Assessment of the effectiveness of body substance precautions as the infection control system of a large teaching hospital

Assessment of the effectiveness of body substance precautions as the infection control system of a large teaching hospital Ian B. R. Duncan, MD, FACP, FRCP(C) Claudette Batchelor, RN, BN, CIC Toronto, Ontario, Canada

Background: Body substance precautions was the name given to the body substance isolation-based infection control system that was introduced in January 1990 at a Canadian university hospital with 650 acute care beds and 570 long-term care beds. When the body substance precautions system was begun, traditional category-specific isolation was discontinued. Methods: After 2 years, we reviewed the incidence of several types of nosocomial infections and the frequency of isolation of hospital strains of bacteria before and after the introduction of body substance precautions to find out whether this system was as effective as the previous system of infection control. Results: Most nosocomial infections did not increase. There was another likely cause for the only one that did. For many years, we had isolated patients infected or colonized by hospital bacteria, limiting their spread throughout the institution. Body substance precautions proved equally effective in doing this. ConcZusion: Our results to date therefore indicate that the body substance precautions system was as successful as category-specific isolation used with other standard infection control techniques in maintaining low rates of nosocomial infections and in controlling the dissemination of hospital strains of bacteria in our institution. Body substance precautions provided a satisfactory alternative to universal precautions and traditional isolation categories for the protection of health care workers against the risk of infection by blood-borne viruses. (AJIC AM J INFECT CONTROL 1993;21:302-9.)

Body substance isolation (BSI) has been advocated’ in recent years for hospitals as an alternative infection control system to the more traditional methods of isolation20 3 of infected or colonized patients. The arguments in favor of BSI are persuasive, but they are based more on commonsense extension of the rationale for traditional isolation than on detailed assessment of the end results after use in practice. Put simply, BSI is a

From the Infection Sunnybrook Health Ontario, Canada.

Control Science

Division, Microbiology Department, Centre, University of Toronto, Toronto,

Reprint Control Science

requests: Claudette Batchelor, RN, BN, CIC, Nurse Manager, Microbiology Dept., Sunnybrook Centre, Toronto, Ontario, M4N 3M5 Canada.

Copyright Control,

0 1993 Inc.

0196-6553193 302

by the Association

$01 .OO + 0.10

for Practitioners

17/46/46676

Infection Health in Infection

method of caring for patients in hospitals in which the techniques of handwashing, gloving, and wearing of gowns, long used for patients placed in traditional isolation, are employed for all patients whenever the nurse or other health care worker (HCW) expects to be in contact with the blood or other body fluids of the patient. Logically, BSI would seem to have two distinct merits. First, it should reduce the risk of cross-infection of patients with exogenous organisms transferred to them from other patients or from HCWs. At the same time, it should reduce the risk of transmission of blood-borne viruses from patients to HCWs. In our hospital, we decided to employ a BSI-type system rather than universal precautions when, in January 1990, we introduced a hospitalwide system for the protection of staff from the risk of infection with blood-borne viruses. At that time, we also phased out category-specific isolation. We called our new system body substance

