A Program for Surveillance of Hospital-Acquired Infections in a General Hospital: A Two-Year Experience

Transcription

1 REVIEWS OF INFECTIOUS DISEASES. VOL. 3, NO.4. JULY-AUGUST by The University of Chicago. All rights reserved /81/0304-Q004$02.00 A Program for Surveillance of Hospital-Acquired Infections in a General Hospital: A Two-Year Experience Nachurn Egoz and Dan Michaeli* From the Infectious Diseases Unit, Chaim Sheba Medical Center, Tel Hashomer, Israel A program for surveillance of nosocomial infections was developed and carried out in six departments of Chaim Sheba Medical Center, Tel Hashomer, Israel. A team of nurseepidemiologists made regular visits to the wards and the microbiology laboratory and recorded all episodes of nosocomial infections, which had been defined by a team of specialists from various hospitals and institutions in Israel. Monthly computerized reports were used to follow the pattern of hospital-acquired infections and the patterns of antibiotic sensitivity of the causative organisms. During a two-year period, 15,207 patients were admitted to the hospital and 1,851 episodes of nosocomial infection were reported-a rate of for such infection. Postoperative wound infections were the most common infections in surgical services, accounting for between one-third and onehalf of the total number of the infections among patients in these units. More than 50% of the infections in the intensive coronary care unit were located at iv sites. In the services for premature and newborn babies the most common infections were conjunctivitis and bacteriuria, the clinical significance of which is not clear. Between one-half and three-fourths of the nosocomial infections were caused by gram-negative organisms. Surgical wound infections were more common after extensive and lengthy operations. The risk of such infections usually increased with age of the patient and was higher after emergency surgery. Since 1971 the infectious diseases unit of the Chaim Sheba Medical Center in Tel Hashomer, Israel, has been involved in various experimental programs of surveillance of nosocomial infections. The objectives of such programs were (1) the collection ofdata on the pattern of nosocomial infections in the hospital and identification of the more prevalent forms of such infections and of the wards, procedures, and groups of patients associated with high risk of acquiring nosocomial infections; (2) the early detection of changes in those patterns; (3) the early detection of outbreaks of nosocomial infections; (4) the routine monitoring of the antibiotic-sensitivity patterns of the bacterial flora causing infections in the hospital; and (5) that the presence and activity of the surveillance teams in the various wards would be This study was supported by a grant from the Chief Scientist, Ministry of Health, Israel. We are deeply grateful to the staff of the Infectious Diseases Unit at Sheba Medical Center and especially to the nurseepidemiologists Mrs. B. Tavdioglo, Mrs. P. Amit, and Mrs. E. Maoz. Please address requests for reprints to Dr. Nachum Egoz, District Health Department, Ministry of Health, Haifa, Israel. * Present address: Tel Aviv Municipal Hospitals, Tel Aviv, Israel. associated with higher awareness by the personnel in these wards of the problem of nosocomial infections and perhaps would bring about desirable changes in attitudes and practices. At the same time, a group of specialists from various hospitals and institutions in Israel was developing a standard set of definitions of nosocomial infections and criteria for their recognition. The aim of this work was to use one standard method of surveillance throughout the country to allow comparisons to be made among hospitals and to establish common preventive programs. After the set of definitions was completed, a systematic program of surveillance of nosocomial infections was started on January 1, 1975, at six services of the Chaim Sheba Medical Center. This report will summarize some of the data collected during the first two years of the program. Patients and Methods Six services participated in the surveillance program: general surgery, orthopedic surgery, chest and heart surgery, the intensive coronary care unit (lccu), and the departments for newborn and premature babies. There were 15,207 admissions 649

