococci in areas of staff movement. Each plate remained in position for the duration of one surgical procedure, average time 3 h.
Culture methods Each swab of the anterior nares and skin over the sternum was cultured on Columbia agar with 5% sheep blood and mannitol salt agar (Difco), and incubated aerobically at 35°C for 48 h. Staphylococcal colonies with representative morphology from each swab were identified and up to 20 presumptive S. epidermidis isolates were typed by susceptibility to 16 antimicrobial agents, plasmid profile and slime production. Agar plates from operating theatres were incubated at 35°C for 48 h and four staphylococcal colonies from each were identified. All presumptive S. epidermidis isolates were tested for susceptibility to oxacillin, and oxacillin-resistant isolates were typed by susceptibility to 16 antimicrobial agents, plasmid profile and slime production.
Identification
of isolates and strain differentiation
Presumptive identification as S. epidermidis was made by Gram’s stain, catalase production and failure to produce acid from trehalose and mannitol.5,6 Identification of probable nosocomial strains and isolates from prosthetic valve endocarditis as S. epidermidis was confirmed by API Staph (France). Strains were typed by antimicrobial susceptibility pattern, plasmid profile and slime production. 2,3,7 Susceptibility to 16 antimicrobial agents was determined by the disc diffusion method on Mueller-Hinton agar (Gibco) as recommended by the National Committee * for Clinical Laboratory Standards.’ Antibiotics tested were penicillin, oxacillin, gentamicin, kanamycin, tobramycin, amikacin, streptomycin, erythromycin, clindamycin, chloramphenicol, tetracycline, rifampicin, vancomycin, trimethoprim/sulphamethoxazole, neomycin and fusidic acid. Slime production was detected by a tube method.’ Plasmid DNA was isolated and separated by a rapid method adapted for S. epidermidis.‘O Isolates were grown on Columbia agar with 5% sheep blood instead of in L broth and bacterial cells were suspended in 10 ml of sterile distilled water to an opacity approximately equal to a number 5 McFarland standard. Lysostaphin (Sigma) was used at double the concentration suggested for
Staphylococcus aweus.
110
R. Menzies
et al.
Results
Characteristics of strains from prosthetic valve endocarditis These strains all had different plasmid and antimicrobial susceptibility profiles. Antimicrobial resistances are shown in Table I. Resistance to penicillin, oxacillin, gentamicin, kanamycin and tobramycin and slime production were common to all strains. Distribution of strains from patient skin swabs and operating theatres Antimicrobial susceptibility, slime production and plasmid profile separated the 215 S. epidermidis isolates obtained from skin swabs into 92 different strains; 49 strains were recovered on admission only, 39 after surgery only and four both on admission and after surgery. The number of strains isolated from each patient at any one time varied from one to eight (average four). Colonization of the anterior nares and skin over the sternum by the same strain was more prevalent after surgery (eight patients) than before surgery (two patients). Identical S. epidermidis strains from postoperative skin swabs, not found on admission swabs, from more than one patient or from one patient and an operating theatre were considered to be nosocomial in origin. Four nosocomial strains were isolated from postoperative swabs from eight
Table
Sex
I. Summary
Age (years)
of the clinical
history and microbiological prosthetic valve endocarditis
Operation
Postoperative
findings
infection
Onset
Treatment
51
21.9.87 Mitral valve replaced with prosthetic valve
2 months
Vancomycin + rifampicin Replacement of infected valve
Male
50
8 days
Male
27
1.7.88 Aortic valve + ascending aorta replaced with prosthetic valve + dacron graft 4.4.89 Porcine mitral valve replaced with prosthetic valve
Vancomycin + gentamicin + rifampicin Replacement of infected valve + graft Vancomycin + rifampicin + flucloxacillin SClindamycin + rifampicin + fusidic acid Discharged on ciprofloxacin
* A, Amikacin; C, chloramphenicol; Cc, clindamycin; oxacillin; P, penicillin; Sxt, trimethoprim/sulphamethoxazole; $ Changed to this regimen because of neutropenia.