AJIC Volume

21, Number

Duncan

6

precatltions (BSP). We report here our infection control results for the first 2 years of BSP and compare them with the last 2 years of our previous system of traditional isolation. Figures were available from past years on the numbers of nosocomial infections detected by our ongoing infection control surveillance program. Data were collected in the same way for the first 2 years of implementation of BSP, and the figures for these 2 years and for the previous 2 years were compared. Some of these nosocomial infections would have been endogenous and others exogenous in origin, and either traditional isolation methods or BSI would only have influenced those with an exogenous source. The figures for infections before and after the institution of a BSIbased system still provide a valid comparison between the two infection control systems because there were no major differences in the patients before and after BSP that would have changed the proportion of endogenous infections. Similarly, data were available from past years on the incidence of “hospital strains” of bacteria; these rates were compared with those in the first 2 years of BSP. Most of the hospital bacteria were gentamicin- and tobramycin-resistant strains of aerobic gram-negative bacilli. We had isolated all patients with aminoglycoside-resistant gram-negative strains since the 1970s when isolation proved successful in halting their spread in the hospita14, 5 Culture practices did not change in any way during the four years of the study. The bacteria were cultured from clinical specimens sent to the laboratory on the instructions of the same medical staff and the same ICPs. No special studies that might have altered the numbers of bacterial isolates took place during the period. The total hospital bacteria cultured and the nosocomial infections diagnosed on the basis of culture could therefore be compared for the 2-year periods before and after the introduction of BSP. We performed the comparison to determine whether BSP was as effective as traditional isolation in preventing the spread of specific nosocomial pathogens and in limiting the total number of nosocomial infections in the institution. We needed the answer for our own infection control planning. We report the information here because published assessments of the use of BSI in hospitals have tended to be process rather than outcome orientated6 or have dealt with changes in the incidence of nosocomial pathogens only after the introduction of BSI,’ rather than comparing the effectiveness of BSI with that of traditional isola-

and Batchelor

303

tion in preventing the dissemination of nosocomial pathogens throughout the hospital. MElwaDs Husplkl

setting

Sunnybrook Health Science Centre is a teaching hospital at the University of Toronto Medical School. It is located in a city of 2 million people. It has two divisions in adjacent but separate buildings. There is a 650-bed acute care tertiary referral division, which has all major medical and surgical services except pediatrics and obstetrics. This division includes the regional trauma unit and a large cancer treatment center. The other division is a 570-bed long-term care facility. The types of patients treated in the hospital during the 4year period did not change, but it seemed to us that there was some increase in the severity of illness of those admitted. We attributed this to hospital bed shortages in our catchment area, which led to the admission of only the more severely ill of those referred to the hospital. This was an impression only; we did not confirm it by measuring indexes of severity of illness at admission. The hospital has a nursing staff of 1200, a full-time medical staff of 200, and 140 residents. Most of the residents rotate through several of the Toronto teaching hospitals during their training, but almost all the nursing staff and the full-time medical staff work only at this hospital. The head of the infection control service of the hospital and of the department of microbiology is a physician. The ICPs are all registered nurses working full-time in infection control. Until 1989, there were two ICPs; since 1990, there have been three. The head of the infection control service reports to senior hospital administration and receives general policy direction from a multidisciplinary infection control committee. The service is responsible for infection control in both acute care and long-term care portions of the hospital and the diagnostic microbiology laboratory also serves both areas. Full surveillance of nosocomial infections is carried out by the ICPs by daily visits to the admitting and emergency departments and to all the acute care units, by daily review of microbiology laboratory reports, and by weekly visits to long-term care units. Centers for Disease Control and Prevention (CDC) definitions8 of nosocomial infections are used. The ICPs carry out a major program of infection control education for nurses and other care givers; this was done both before and after the institution of BSP. The program includes formal teaching programs

304

AJIC December 1993

Duncan and Batchelor

for all staff, special orientation of new staff, and much informal teaching in the course of daily visits to nursing units and when responding to specific questions asked by staff members. At nights and on weekends, coverage for urgent infection control problems is provided by the physician on call for the microbiology diagnostic service. The hospital has a drug formulary system, and the aminoglycoside specified in the formulary for general use during the study period was tobramycin.

Throughout the 1980s infection control measures to prevent the dissemination of pathogens were based on the CDC system3 of categoryspecific isolation. All the recommended categories were employed except drainage/secretion precautions. In addition, specific nursing precautions were enforced in cases where a multiple drugresistant bacterium of a strain known to be troublesome in the hospital was cultured from a urine sample or from sputum from a patient with a tracheostomy or mechanical ventilator-y support. These precautions included the use of gloves or gowns when they were considered necessary and nursing the patient in a private room (if one was available) when the organism had been isolated from sputum. This was essentially a form of isolation applied to patients after the diagnosis of infection had been made. The isolation methods employed for patients with the nosocomial pathogens listed in Table 1 varied. Patients with Clostridium dificile were nursed in enteric isolation, those with a soft-tissue infection or colonization were nursed in contact isolation, and those with an aminoglycoside-resistant gram-negative bacillus in urine or sputum were nursed with the special precautions mentioned.