2 650 Egoz and Michaeli to the participating wards during the two-year period covered by this report. The surveillance was carried out by nurse-epidemiologists from the infectious diseases unit. Each nurse was permanently assigned to several wards and made daily visits to the wards, checking the patients' charts and the nursing reports. Each case of fever or diagnosis of infection made by the staff in the ward was recorded on a special form by the nurse-epidemiologist. Entered on this form were the patient's identifying data, e.g., age, sex, ethnic group, dates of admission and discharge; details of underlying disease, diagnostic and therapeutic procedures, antibiotic treatment, and infection, i.e., date of onset, site, and, if identified, the causative organism and its antibiotic sensitivity. A weekly inquiry was made at the microbiology laboratory. The nurse-epidemiologist followed up each patient affected by a nosocomial infection until s/he was discharged from the department under surveillance. All the forms were checked and discussed at regular meetings of the team of nurses and the epidemiologist of the unit. The data were analyzed by computer. A monthly computer report was discussed and evaluated by the staff of the infectious diseases unit and was used to identify trends and to detect changes in the pattern of nosocomial infections in the participating wards. Definitions Nosocomial infection. A nosocomial infection was defined as a clinically evident or bacteriologically proven infection in a hospitalized patient, the onset of which occurred at least one incubation period after the admission of the patient or, when the period of incubation was not known, at least 48 hr after admission of the patient. The isolation of a new pathogen from a site where another pathogen had previously been isolated was recorded as a separate episode of infection. Postoperative fever. Postoperative fever was considered a nosocomial infection if within five days after an operation the patient's temperature rose to 38 C or higher and this fever persisted for at least 48 hr but no specific infectious process was diagnosed. Surgical wound infection. A surgical wound infection was recorded when signs of inflammation were present at the wound area or in deep organs that had been exposed during the operation and the inflammation was associated with either a purulent discharge or positive bacteriologic findings. Septicemia. A diagnosis of septicemia was made when symptoms and signs of inflammation were present concomitantly at several sites. Urinary tract infection. A diagnosis of urinary tract infection was based on typical clinical signs (e.g., dysuria, fever, frequency of urination, and loin tenderness) associated with pyuria or, when clinical signs were absent, on two consecutive positive urine cultures (~104 organisms/ml of urine). Infection in an iv site. Such an infection was recorded when local signs of inflammation were present and accompanied by purulent discharge or a positive bacteriologic finding. Conjunctivitis in a newborn. Conjunctivitis in a newborn was defined on the basis of clinical findings (purulent discharge) or positive bacteriologic findings. Results The rate of acquired infections in each of the six participating departments is shown in table 1. A total of 1,851 episodes of infection was recorded, a number representing a rate of This rate, of course, does not represent the true rate of nosocomial infections at Chaim Sheba Medical Center, as these six services under surveillance and the patients in these units do not form a representative sample of the hospital. Each of the services must be looked at separately. The chest and heart surgery service had 1,210 admissions during the two-year period; there were Table 1. Incidence of hospital-acquired infections at Chaim Sheba Medical Center, Tel Hashomer, Israel, Service Chest and heart surgery General surgery Orthopedics Intensive coronary care Premature babies Newborns Total No. of admissions 1,210 1,829 2, ,667 15,207 No. of episodes of acquired infections (010 )* 397 (32.8) 247 (13.5) 234 (10.7) 216 (25.3) 164 (35.5) 593 (6.8) 1,851 (12.2) Rate given is number of episodes per 100 patients admitted to service.

3 Surveillance of Hospital-Acquired Infections 651 Table 2. Episodes of hospital-acquired infections among 1,210 patients admitted to the chest and heart surgery service at Chaim Sheba Center, Tel Hashomer, Israel, : identification of isolates and their distribution by type of infection. Surgical Postoperative Respiratory Total Isolate Septicemia wound fever tract Other (070)* Staphylococcus aureus (2.3) Escherichia coli (0.8) Klebsiella (1.7) Proteus (1.2) Pseudomonas (1.8) Other (3.6) Sterile culture (4.9) Not cultured (16.5) Total (%)* 13 (1.1) 108 (8.9) 184 (15.2) 30 (2.5) 62(5.1) 397 (32.8) * Total number of episodes (per 100 patients). 397 episodes of acquired infections - a rate of 32.80/0. Table 2 shows the distribution of those infections by site and identifies the common pathogens. The largest group of infections, accounting for 184 episodes, was that of postoperative fever. We suppose that most of these cases were not of infectious etiology but were cases of postcardiotomy syndrome, although all these cases conformed to the definition of postoperative fever that was agreed on. If cases defined as postoperative fever were excluded, only 213 cases of nosocomial infection would be left, and the rate of genuine infections would decrease to 17.6%. Half of these remaining infections (l08 cases) were surgical wound infections, which will be discussed later in more detail. Gram-negative enteric bacteria (Escherichia coli, Klebsiella, Pseudomonas, and Proteus) were responsible for approximately one-half of the infections, and Staphylococcus aureus was responsible for one-fifth. Table 3 shows the data from the general surgery ward. Two hundred forty-seven infections were recorded-13.5% of 1,829 admissions. The largest group was that of surgical wound infections, which formed almost one-third of the acquired infections. Urinary tract infections were the second largest group, accounting for 25.50/0 of all the infections in this ward. There were 17 cases of septicemia, i.e., almost 1% of the admissions. Seventy-eight percent of the pathogens (159 of 203) were enteric, gram-negative bacteria. The orthopedic service (table 4) had 2,185 ad- Table 3. Episodes of hospital-acquired infections among 1,829 patients admitted to the general surgery ward at Chaim Sheba Medical Center, Tel Hashomer, Israel, : identification of isolates and their distribution by type of infection. Surgical Postoperative Respiratory Urinary Total Isolate Septicemia wound fever tract tract Other (%)* Staphylococcus aureus (0.8) Escherichia coli I (3.5) Klebsiella (3.0) Proteus (0.8) Pseudomonas I 0 15 (0.8) Other (2.1) Sterile culture (0.1) Not cultured (2.3) Total (%)* 17 (0.9) 78 (4.3) 20(1.1) 41 (2.3) 63 (3.4) 28 (1.5) 247 (3.5) * Total number of episodes (per 100 patients).