G, gentamicin; To,
the three
patients
with
Microbiological results
Male
7 days
for
S. epidermidis from blood cultures + prosthetic valve. Resistant to P,Ox,G,K,To,N,Sxt* S. epidermidis from blood cultures + prosthetic valve + graft. Resistant to P,Ox,G,K,To,Cc,C,Sxt* S. epidermidis from blood cultures. Resistant to P,Ox,G,K,To,A,N,C,Sxt*
K, kanamycin; tobramycin.
N, neomycin;
Ox,
P.V.E. due
to
111
S. epidermidis
patients and one operating theatre. One strain was isolated from three patients and two strains each from one pair of patients. None of these patients were resident in the same area of the hospital at the same time. Operating theatres yielded one strain which was identical to an isolate recovered from a patient discharged from hospital 3 months previously. Antimicrobial susceptibility of strains from patient survey swabs and operating theatres Antimicrobial susceptibility tests of S. epidermidis strains isolated on admission and after surgery showed a very highly significant difference in antimicrobial susceptibility (Table II), P
Table II. epidermidis
Antimicrobial susceptibility of Staphylococcus isolated from patient skin swabs (13 patients, 215 isolates, 92 strains)
Antimicrobial agent
No.
of patients with strains susceptible
On admission (53 strains) Penicillin Oxacillin Erythromycin Tetracycline Clindamycin Kanamycin Gentamicin, tobramycin Cotrimoxazole Streptomycin Chloramphenicol Fusidic acid, Rifampicin Vancomycin
all
After surgery (43 strains)
5 13 10 12 12 13
0 1 6 6 8 3
13 13 13 13 13 13 13
4 6 8 9 11 12 13
neomycin,
amikacin
112
R. Menzies Table III. epidermidis
Nosocomial
et al.
Antimicrobial resistance of nosocomial strains isolatedpostoperativelyfrompatient and operating theatres strain
Staphylococcus skin swabs
Source
Antimicrobial resistance
.2
3 patients 2 patients
3
2 patients
4
1 patient + operating theatre
P,Ox,G,K,To,N* P,Ox,G,K,To, S,E.T,Sxt* P,OX,G,K, To.A.N.C,Sxt* P,Ox;G;K;Tb,N*
1
* A, Amikacin; C, chloramphenicol; E, erythromycin; G, gentamicin; K, kanamycin; N, neomycin; Ox, oxacillin; P, penicillin; S, streptomycin; Sxt, trimethoprim/sulphamethoxazole; T, tetracycline; To, tobramycin. Additional note: Isolates obtained on admission were all sensitive to: P, Ox, FG, K, To.
Slime production Table There strains
IV shows the number was a highly significant that produced slime P=O.O09 (X2 test).
of S. epidermidis strains that produced slime. difference in the proportion of S. epidermidis on admission to hospital and after surgery, Discussion
Antimicrobial resistances expressed by S. epidermidis strains recovered from patients’ skin swabs taken after surgery differed significantly from the resistances expressed by strains recovered on admission to the cardiac surgery unit. For most patients an oxacillin-resistant population had to other antibiotics had replaced an oxacillin-susceptible one. Resistance also increased (Table II) and the percentage of strains that produced slime had almost doubled (Table IV). We considered that isolates from admission swabs would give an indication of the characteristics of patients’ normal skin flora. These isolates produced slime at a rate of one in three and never expressed resistance to oxacillin, gentamicin, kanamycin and tobramycin. These findings indicated that patients’ normal flora was an unlikely source of isolates responsible for sporadic cases of early prosthetic valve endocarditis. Some antibiotic-resistant S. epidermidis strains may colonize patients’ skin in very small numbers on admission to hospital and can emerge during cardiac surgery with antimicrobial prophylaxis.“T’2 We did not identify any such strains. The strains termed nosocomial in our study were only isolated after surgery and from more than one patient or from a patient and the environment. Survey patients were not in the same area at the same time, therefore the environment, other patients and staff of the cardiac surgery unit were the most likely sources of these isolates. In one case contamination of the
P.V.E. due
Table IV. Slime production midis strains isolatedfrompatient to hospital S. epidermidis
Total
On admission After surgery Nosocomial Prosthetic valve endocarditis
53 43 4 3
Staphylococcus
by
epiderskin swabs on admission cardiac surgery
no.