produced a handbook that would be an educational tool for staff training and a reference manual for hospital staff members in their work settings. When BSP was actually implemented, the handbook was incorporated into the infection control manual distributed to all nursing units. The task force also made the decision to call the new system BSP, rather than BSI, because they believed that the term “precautions” would be better accepted than “isolation” by our patients and staff. We believe that this was a correct decision, and the term BSP continues to be used in our hospital. After completing the handbook, the task force handed over further implementation of BSP to the hospital epidemiologist and the ICPs, who worked closely with the infection control committee. From September until December 1989, an intensive educational program was developed to teach everyone working in the hospital about BSP. The handbook, plus audiovisual materials based on it, were widely used in teaching. By December 1989, the entire hospital staff had been instructed in BSP. BSP was introduced and the former categoryspecific isolation system was discontinued at the beginning of January 1990. BSP has remained the infection control system of the hospital since that time. All new staff members and trainees are taught BSP as part of their orientation. In addition, the ICPs provide regular reinforcement teaching to all members of the hospital staff. Seminars are given in any area where, on their regular daily unit visits, the ICPs observe inadequate staff performance of BSP. Since BSP started in January 1990, patients with C. dificile and aminoglycoside-resistant aerobic gram-negative bacilli have not been transferred to an isolation room but have been nursed with our standard BSP techniques on the unit in which they were located.

BSP In 1990 and 1991

Bacteriologic

BSP was introduced in the hospital in January 1990. This introduction was preceded by a lengthy preparatory period. Early in 1988, it was decided that either universal precautions or a BSI-based system would be introduced as soon as preparations could be completed. A BSI type system was selected because it appeared both to protect hospital staff members against blood-borne pathogens and to reduce nosocomial infection among patients. The preparatory period lasted approximately 18 months. First, a multidisciplinary task force including infection control was set up to get the initial preparations under way. The task force

All specimens were routine specimens sent for clinical purposes. Standard cultural methods’ were employed with one exception. This exception was that we cultured all stools sent for C. difficile testing on the cefotoxin- cycloserine fructose agar selective medium,‘O as well as referring them to another laboratory for toxin testing. The data on C. dijjkile in the Results section are drawn from these cultural studies.

lsolatlon

methods

used until end of 1989

Statlstical

methods

analysis

The x2 test of individual proportions with the criterion of significance for p set at 0.05 was used

AJIC Volume 21, Number 6

Table

Duncan

1. Numbers

of hospital

bacteria

isolated

from acute care patients

and Batchelor

305

before and after the introduction

of BSP With BSP

Before BSP Bacterium

No. of strains

GR and TR Klebsiella Other TR Enterobacteriaceae TR P. aeruginosa

16 67 86

C. difficile GR, Gentamicln-resistant; Total patient population *Calculated by x2 test.

Table

131 Tf7, tobramycin-resistant. before BSP was 33,751 and with BSP 35,426.

2. Nosocomial

infections

No. of strelns

Rate 0.47

18

1.99

94

2.55 3.88

87 102

Rates expressed

in acute care patients

Infection

before and after the introduction

before

of BSP with BSP

Rate*

No. of cases

Patlent total

Rate*

pt

332 90 135 171

33,751 33,751 33,751 9,101

9.80 2.67 4.00 1.88

415 108 131 181

35,426 35,426 35,426 10,128

11.71 3.05 3.70 1.79

0.019 0.385 0.562 0.681

per 1000 patients,

of nosoeomial and mfter BSP

0.976 0.081 0.868 0.027

Patient total

except

for surgical

wound

to compare the number of acute care patients with a nosocomial infection or harboring a hospital strain of bacterium per total acute care patients for the 2 years before with that for the first 2 years after BSP was introduced. Because there were so few admissions and discharges each year in the long-term care part of the hospital, patient activity there was measured in patient days and the infection rates in terms of patient days were compared before and after BSP was introduced, according to a method specified by Kleinbaum and associates. ’ ’ RESULTS comparlsort

0.51 2.65 2.46 2.88

No. of cases

Primary bloodstream intravenous site Conjunctivitis Surgical wounds in clean cases *Infection rates are expressed tBy x2 test.

