Bums (1992) 18, (4), 332-335
Printed in Great Britain
Predominance of staphylococcal organisms in infections occurring in a burns intensive care unit G. D. Taylor, P. Kibsey, T. Kirkland, E. Burroughs and E. Tredget University
of Alberta Hospitals, Edmonton,
To assessthesites,incidence,and bacteriology of infecfions in intensive care burn patients, a prospective sumey of all admissions to a tertiary care institution burn unitwas car&d out over a 12-month period. One hundred and sixfeenpafienfswereadmitted, 106 with a diagnosisofthermal burns. Fortypatientsdeveloped90 infections.Only fwo deaths occurred, one in a patient with sepsis. In order of frequency, pneumonia, burn infection, UT1 and primay bacteraemia were most common. Sfaphylococcal species accounted fora majority of infecfio+zsat all body sites exceptUT1 (47 per cent of all infections, including II of 14 bacferaemic infections). Staph. aureus sepsis was more common in those carrying the organism on admission. Strain typing of paired admission and subsequent clinical isolates in 19 patients with Staph. aureus sepsis indicated that eight (42 per cent) became infected with a strain they carried on admission. Further red&ions in septic complications of burns in our center would be best directed at sfaphylococcal species, particularly Staph. aureus. Both eradication of carrier state, andprevention ofacquisition of Staph. aureus sfrains could be explored.
Introduction Infection is amajor problem in the treatment of bum patients (Luterman, 1986). As a result of the extensive changes in bum care over the last few decades, including prevention of infection, bum survival has greatly increased (Deitch, 1990). No recent comprehensive survey has been carried out to study the effect of those changes on the infection problem in this group of patients, to allow those providing bum care to plan future steps to reduce further the incidence of sepsis. Consequently, we carried out a l&month survey of all patients admitted to our regional burns center to determine incidence, sites of infection and bacteriology of infections in this population. A preliminary chart review had suggested that the principal pathogen in our unit was Staph. auretls. In addition to confirming this finding we wished to explore further the epidemiology of Staph. aUreW in our bum population.
Methods The University of Alberta Hospital is a IlOO-bed tertiary care institution providing all bum care services to the areas of northern Alberta, northern British Columbia and the Northwest Territories. A surveillance project identifying all 0 1992 Butterworth-Heineman 030%4179/92/040332-04
episodes of nosocomial infection occurring in patients admitted to our Bum Unit was carried out over a l&month period. One of the authors (T.K.) reviewed patients and charts to document infections. All infections were reviewed by a physician (G.T.) prior to inclusion. Criteria established for identification of infection were derived from the Centre for Disease Control definitions for nosocomial infections (Gamer et al., 1988). The Bum Unit is a IO-bed unit providing intensive care services to an average of 160 patients per year. The great majority of patients are admitted to the unit with a diagnosis of burns; a small number are admitted with a non-bum diagnosis such as frostbite or trauma. Patients remain in the bum unit until they no longer require critical care, though they may remain for rehabilitation care depending on bed availability. Routine practices within the bum unit include early surgical wound closure wherein wound excision is begun 2-4 days postburn and wound closure is obtained with autografts, as a priority, or banked cadaveric allografts. Sequential debridement and grafting continues every 4-5 days when possible until wound closure is achieved. Hydrotherapy is not used for our patient population until wound closure has occurred. Cleansing of the bum wound is accomplished by washing with sterile normal saline and an antimicrobial soap (chlorhexidine), followed by rinsing with sterile normal saline to accomplish debridement. Routine surveillance cultures are obtained from bum sites at the time of patient admission to identify colonizing organisms and from the anterior nares to detect Sfaph. aweus colonization. Topical antimicrobial agents including silver sulphadiazine (alone or in combination with chlorhexidine), mafenide acetate, silver nitrate and gentamicin are prescribed by the attending physician dependent on organisms colonizing the bum site and the location of the bum wound. Penicillin G and cloxacillin are given prophylactically immediately preoperatively and postoperatively for a 48-h period of time for surgical procedures involving skin surfaces. There is no endemic methicillin-resistant Staph. aureus (MRSA) within our institution, and during the study period no outbreak of MRSA occurred. To determine whether admission Staph. aureus carrier strains were identical with those responsible for subsequent sepsis, and whether ‘hospital acquired’ Staph. aureus isolates were of a single strain type, all admission colonizing strains
Taylor et al.: Staphyloccocal TableI.