No. slime producers
and after
strains
113
to S. epidermidis
18 (34%)
26 (61%) 3 3
environment of the operating theatre with the same strain isolated from a patient discharged from hospital 3 months previously was established. Our results support reports’3”4 that hospital reservoirs of antibiotic-resistant coagulase-negative staphylococci can be the source of antibiotic-resistant strains found on patients’ skin. Our study may have under reported the number and distribution of nosocomial strains in our cardiac surgery unit. Nosocomial strains were characterized by resistance to gentamicin, kanamycin and tobramycin. Genes encoding resistance to these antibiotics have been recorded in staphylococcal populations in Australia, USA and Europe. I5 Such a gene app ears to be established in the nosocomial S. epidermidis population in our cardiac surgery unit and to be expressed by the isolates responsible for early prosthetic valve endocarditis. Most nosocomial strains and all strains that infected prosthetic heart valves produced slime. These observations support the hypothesis that slime production by S. epidermidis protects against host defences and antibiotic activity16 and is associated with infection of prosthetic devices.17 The nosocomial S. epidermidis population exhibited the same characteristics that distinguished isolates responsible for prosthetic valve endocarditis; a variety of strains, slime production, resistance to oxacillin, gentamicin, kanamycin and tobramycin, and appeared to have been the source of these isolates. Archer” has postulated that a reservoir of antibiotic-resistant S. epidermidis strains and their genes can become established on patients’ skin when normal flora is disrupted by antimicrobial prophylaxis. From this reservoir, strains are disseminated by hospital workers to the operating theatre environment where they gain access to prosthetic heart valves during implantation. We have established that a reservoir of antibiotic-resistant S. epidermidis strains was established in our cardiac surgery unit and caused serious infection. The gene pool available for adaptation and survival in such a population would be enormous and available for transfer to Staphylococcus aweus and Streptococcus spp. *’ The recognition of S. epidermidis as an important pathogen in cardiac surgery units means that strategies for the identification and control of reservoirs of strains must be devised if nosocomial S. epidermidis prosthetic valve endocarditis is to be avoided.
114 This study was approved Whitlock for statistical
R. Menzies
et al.
by Green Lane Hospital analysis of results.
Ethics
Committee.
We thank
Dr
R. M.
L.
References 1. Lowv 2. 3. 4. 5. 6.
7.
8.
9.
10.
11.
12.
13. 14.
15. 16. 17.
FD. Hammer SM. Staahvlococcus eaidermidis infections. Ann Intern Med 1983: 99: 834-839. Parisi JT. Coagulase-negative staphylococci and the epidemiological typing of Staphylococcus epidermidis. Microbial Rev 1985: 49: 126-139. Pfaile; MA, Hkrwaldt LA. Laboratory, clinical and epidemiological aspects of coagulase-negative staphylococci. Clin Microbial Rev 1988; 1: 281-299. Archer GL. Molecular epidemiology of multiresistant StaPhvlococcus epidermidis. 7 _ Antimicrob Chemother 1988; 21 (SuppI. C): 133-138. Kloos WE. Schleifer KH. Simnlified scheme for routine identification of human Staphyloco&us species. J Clin M&robiol 1975; 1: 82-88. Stevens DL, Jones C. Use of trehalose-mannitol-phosphatase agar to differentiate Staphylococcus epidermidis and Staphylococcus saprophyticus from other coagulase-negative staphylococci. J Clin Microbial 1984; 20: 977-980. Archer GL, Karchmer AW, Vishniavsky N, Johnston JL. Plasmid-pattern analysis for the differentiation of infecting from non-infecting Staphylococcus epidermidis. J Infect Dis 1984; 149: 9133920. National Committee for Clinical Laboratory Standards.. Performance standards for antimicrobial disk susceptibility tests, 4th ed, M2-T4. Villanova, Pa: National Committee for Clinical Laboratory