P

per 1000 patients.

Before BSP Type of nosocomial

Rate

infection

rates

We limited our comparison of nosocomial infections detected by our routine surveillance program before and after the introduction of BSP to those infections for which we are certain that our surveillance criteria did not alter during the period of the comparison. This limitation was essential because during these years we had been upgrading some of our surveillance methods to comply with the 1988 CDC definitions.’ The only skin and soft-tissue infections that we used in the comparison were therefore skin infections at intravenous sites. Similarly, conjunctivitis was the

rates, which are given as percentages

of clean surgical

cases

only infection used in the eye, ear, nose, throat, and mouth infection group. The comparative data for nosocomial infections in the acute care section of the hospital are shown in Table 2 for the first 2 years that BSP was in operation and for the preceding 2 years. Surgical wound infection data are given only for the clean category12 of wounds. The data for nosocomial infections in long-term care patients are given in Table 3. Three of the four types of nosocomial infections in acute care patients and both types in long-term care patients showed no statistically significant change after BSP was introduced. Primary bloodstream infections in acute care, patients showed an increase that was found to be significant by the x2 test. This increase is discussed later. We also reviewed our surveillance data for evidence of cross-infection of other patients from patients admitted with infectious diseases already present. Patients with communicable diseases on admission were placed immediately in the appropriate form of isolation in 1988 and 1989 and were admitted to an ordinary nursing area and nursed with the BSP system in 1990 and 1991. Most patients admitted to our hospital with a communicable infection are patients with AIDS or hepatitis B. These were not analyzed in detail because cases of cross-infection with these agents would be

AJIC

366

Duncan and Batchelor

Table

3. Nosocomial

infections

December 1993 in long-term

care patients

before and after the introduction

Before BSP Type of nosocomial infection Primary bloodstream Conjunctivitis

No. of cases 28 80

Rate per 1000 days 0.086 0.245

of BSP

With BSP No. of cases 14 61

Rate per 1000 days

2 statistic

0.038 0.166

p (two-tailed)

2.54 2.29

0.011 0.022

Total patient days before BSP, 326,456; with BSP, 366,572. Z statistic and p value obtained by method of Kleinbaum.”

so long in appearing that they would be unlikely to be known to us. The few other infections were reviewed. Communicable gastrointestinal infections, such as Salmonella and Campylobacter infections, ranged in number from 12 to 29 per year and other communicable infections, such as meningitis, ranged from 14 to 23. The numbers are small because the hospital does not admit pediatric patients. In no case either before or after the introduction of BSP was there evidence of any cross-infection. BSP thus functioned as well as the traditional isolation system in preventing any spread in the hospital of these community-acquired infections. Comparison of incidence before and after BSP

of hospital

bacteria

The incidences of hospital bacteria of the types that have been troublesome in our institution in the past are shown in Table 1. Before BSP was introduced, patients harboring any of these organisms were moved to isolation rooms. In 1990 and 1991 they were not moved but were instead treated like all other patients with BSP methods. The rates for these organisms before and after BSP was begun are shown in Table 1. There were no increases in isolates among any of the hospital bacteria that we studied. The three groups of tobramycin-resistant bacteria showed no significant difference in numbers before and after BSP, as determined by x2 tests. C. difjficile numbers were significantly less with BSP.