infections in a bums KU
Patient characteristics in 116 burn intensive care unit
patients No. Sex M F Type of injury Burn Chemical Electrical House/campfire Scald Gas/propane fire Other Non-burn Inhalation injury Initial total body surface burn (burn patients only) l-20% 21-40% 41-60% > 60% Initial full thickness burn (burn patients only) O-20% > 20%
2 10 11 31 44 8 10 28 71 20 10 5 95 11
Mean age= 32 years.
Table II. Infections in intensive site
care unit (KU) bum patients by
Infection rate/ 1000 patient days in KU
Infection rate/100 admissions
Burn Pneumonia UTI I’ blood stream” Vein Cutaneous Other resp. tract** Surgical wound Donor sites Gastrointestinal
21 22 18 9 5 5 4 3 2 1
23 24 20 10 6 6 4 3 2 1
18 19 16 8 4 4 3 3 2 1
7.6 8.0 6.6 3.2 1.8 1.8 1.5 1 .l 0.7 0.4
“Primary blood stream infections with no evident source. *“Respiratory tract infections other than pneumonia, e.g. tracheitis and bronchitis.
and later clinical isolates underwent strain typing. In addition, 23 health care workers employed in the bum unit submitted swabs of anterior nares to detect Staph. aureus carriage; isolates from this source were also strain typed. Sfaph. aurezts isolates were typed by three different methods: phage typing, plasmid profiles and chromosomal DNA finger printing. Phage typing was carried out in the Laboratory Center for Disease Control, Ottawa, Canada as a single batch (Blair and Williams, 1961). Plasmid DNA was obtained and typed by the method of Paris and Hecht (1980). Chromosomal genomic DNA was obtained by the lysostaphin SDS method of Pasloske et al. (1988) using repeated phenol extraction and alcohol precipitation. Gibco BRL recommendations for Xbal enzyme digestion were followed.
Results Over the 12 months of the study period, 116 patients were admitted to the bums intensive care unit, 106 with a bum injury and 10 with a non-bum diagnosis. Table1 illustrates the characteristics of the 116 patients. Non-bum diagnoses included frostbite (three), cellulitis (one), facial injury (three), mandibular fracture (two) and laceration (one). Forty of 116 patients (34 per cent) developed 90 infections. Table II shows the site of these infections. Bum infection, pneumonia, urinary tract infections (UTI) and primary bacteraemia accounted for 77 per cent of all infections. There were 14 bacteraemic infections. The source of these bacteraemic infections were: primary 10; pneumonia, three; bum, one; UTI, one. Two patients (1.7 per cent) died during the ICU phase of their hospital stay, one from respiratory failure secondary to inhalation injury and one bum patient with acute renal failure who died as a result of a sudden gastrointestinal haemorrhage while at the same time experiencing a Cadida blood stream infection. Table III demonstrates the organisms isolated from infected patients. Overall, Staph. aurew was the most important pathogen found in 3 7 per cent of all infections. It was also the most important pathogen at every individual site, except ‘primary’ blood stream infections where coagulase-negative staphylococcal strains predominated. Of the Gram-negative bacilli, only Escher&a coli accounted for substantial numbers of infection, principally urinary tract infections. Pseudomonas species were only rarely encountered (4 per cent of all infections). Admission Staph. adieus carrier state was a predictor of
Table III. Organisms isolated from selected site infections All infections Pneumonia Staph. aureus Coag. neg. staph. E. coli Pseudomonas sp. Other enteric Gram-negative Candida sp. Other Total
33 9 13 4 7 7 16
37 10 14 4 8 8 19
7 0 3 1 5 0 6
15 0 2 1 0 1 2
0 5 2 0 4 4
6 5 1 0 0 1 2
Burns (1992) Vol. M/No. 4
Table IV. Effect of admission carrier state on subsequent Staph. uureus sepsis by site of infection
Staph. aureus admission cultures’ Positive (36) Negative (63) Total (99)
All sites * ??