Standards 1988. Christensen GD, Simpson WA, Bisno AL, Beachey EH. Adherence of slime-producing strains of Staphylococcus epidermidis to smooth surfaces. Infect Immun 1982; 37: 318-326. Lyon BR, May JW, Skurray RA. Analysis of plasmids in nosocomial strains of multiple-antibiotic-resistant Staphylococcus aureus. Antimicrob Agents Chemother 1983; 23: 817-826. Etienne J, Brun Y, El Solh N et al. Characterization of clinically significant isolates of Staphylococcus epidermidis from patients with endocarditis. J Clin Microbial 1988; 26: 6133617. Kernodle DS, Barg NL, Kaiser AB. Low-level colonization of hospitalized patients with methicillin-resistant coagulase-negative staphylococci and emergence of the organisms Antimicrob Agents Chemothev 1988; 32: during surgical antimicrobial prophylaxis. 202-208. Archer GL, Armstrong BC. Alteration of staphylococcal flora in cardiac surgery patients receiving antibiotic prophylaxis. J Infect Dis 1983; 147: 642-649. Houang ET, Marples RR, Weir I, Mourant AJ, de Saxe MJ, Singleton B. Problems in the investigation of an apparent outbreak of coagulase-negative staphylococcal septicaemia following cardiac surgery. J Hosp Infect 1986; 8: 224-232. Lyon BR, Skurray R. Antimicrobial resistance of Staphylococcus aweus: genetic basis. Microbial Rev 1987; 51: 88-134. Peters G. New considerations in the pathogenesis of coagulase-negative staphylococcal foreign body infections. J Antimicrob Chemother 1988; 21 (Suppl. C): 139-148. Davenport DS, Massanari RM, Pfaller MA, Bale MJ, Streed SA, Hierholzer Jr. WJ. Usefulness of a test for slime production as a marker for clinically significant infections with coagulase-negative staphylococci. J Infect Dis 1986; 153: 332-339.
of Hospital
Infection
Investigation endocarditis
(1991)
19, 107-l
14
of nosocomial prosthetic due to antibiotic-resistant
Staphylococcus R. Menzies, Microbiology
epidermidis
D. MacCulloch
Department,
and
Green Lane Hospital, 3, New Zealand
Accepted
valve
for publication
B. Cornere Greenlane
26 j%ly
West,
Auckland
1991
Summary: A reservoir of antibiotic-resistant Staphylococcus epidermidis strains in our cardiac surgery unit appeared to be the source of organisms responsible for three cases of early prosthetic valve endocarditis. Staphylococcus epidermidis isolates recovered from the skin of 13 patients before and after surgery were compared. All were typed by plasmid profile, antimicrobial susceptibility and slime production. The three strains from early prosthetic valve endocarditis resembled the antibiotic-resistant nosocomial strains recovered from the skin of eight patients following surgery and the environment of the operating theatres. These strains expressed resistance to oxacillin, gentamicin, kanamycin and tobramycin and most produced slime, whereas those isolated from the skin of patients at the time of admission were predominantly susceptible to antibiotics and few produced slime. Keywords: Nosocomial cus epidermidis; slime
infection; production.
prosthetic
valve
endocarditis;
Stuphylococ-
Introduction Staphylococcus epidermidis is an important cause of prosthetic heart valve infection.lm3 It is ubiquitous on human skin which is the apparent source of early postoperative infections. l-3 Some evidence indicates that strains of S. epidermidis associated with early prosthetic valve endocarditis may be acquired during hospitalization and that these strains are different from those colonizing patients’ skin at the time of admission to hospital.3B4 Strains present at the time of admission are usually susceptible to antibiotics whereas those cultured from hospital patients following surgery are frequently multiply-resistant.3s4 We investigated three cases of prosthetic valve endocarditis due to multiply antibiotic-resistant S. epidermidis. Each infection occurred within 2 months of surgery at intervals of c. 6 months. Because antimicrobial Correspondence West, Auckland 0195-670t/91/100107+08
to: Rosalie Menzies, 3, New Zealand.