types 22 and 68, had spread widely in the hospital by 1977; this led us to isolate all patients infected or colonized by aminoglycoside-resistant Klebsiella (or by other aerobic gram-negative bacilli) from 1978 onward. As shown in Fig. 1, which gives numbers of cases in all areas of the hospital, isolation led to a marked reduction in the numbers of these strains. There were 58 resistant infections among 13,462 patients at risk in 1977 and 12 resistant infections among 13,723 patients in 197 8 after the introduction of isolation at the beginning of that year. This decrease was statistically highly significant (p < 0.00001). Fig. 1 shows that continued use of category-based isolation maintained the low numbers of infections with .&theseorganisms until the end of 1989 and that there was no resurgence of the resistant strains after the change was made in 1990 from traditional isolation to BSP, In the 4 years 1989 to 199 1, all but one of the resistant strains were isolated from acute care patients; Table 1 shows that there was no statistically significant difference in the numbers from acute care patients for the 2-year periods before and after BSP was introduced. Tobramycin-resistant genera. The numbers

Entbreobacteriaceae

of other

of isolates of any one genus were small as had been the case ever since the policy of isolating aminoglycoside-resistant gramnegative rods was instituted in 1978. The numbers of all genera other than Klebsiella with tobramytin-resistant strains (Citvobacter, Enterobacter, Escherichia,

Morganella,

Proteus,

Providencia,

Gentamicin- and tobramycin-resistant Mebsiella. Klebsiella species are described separately from

and Sewatia) were therefore combined sis. They are shown in Table 1.

complete other Enterobacteriaceae because records have been kept of all aminoglycosideresistant strains of this species isolated from both acute care and long-term care patients since plasmid-determined resistance to gentamicin appeared in Klebsiella in our hospital in 1974.4 The resistance was caused by an adenylating enzyme13 that inactivated gentamicin and tobramycin to an equal extent. Two serotypes of KlebsielZa, capsular

The figures for tobramycin-resistant P. aeruginosa were calculated separately from those for tobramycin-resistant Enterobacteriaceae because resistant P. aeruginosa remained more frequent than other resistant gram-negative organisms in our institution despite isolation precautions. Others14 have also found that drug-resistant strains of P. aeruginosa are more difficult to control by

Tobramycin-resistant

Pseudomonas

for analyaeruginosa.

AJlC Volume

21,

Number

Duncan and Batchelor

6

ISOLATION

307

BSP

Fig. 1. Numbers of acute and long-term care patients per year with gentamicin- and tobramycin-resistant Klebsieh each year from 1972 to 1991. Category-specific isolation was begun in January 1978 and BSP was begun in January 1990.

isolation bacteria.

measures than other resistant hospital The data are presented in Table 1. C. difncik Although some cases of infection with C. difFciZe in hospitals arise as a result of antibiotic therapy suppressing other bowel bacteria and allowing the patient’s own colonizing strains of C. dificile to multiply, it is accepted that this organism also commonly spreads in hospital by cross-infection from patient to patient.‘5-‘7 Cases of C. dificile enteritis in hospitals may therefore be of either endogenous or exogenous origin. However, it seemed unlikely to us that there would be major changes from year to year in the proportion of our patients harboring C. difitile on admission and those cross-infected after admission. We therefore compared the total numbers of patients from whose stools C. dificile was cultured for the 2 years before and the 2 years after the introduction of BSP, on the assumption that endogenous cases would have remained relatively constant in number and major changes would reflect the amount of exogenous infection by hospital strains. The figures are shown in Table 1. When BSP was in operation, there was a significant reduction in patients with C. difficile in their stools from the previous 2 years. DISCUSStON

Until the end of 1989, we followed the CDC guidelines for category-based isolation as a major part of our infection control system. We used

isolation successfully to control the spread of hospital strains of bacteria. We decided to change to BSP in January 1990. Because the ICPs, in the course of their regular surveillance activities, carefully monitored the performance of HCWs throughout the hospital, we are confident that BSP was in fact consistently carried out. Our preliminary assessment of our results after 2 years of BSP was intended to show us whether a BSI-based system such as BSP was as effective as our previous isolation system in controlling nosocomial infections in general and in preventing the dissemination of specific hospital bacteria. We believe that the results presented in this article show that BSP is at least as satisfactory in accomplishing our objectives as our previous infection control methods. There was no statistically significant increase in five of the six categories of nosocomial infections that we compared. The sixth, primary bloodstream infections in acute care patients, showed an increase that we believe but did not formally prove, was caused by an increase in severity of illness, which would have led to a higher proportion of bacteremias of endogenous origin in the patients admitted in the second 2 years of the study. There was no increase in the numbers of any of the hospital bacteria that we followed, and the numbers of C. di#iciZe actually declined. We would like a longer period of observation before trying to assess whether BSP is better than-as distinct from as good as - stan-