‘No admission cultures in 17 patients. * P= 0.03 by Chi square, Odds ratio = 3.025.95 ??
per cent confidence limits= 1.06 and 8.6.
subsequent development of Staph. aureus sepsis (Table IV): 31 per cent of carriers but only 13 per cent of non-carriers later developed Staph. aureus sepsis. Strain typing of paired admission and subsequent clinical isolates in the 11 carriers who later developed Staph. atlreus sepsis showed eight (42 per cent of all patients developing Staph. aureus sepsis) to be infected by a strain identical with their admission isolate. Strain typing of the ‘hospital acquired isolates (i.e. eight patients cultured negative on admission who later developed Staph. aureus infection and three whose admission strain and later clinical isolate was discordant) showed multiple different strains with no single strain predominating. Likewise typing of eight isolates from the 23 health care workers who were cultured showed multiple different strains (between two and seven depending on the strain typing method). Genomal DNA typing proved to be the most reliable technique; only 35 per cent of the typed isolates contained plasmids and 27 per cent of the isolates were non-typeable by phage methodology.
Discussion Previous comprehensive surveys for all types of infections have not been recently published, however, burn wound infection, pneumonia and UT1 have been recognized as collectively accounting for most infections (Deitch, 1990). Our results are very much in keeping with this experience: 67 per cent of all our infections occurred at one of these three sites. Kagan et al. (1985), in a survey of bum patients from 1981 to 1983, found a bum wound infection rate of O.O9/patient-week. Calculated in this manner our bum wound infection rate was O.O5/patient-week. However, different patient populations, definitions of infection and intensity of survey techniques make interhospital study comparisons problematical. Primary bacteraemia, generally regarded as largely due to intravascular catheters, has been recognized as a potential problem in bum patients (Franceschi et al., 1989). Our 8 per cent primary bacteraemia rate is, however, essentially the same as for non-bum patients (Maki et al., 1977), and below the rates previously reported in other bum populations (Samaoondar et al., 1985). An important finding of our study is an apparent change in the microbiology of hospital infections in bum patients away from the traditionally important Gram-negative rods to Gram-positive cocci, especially Staph. aureus and coagulase-negative staphylococci. Similar changes are occurring in nosocomial infections in non-bum patients (Stillman et al., 1987). Overall, Staph. uureus was present in 37 per cent and coagulase-negative staphylococci in 10 per cent of our infections. Since determining which organisms are responsible for infections at contaminated sites, such as bums, or when contamination occurs during specimen collections, such as pneumonia, is difficult, it may be more
useful to determine the bacteriology of blood stream infections. Staphylococcus species accounted for 11/14 (79 per cent) of bacteraemic infection in our population. Gramnegative rod infections accounted for only one of the bacteraemic infections. A striking feature of the bacteriology of our bum population in this study period is the relative absence of Pseudomonas infections - present in only 4 per cent of our 90 infections, and none of the bacteraemic infections. This differs markedly from other published series. McManus et al. (1985) in a review of 25 years of experience (1959-81) found Pseudomonas to account for 18 per cent of bacteraemic infections (and Gram-negative bacilli as a group 50 per cent). In another survey from the same institution, Pseudomonus was present in 51 per cent of respiratory tract specimens over a l-year period (Pruitt et al., 1983). This apparent change in bacteriology of bum infections may be attributable to the increased use of intravascular catheters, thereby increasing the frequency of bacteraemic infections from staphylococcal species, or possibly our policy of avoiding hydrotherapy until such time as wounds are closed, thereby reducing cutaneous contamination by water-living organisms, such as Pselldomonas species. The change in the bacteriology of sepsis in bum patients may require a redirection of research or preventive efforts away from Gram-negative bacteria and toward staphylococcal species. Admission cultures for the presence of Staph. aureus, and subsequent strain typing of these and clinical isolates, was able to elucidate the source of the Staph. aureus infections. Of the 19 patients who developed Staph. uureus sepsis, 11 were culture positive on admission. Eight of 11 paired admission and subsequent clinical isolates were identical, implying that infection with endogenous flora was responsible for 42 per cent of Staph. uurues sepsis. Imaginative approaches, such as eradication of Staph. aureus carrier states, successfully applied to other populations (Yu et al., 1986) are potential approaches for this group. For the remaining patients with ‘hospital acquired Sfuph. aureus infections, it was reassurring that transmission of a single strain within the unit was not occurring. On the other hand it is disconcerting to find that, in a group of patients to whom barrier isolation practices are rigorously applied, hospital acquisition of flora still frequently occurs. For this group, surveillance cultures and eradication of the carrier state when it occurs may be an option to be examined.