Microbiology
Department,
SOS.OO,O
Green
Lane
Hospital,
0 1991 The Hospital
107
Greenlane
Infection
Society
,
108
R. Menzies
et al.
susceptibility and plasmid profiles indicated that each isolate was a different strain, and a different surgical team was responsible for each heart valve replacement, it seemed unlikely that a common source of infection such as a carrier was involved. Our aim was to determine whether the source of these strains was the patients’ own flora or a reservoir of organisms in the cardiac surgery unit. The method of investigation was a survey of S. epidermidis strains cultured from skin swabs taken from 13 patients on admission to hospital and 7 days after cardiac by-pass surgery. At the same time, the operating theatre environment was sampled for the presence of antibiotic-resistant S. epidermidis during surgical procedures. We felt that a wider investigation could not be justified in the absence of an outbreak of infection. Materials
and
methods
Patients for cardiac surgery were admitted to the general ward of the cardiac surgery unit no earlier than the day before surgery. After surgery they were transferred to an Intensive Care Unit for at least 24 h before transfer to a special care room in the general ward. At discharge they were resident in the area of the general ward which acted as a reception area for incoming patients. An average of 910 cardiac by-pass operations are carried out each year and 220 of these are heart valve replacements, with prosthetic valves. Since 1977 patients for cardiac by-pass surgery have received antibiotic prophylaxis with cephradine (E. R. Squibb & Sons), 1 g iv. on anaesthetic induction, 1 g i.v. at the start of by-pass surgery and 1 g i.v. on return to the Intensive Care Unit. This was followed by 500 mg q 6 h p.o. or i.v. for 4 days. The night before surgery patients were shaved and had a total body wash with 4% chlorhexidine gluconate. Preoperatively the body wash was repeated and 2.5% tincture of iodine was applied to the incision site. In theatre the iodine tincture application was repeated before surgery, and following surgery the wound was covered with a sterile dry dressing. Between 1987 and 1989 three patients developed early prosthetic valve endocarditis. All patients recovered; two after replacement of the infected prosthetic material and antibiotic treatment, and the third after treatment with antibiotics only. Their details are summarized in Table I.
Patient skin survey The anterior nares and skin over the sternum of 13 patients admitted to hospital between 18 January 1989 and 29 May 1989 were swabbed on admission to hospital and 7 days after cardiac by-pass surgery; heart valve replacement (six patients), coronary artery grafting (six patients) and repair of aneurysm (one patient). The first patient available each week was swabbed. All patients were from the elective surgery waiting list, 18 to 65 years of age and had not received antibiotics or been in hospital within the preceding 4 weeks. The nose was swabbed with a dry swab and the skin over
P.V.E.
due
the sternum with a swab moistened Culture swab transport system).
109
to S. epidermidis
with
Amies
transport
medium
(Difco
Operating theatres During the patient survey the two operating theatres used for cardiac surgery were sampled for the presence of antibiotic-resistant S. epidermidis strains during surgical procedures. Twenty agar plates, Columbia agar (Gibco) with 5% sheep blood, were placed to detect staphylococci in areas of staff movement. Each plate remained in position for the duration of one surgical procedure, average time 3 h.
Culture methods Each swab of the anterior nares and skin over the sternum was cultured on Columbia agar with 5% sheep blood and mannitol salt agar (Difco), and incubated aerobically at 35°C for 48 h. Staphylococcal colonies with representative morphology from each swab were identified and up to 20 presumptive S. epidermidis isolates were typed by susceptibility to 16 antimicrobial agents, plasmid profile and slime production. Agar plates from operating theatres were incubated at 35°C for 48 h and four staphylococcal colonies from each were identified. All presumptive S. epidermidis isolates were tested for susceptibility to oxacillin, and oxacillin-resistant isolates were typed by susceptibility to 16 antimicrobial agents, plasmid profile and slime production.
Identification
of isolates and strain differentiation
Presumptive identification as S. epidermidis was made by Gram’s stain, catalase production and failure to produce acid from trehalose and mannitol.5,6 Identification of probable nosocomial strains and isolates from prosthetic valve endocarditis as S. epidermidis was confirmed by API Staph (France). Strains were typed by antimicrobial susceptibility pattern, plasmid profile and slime production. 2,3,7 Susceptibility to 16 antimicrobial agents was determined by the disc diffusion method on Mueller-Hinton agar (Gibco) as recommended by the National Committee * for Clinical Laboratory Standards.’ Antibiotics tested were penicillin, oxacillin, gentamicin, kanamycin, tobramycin, amikacin, streptomycin, erythromycin, clindamycin, chloramphenicol, tetracycline, rifampicin, vancomycin, trimethoprim/sulphamethoxazole, neomycin and fusidic acid. Slime production was detected by a tube method.’ Plasmid DNA was isolated and separated by a rapid method adapted for S. epidermidis.‘O Isolates were grown on Columbia agar with 5% sheep blood instead of in L broth and bacterial cells were suspended in 10 ml of sterile distilled water to an opacity approximately equal to a number 5 McFarland standard. Lysostaphin (Sigma) was used at double the concentration suggested for
Staphylococcus aweus.