308

Duncan and Batchelor

dard isolation methods. In this article, we give the surveillance figures only for those types of nosocomial infections for which we were confident that there was no change in surveillance methods during the 4-year period. The numbers of other nosocomial infections did not appear to change after the introduction of BSP, but we did not consider that these data had been collected in such a way that a formal statistical comparison could be made. The finding that BSP was as good as previous methods at limiting the spread of aminoglycoside-resistant gram-negative bacilli was particularly satisfying because we had been successful in the past with category-based isolation in preventing the spread of these organisms, maintaining the usefulness of tobramycin and gentamicin. One of the major differences in our BSP system from BSI as originally described’ was our retention of isolation for patients with methicillinresistant Staphylococcus aweus (MRSA). We continued to isolate patients with MRSA because of our concern that staphylococcal infections may be spread in part by the airborne route, as well as by the hands of care givers. The older literature, from the time when multidrug-resistant hospital staphylococci were the principal organisms of nosocomial infection, presents clear evidence of the spread of staphylococci from infected patients into the hospital air. l8 We are aware that there is considerable controversy”, ‘O in infection control circles regarding the amount of effort that should be devoted to containing and controlling MRSA in a hospital, but we decided to continue.to employ the methods of control that had proved successful for 23 years in preventing MRSA introduced from outside the hospital from becoming established in it. We have maintained this policy through the years, and we have been able to retain the usefulness of the isoxazolyl penicillins to treat patients in our hospital with presumed and proven S. aureus infections. In the light of the controversies about the advisability of attempts to control the spread of MRSA in hospitals and uncertainty about the importance of airborne spread of staphylococci in practice, we are currently reviewing our entire infection control policy for dealing with MRSA. We believe that BSP has been at least as successful as traditional isolation methods in preventing the spread of infection in our hospital. The object of this investigation was to find out whether this was the case. However, we intro-

December

AJIC 1993

duced BSP not only for the control of infections in patients but also as an alternative to universal precautions for the protection of hospital staff from infection by blood-borne viruses. We have not attempted to compare the two systems from that standpoint, but the nursing department, the occupational health department, and the infection control service have found the BSP system relatively easy to teach to hospital staff members. We also believe that it is simpler to implement in practice than a combination of universal precautions and category-based isolation. We thank Dr. J. P. &alai and Mr. M. J. Katie the statistical analysis of the data.