4 652 Egoz and Michaeli Table 4. Episodes of hospital-acquired infections among 2,185 patients admitted to the orthopedic service at Chaim Sheba Medical Center, Tel Hashomer, Israel, : identification of isolates and their distribution by type of infection. Isolate Surgical wound Postoperative fever Urinary tract Other Total (010)* Staphylococcus aureus (0.7) Escherichia coli (1.4) Klebsiella (0.9) Proteus (1.0) Pseudomonas (1.5) Other (1.5) Sterile culture (0.04) Not cultured (3.8) Total (%)* 99 (4.5) 73 (3.3) 31 (1.4) 31 (1.4) 234 (10.7) * Total number of episodes (per 100 patients). missions with 234 acquired infections, a rate of The distribution by site of infection is similar to that seen in general surgery; postoperative wound infections (99 cases) form the largest group (42% of the acquired infections). Postoperative fever accounted for almost one-third of the infections (73 cases). Third in frequency of occurrence was the group of urinary tract infections (31 cases, 13% of the acquired infections). Of the pathogens isolated from cases of nosocomial infections, 72% were gram-negative enteric bacteria. The ICCU presented a different picture (table 5). Among 854 admissions, 216 cases (25.3%) of acquired infections were recorded. More than 50% of the infections were cases of phlebitis at the iv site. Most of these infections were not cultured, and therefore their etiology was unknown. Infections of the respiratory and urinary tracts came next in the order of frequency. There were nine cases of sepsis. Out of 48 infectious episodes for which cultures were positive, 28 (58%) were caused by gram-negative enteric bacteria. One hundred sixty-four infections were acquired by the 462 admissions to the ward for premature babies (table 6), a rate of In 61 cases (37% of the infections) bacteriuria was recorded. In 52 of these cases E. coli was isolated. Another large group-approximately one-fifth of the acquired infections-was that of clinical conjunctivitis for which the culture was sterile. In an additional 22 cases of conjunctivitis (13.4% of the infections), a bacterial agent was isolated: S. aureus in eight cases, Klebsiella in eight cases, and other bacteria in six cases. Most of the isolates, i.e., 78% out of 126, were gram-negative enteric bacteria. The newborn department (table 7) had 8,667 ad- Table 5. Episodes of hospital-acquired infections among 854 patients admitted to the intensive coronary care unit at Chaim Sheba Medical Center, Tel Hashomer, Israel, : identification of isolates and their distribution by type of infection. Isolate At iv site Respiratory tract Urinary tract Other Total (%)* Staphylococcus aureus (0.6) Escherichia coli (0.9) Klebsiella (0.9) Proteus (0.6) Pseudomonas (0.8) Other (1.8) Sterile culture (5.3) Not cultured (14.4) Total (%)* 138 (16.2) 24 (2.8) 22 (2.6) 32 (3.8) 216 (25.3) * Total number of episodes (per 100 patients).