110
R. Menzies
et al.
Results
Characteristics of strains from prosthetic valve endocarditis These strains all had different plasmid and antimicrobial susceptibility profiles. Antimicrobial resistances are shown in Table I. Resistance to penicillin, oxacillin, gentamicin, kanamycin and tobramycin and slime production were common to all strains. Distribution of strains from patient skin swabs and operating theatres Antimicrobial susceptibility, slime production and plasmid profile separated the 215 S. epidermidis isolates obtained from skin swabs into 92 different strains; 49 strains were recovered on admission only, 39 after surgery only and four both on admission and after surgery. The number of strains isolated from each patient at any one time varied from one to eight (average four). Colonization of the anterior nares and skin over the sternum by the same strain was more prevalent after surgery (eight patients) than before surgery (two patients). Identical S. epidermidis strains from postoperative skin swabs, not found on admission swabs, from more than one patient or from one patient and an operating theatre were considered to be nosocomial in origin. Four nosocomial strains were isolated from postoperative swabs from eight
Table
Sex
I. Summary
Age (years)
of the clinical
history and microbiological prosthetic valve endocarditis
Operation
Postoperative
findings
infection
Onset
Treatment
51
21.9.87 Mitral valve replaced with prosthetic valve
2 months
Vancomycin + rifampicin Replacement of infected valve
Male
50
8 days
Male
27
1.7.88 Aortic valve + ascending aorta replaced with prosthetic valve + dacron graft 4.4.89 Porcine mitral valve replaced with prosthetic valve
Vancomycin + gentamicin + rifampicin Replacement of infected valve + graft Vancomycin + rifampicin + flucloxacillin SClindamycin + rifampicin + fusidic acid Discharged on ciprofloxacin
* A, Amikacin; C, chloramphenicol; Cc, clindamycin; oxacillin; P, penicillin; Sxt, trimethoprim/sulphamethoxazole; $ Changed to this regimen because of neutropenia.
G, gentamicin; To,
the three
patients
with
Microbiological results
Male
7 days
for
S. epidermidis from blood cultures + prosthetic valve. Resistant to P,Ox,G,K,To,N,Sxt* S. epidermidis from blood cultures + prosthetic valve + graft. Resistant to P,Ox,G,K,To,Cc,C,Sxt* S. epidermidis from blood cultures. Resistant to P,Ox,G,K,To,A,N,C,Sxt*
K, kanamycin; tobramycin.
N, neomycin;
Ox,
P.V.E. due
to
111
S. epidermidis
patients and one operating theatre. One strain was isolated from three patients and two strains each from one pair of patients. None of these patients were resident in the same area of the hospital at the same time. Operating theatres yielded one strain which was identical to an isolate recovered from a patient discharged from hospital 3 months previously. Antimicrobial susceptibility of strains from patient survey swabs and operating theatres Antimicrobial susceptibility tests of S. epidermidis strains isolated on admission and after surgery showed a very highly significant difference in antimicrobial susceptibility (Table II), P
Table II. epidermidis
Antimicrobial susceptibility of Staphylococcus isolated from patient skin swabs (13 patients, 215 isolates, 92 strains)
Antimicrobial agent
No.
of patients with strains susceptible
On admission (53 strains) Penicillin Oxacillin Erythromycin Tetracycline Clindamycin Kanamycin Gentamicin, tobramycin Cotrimoxazole Streptomycin Chloramphenicol Fusidic acid, Rifampicin Vancomycin
all
After surgery (43 strains)
5 13 10 12 12 13
0 1 6 6 8 3
13 13 13 13 13 13 13
4 6 8 9 11 12 13
neomycin,
amikacin
112
R. Menzies Table III. epidermidis
Nosocomial
et al.