for assistance

with

References 1. Lynch P, Jackson MM, Cummings MJ, Stamm WE. Rethinking the role of isolation practices in the prevention of nosocomial infections. Ann Intern Med 1987; 107:243-6. 2. Centers for Disease Control. Isolation techniques for use in hospitals. DHEW pub no (PHS) 70-2504. Washington, DC: Government Printing Office, 1970. 3. Garner JS, Simmons BP. Guideline for isolation precautions in hospital. In: Centers for Disease Control. Guidelines for prevention and control of nosocomial infections. Atlanta: Centers for Disease Control; 1983:1-24. 4. Rennie RP, Duncan IBR. Emergence of gentamicinresistant Klebsiella in a general hospital. Antimicrob Agents Chemother 1977;11:179-84. 5. Duncan IBR, Rennie RP, Duncan NH. A long-term study of gentamicin-resistant Pseudomonas aeruginosa in a general hospital. J Antimicrob Chemother 1981;7:147-55. 6. Troya SH, Jackson MM, Lovrich-Kerr M, McPherson DC. A survey of nurses’ knowledge, opinions, and reported uses of the Body Substance Isolation System. AM J INFECT CONTROL 1991;19:268-76. Lynch P, Cummings MJ, Roberts PL, Herriott MJ, Yates B, Stamm WE. Implementing and evaluating a system of generic infection precautions: body substance isolation. AM J INFECT CONTROL 1990;18:1-12. Garner JS, Jarvis WR, Emori TG, Horan TC, Hughes JM. CDC definitions for nosocomial infections, 1988. AM J INFECT CONTROL 1988;16:128-40. Lennette EH, Balows A, Hausler WJ, Shadomy HJ. Manual of clinical microbiology. 4th ed. Washington, DC: American Society for Microbiology, 1985. 10. George WL, Sutter VL, Citron D, Finegold SM. Selective and differential medium for isolation of Clostridium dificile. J Clin Microbial 1979;9:2 14-9. 11. Kleinbaum GD, Kupper LL, Morgenstem H. Epidemiologic research: principles and quantative methods. New York van Nostrand Reinhold, 1982:285-8. 12. Ad Hoc Committee of the Committee on Trauma, Division of Medical Sciences, National Research Council. Postoperative wound infections. Ann Surg 1964;16O(Suppl):23. 13. Nugent ME, Bone DH, Datta N. A transposon, Tn 732, encoding gentamicititobramycin resistance. Nature 1979; 282~422-3. 14. Weinstein RA, Kabins SA. Strategies for prevention and

AJIC Volume

21,

Number

Duncan

6

and Batckebr

Tabaqchali S. The epidemiology of Clostridium difjicile with use of a typing scheme: nosocomial acquisition and cross-infection among immunocompromised patients. J Infect Dis 1986;153:159-62. 18. Noble WC. The dispersal of staphylococci in hospital wards. J Clin Path01 1962;15:552-58. 19. Boyce JM. Should we vigorously try to contain and control methicillin-resistant Staphylococcus aureus? Infect Control Hosp Epidemiol 1991;12:46-54. 20. Haley RW. Methicillin-resistant Staphylococcus aweus: do

control of multiple drug-resistant nosocomial infection. Am J Med 1981;70:449-54. 15. Hughes JM, Jarvis WR. Nosocomial gastrointestinal infections. In: Wenzel RP, ed. Prevention and control of nosocomial infections. Baltimore: Williams & Wilkins, 1987:405-39. 16. McFarland LV, Mulligan ME, Kwok RYY, Stamm WE. Nosocomial acquisition of Clostridium dificile infection. N Engl J Med 1989;320:204-10. 17. Heard SR, O’Farrell S, Holland D, Crook S, Barnett MJ,

ERRATUM In the article “Job analysis 1992: Infection control practitioner,” by Bjerke and coworkers, which appeared in the April issue of the JOURNAL (1993;2 1:5 l-7), there is an error in Table 17 on page 56. The data in the “old” and “new” columns were inadvertantly transposed. The corrected table is shown below:

Table content

17. Comparison outlines

of old and new test

for the CBIC

examination

No.of Content

Items old

area

I. The infectious process A. Host factors 8. Microbiologic agents C. Clinical findings and diagnostic tests TOTAL II. Surveillance and epidemiologic

No. of Items new

5 9

6 6

7 21

11 23

investigation A. Data collection 8. Data analysis

19

23

11

18

C. Data reporting

1 31

4 4.5

22 36 60

15 30 45

4

3

2 13 5

6 9 2

5 29

5 25

2 4 3 9

2 7 3 12

TOTAL Ill. Transmission of infection A. Risks 8. Prevention and intervention TOTAL IV. Management and communication A. Influence of organization and regulation of the health care system B. Role of infection control personnel

C. Program planning and evaluation D. Human relations skills E. Personnel

and financial

resource

agement TOTAL V. Education A. Principles of adult learning B. Program planning and evaluation C. Instructional methods TOTAL

We regret any confusion

309

caused by this error.

man-