Acknowledgements This study was supported by a grant from the Firefighters’ Bum Trust Fund. We wish to thank Dr M. Joffres for assistance with statistical analysis. This data has previously been presented in part at the 3rd Decennial Conference on Nosocomial Infections, Atlanta, Georgia, July 1990.
Taylor et al.: Staphyloccocal
infections in a bums ICU
References Blair J. E. and Williams R. E. 0. (1961) Phage typing of Staphylococci. Bull WHO 24, 771. Deitch E. A. (1990) The management of bums. N. Engl. 1, Med. 323,1249. Franceschi D., Gerding R. L., Phillips G. et al. (1989) Risk factors associated with intravascular catheter infection in burned patients: a prospective, randomized study. 1, Trauma 29, 811. Gamer J. S., Jarvis W. R., Emori T. G. et al. (1988) CDC definitions for nosocomial infections. Am 1. Infect. Confrol 16, 128. Kagan R. J., Matsuda, T., Hanumadass M. et al. (1985) Serious wound infections in burned patients. Surgery 98, 640. Luterman A., Dacso C. C. and Curreri P. W. (1986) Infections in burn patients. Am. J Med. 81, 45. Maki D. G., Weise C. E. and Sarafin H. W. (1977) A semiquantitative culture method for identifying intravenous-catheterrelated infection. N. Etzgl.J Med. 296, 1305. McManus A. T., Mason A. D. Jr, McManus W. F. et al. (1985) Twenty-five year review of Pseudomonas aeruginosa bacteremia in a burn center. Eur. J Chn. Microbial. 4, 219. Parisi J. T. and Hecht D. W. (1980) Plasmid profiles in epidemiologic studies of infections by Staphylococcus epidwmidis. J Infect. Dis. 141, 637.
Pasloske B. L., Joffe A. M., Sun Q. et al. (1988) Serial isolates of Psetiomonas aeraginosa from a cystic fibrosis patient have identical pilin sequences. Infect. Immunol. 56, 665. Pruitt B. A. Jr Lindberg R. B., McManus W. F. et al (1983) Current approach to prevention and treatment of Pseudomonas aeruginosa infections in burned patients. Rev. Infecf. Dis. 5, (Suppl.), S889. Samaoondar W., Freeman J. B., Coultish I. et al. (1985) Colonization of intravascular catheters in the intensive care unit. Am. 1, Surg. 149, 730. Stillman R. I., Wenzel R. P. and Donowitz L. C. (1987) Emergence of coagulase negative staphylococci as major nosocomial blood stream pathogens. Infect. Control 8, 108. Yu V. L., Goetz A., Wagener M. et al. (1986) Staphylococcus aurew nasal carriage and infection of patients on hemodialysis. Efficacy of antibiotic prophylaxis. N. Engl. J Med. 315, 91.
Correspondence should be addressed to: Dr G. D. Taylor, 2E4.11 Walter Mackenzie Center, University of Alberta, Edmonton, Alberta, Canada T6G 2B7.