Antimicrobial resistance of nosocomial strains isolatedpostoperativelyfrompatient and operating theatres strain
Staphylococcus skin swabs
Source
Antimicrobial resistance
.2
3 patients 2 patients
3
2 patients
4
1 patient + operating theatre
P,Ox,G,K,To,N* P,Ox,G,K,To, S,E.T,Sxt* P,OX,G,K, To.A.N.C,Sxt* P,Ox;G;K;Tb,N*
1
* A, Amikacin; C, chloramphenicol; E, erythromycin; G, gentamicin; K, kanamycin; N, neomycin; Ox, oxacillin; P, penicillin; S, streptomycin; Sxt, trimethoprim/sulphamethoxazole; T, tetracycline; To, tobramycin. Additional note: Isolates obtained on admission were all sensitive to: P, Ox, FG, K, To.
Slime production Table There strains
IV shows the number was a highly significant that produced slime P=O.O09 (X2 test).
of S. epidermidis strains that produced slime. difference in the proportion of S. epidermidis on admission to hospital and after surgery, Discussion
Antimicrobial resistances expressed by S. epidermidis strains recovered from patients’ skin swabs taken after surgery differed significantly from the resistances expressed by strains recovered on admission to the cardiac surgery unit. For most patients an oxacillin-resistant population had to other antibiotics had replaced an oxacillin-susceptible one. Resistance also increased (Table II) and the percentage of strains that produced slime had almost doubled (Table IV). We considered that isolates from admission swabs would give an indication of the characteristics of patients’ normal skin flora. These isolates produced slime at a rate of one in three and never expressed resistance to oxacillin, gentamicin, kanamycin and tobramycin. These findings indicated that patients’ normal flora was an unlikely source of isolates responsible for sporadic cases of early prosthetic valve endocarditis. Some antibiotic-resistant S. epidermidis strains may colonize patients’ skin in very small numbers on admission to hospital and can emerge during cardiac surgery with antimicrobial prophylaxis.“T’2 We did not identify any such strains. The strains termed nosocomial in our study were only isolated after surgery and from more than one patient or from a patient and the environment. Survey patients were not in the same area at the same time, therefore the environment, other patients and staff of the cardiac surgery unit were the most likely sources of these isolates. In one case contamination of the
P.V.E. due
Table IV. Slime production midis strains isolatedfrompatient to hospital S. epidermidis
Total
On admission After surgery Nosocomial Prosthetic valve endocarditis
53 43 4 3
Staphylococcus
by
epiderskin swabs on admission cardiac surgery
no.
No. slime producers
and after
strains
113
to S. epidermidis
18 (34%)
26 (61%) 3 3
environment of the operating theatre with the same strain isolated from a patient discharged from hospital 3 months previously was established. Our results support reports’3”4 that hospital reservoirs of antibiotic-resistant coagulase-negative staphylococci can be the source of antibiotic-resistant strains found on patients’ skin. Our study may have under reported the number and distribution of nosocomial strains in our cardiac surgery unit. Nosocomial strains were characterized by resistance to gentamicin, kanamycin and tobramycin. Genes encoding resistance to these antibiotics have been recorded in staphylococcal populations in Australia, USA and Europe. I5 Such a gene app ears to be established in the nosocomial S. epidermidis population in our cardiac surgery unit and to be expressed by the isolates responsible for early prosthetic valve endocarditis. Most nosocomial strains and all strains that infected prosthetic heart valves produced slime. These observations support the hypothesis that slime production by S. epidermidis protects against host defences and antibiotic activity16 and is associated with infection of prosthetic devices.17 The nosocomial S. epidermidis population exhibited the same characteristics that distinguished isolates responsible for prosthetic valve endocarditis; a variety of strains, slime production, resistance to oxacillin, gentamicin, kanamycin and tobramycin, and appeared to have been the source of these isolates. Archer” has postulated that a reservoir of antibiotic-resistant S. epidermidis strains and their genes can become established on patients’ skin when normal flora is disrupted by antimicrobial prophylaxis. From this reservoir, strains are disseminated by hospital workers to the operating theatre environment where they gain access to prosthetic heart valves during implantation. We have established that a reservoir of antibiotic-resistant S. epidermidis strains was established in our cardiac surgery unit and caused serious infection. The gene pool available for adaptation and survival in such a population would be enormous and available for transfer to Staphylococcus aweus and Streptococcus spp. *’ The recognition of S. epidermidis as an important pathogen in cardiac surgery units means that strategies for the identification and control of reservoirs of strains must be devised if nosocomial S. epidermidis prosthetic valve endocarditis is to be avoided.
114 This study was approved Whitlock for statistical
R. Menzies
et al.
by Green Lane Hospital analysis of results.
Ethics
Committee.
We thank
Dr
R. M.
L.
References 1. Lowv 2. 3. 4. 5. 6.
7.
8.
9.
10.
11.
12.
13. 14.
15. 16. 17.
FD. Hammer SM. Staahvlococcus eaidermidis infections. Ann Intern Med 1983: 99: 834-839. Parisi JT. Coagulase-negative staphylococci and the epidemiological typing of Staphylococcus epidermidis. Microbial Rev 1985: 49: 126-139. Pfaile; MA, Hkrwaldt LA. Laboratory, clinical and epidemiological aspects of coagulase-negative staphylococci. Clin Microbial Rev 1988; 1: 281-299. Archer GL. Molecular epidemiology of multiresistant StaPhvlococcus epidermidis. 7 _ Antimicrob Chemother 1988; 21 (SuppI. C): 133-138. Kloos WE. Schleifer KH. Simnlified scheme for routine identification of human Staphyloco&us species. J Clin M&robiol 1975; 1: 82-88. Stevens DL, Jones C. Use of trehalose-mannitol-phosphatase agar to differentiate Staphylococcus epidermidis and Staphylococcus saprophyticus from other coagulase-negative staphylococci. J Clin Microbial 1984; 20: 977-980. Archer GL, Karchmer AW, Vishniavsky N, Johnston JL. Plasmid-pattern analysis for the differentiation of infecting from non-infecting Staphylococcus epidermidis. J Infect Dis 1984; 149: 9133920. National Committee for Clinical Laboratory Standards.. Performance standards for antimicrobial disk susceptibility tests, 4th ed, M2-T4. Villanova, Pa: National Committee for Clinical Laboratory Standards 1988. Christensen GD, Simpson WA, Bisno AL, Beachey EH. Adherence of slime-producing strains of Staphylococcus epidermidis to smooth surfaces. Infect Immun 1982; 37: 318-326. Lyon BR, May JW, Skurray RA. Analysis of plasmids in nosocomial strains of multiple-antibiotic-resistant Staphylococcus aureus. Antimicrob Agents Chemother 1983; 23: 817-826. Etienne J, Brun Y, El Solh N et al. Characterization of clinically significant isolates of Staphylococcus epidermidis from patients with endocarditis. J Clin Microbial 1988; 26: 6133617. Kernodle DS, Barg NL, Kaiser AB. Low-level colonization of hospitalized patients with methicillin-resistant coagulase-negative staphylococci and emergence of the organisms Antimicrob Agents Chemothev 1988; 32: during surgical antimicrobial prophylaxis. 202-208. Archer GL, Armstrong BC. Alteration of staphylococcal flora in cardiac surgery patients receiving antibiotic prophylaxis. J Infect Dis 1983; 147: 642-649. Houang ET, Marples RR, Weir I, Mourant AJ, de Saxe MJ, Singleton B. Problems in the investigation of an apparent outbreak of coagulase-negative staphylococcal septicaemia following cardiac surgery. J Hosp Infect 1986; 8: 224-232. Lyon BR, Skurray R. Antimicrobial resistance of Staphylococcus aweus: genetic basis. Microbial Rev 1987; 51: 88-134. Peters G. New considerations in the pathogenesis of coagulase-negative staphylococcal foreign body infections. J Antimicrob Chemother 1988; 21 (Suppl. C): 139-148. Davenport DS, Massanari RM, Pfaller MA, Bale MJ, Streed SA, Hierholzer Jr. WJ. Usefulness of a test for slime production as a marker for clinically significant infections with coagulase-negative staphylococci. J Infect Dis 1986; 153: 332-339.