5 Surveillance of Hospital-Acquired Infections 653 Table 6. Episodes of hospital-acquired infections among 462 babies admitted to the premature babies ward at the Chaim Sheba Medical Center, Tel Hashomer, Israel, : identification of isolates and their distribution by type of infection. Isolate Urinary tract Conjunctiva Respiratory tract Other Total (070)* Staphylococcus aureus (3.9) Escherichia coli (14.3) Klebsiella (3.7) Other (5.4) Sterile culture (7.6) Not cultured (0.6) Total (070)* 61 (13.2) 59 (12.8) 10 (2.2) 34 (7.3) 164 (35.5) * Total number of episodes (per 100 babies). missions of newborn babies during thetwo-year period. Five hundred ninety-three episodes of nosocomial infection were recorded, a rate of The large number (449) of conjunctival infections stands out. Almost three-fourths of these infections were sterile on culture, many after repeated attempts to culture bacteria or to isolate virus. The definition that was used for conjunctival infection did not require the isolation of a pathogen from purulent discharge and, therefore, did not exclude the common cases of chemical conjunctivitis in newborns. If these cases of sterile conjunctivitis are excluded, the rate of nosocomial infections in this service decreases to 3.0% (264 cases out of 8,667 admissions); this rate would take into account only 57 cases of conjunctivitis with proven bacterial etiology and 67 cases of bacteriuria. There were only eight cases of skin infections in two years; two of these were caused by S. aureus. One of the first problems that we analyzed in more detail was that of surgical wound infections, which appeared to be the most common infection among patients in the surgical wing. In the orthopedic service 1,527 operations were performed during the two-year period. Surgical wound infections developed in 65 patients who underwent surgery, a rate of More than one pathogen was isolated from a number of patients; the distribution of the 99 total isolates is given in table 4. Approximately two-thirds of the isolates were enteric, gram-negative bacteria, Pseudomonas aeruginosa accounting for 27.3% and S. aureus for 13.1%. The rate of wound infections among these patients was significantly higher after emergency operations (19/297) than after elective surgery (46/1,230) (6.4% and , respectively; X 2 = 4.1, P < 0.05). The rates were almost identical for both sexes and increased with age; the rate was 3.7% for patients from birth through 44 Table 7. Episodes of hospital-acquired infections among 8,667 babies in the newborn unit at Chaim Sheba Medical Center, Tel Hashomer, Israel, : identification of isolates and their distribution by type of infection. Isolate Conjunctiva Urinary tract Umbilicus Septicemia Other Total (070)* Staphylococcus aureus (0.3) Streptococci (0.2) Escherichia coli (1.0) Klebsiella (0.3) Proteus (0.3) Other (0.3) Sterile culture (3.8) Not cultured (0.8) Total (070)* 449 (5.2) 67 (0.8) 30 (0.3) 18 (0.2) 29 (0.4) 593 (6.8) * Total number of episodes (per 100 babies).

6 654 Egoz and Michae/i Table 8. Rates of surgical wound infection by age, sex, and site of elective operations among patients who underwent surgery in the orthopedic service of Chaim Sheba Medical Center, Tel Hashomer, Israel, No. of infections/no. of operations (010) for indicated age group (years) Procedure Total Unknown All operations 65/1,527 (4.3) 36/963 (3.7) 24/306 (7.8) 5/60 (8.3) 0/198 Males 39/906 (4.3) 27/650 (4.2) 11/120 (9.2) 1/8 (12.5) 0/128 Females 26/621 (4.2) 9/313 (2.9) 13/186 (7.0) 4/52 (7.7) 0/70 Elective hip operations 13/162 (8.0) 4/69 (5.8) 8/69 (11.6) 1/16 (6.3) 0/8 All other elective operations 33/1,068 (3.1) 24/727 (3.3) 8/172 (4.7) 1/12 (8.3) 0/157 years, 7.8% for patients 45-74, and 8.3% for patients 75 and older (table 8). The same trend was found when emergency and elective operations were analyzed separately. Table 9 shows the risk of surgical wound infection after various types of elective procedures. Operations carrying a high risk of infection were those of the hip (8070), the spine (9.6%), and those correcting limb malformations (8%; this figure should be viewed with reservation due to the small number of procedures performed). Operations associated with moderate risk were knee operations (2.6%), various osteotomies (4%), and "other" surgeries (mostly open reductions; ). There was a low risk of infection after hand (1.6%) and soft tissue (1.6%) operations. The high risk of infection associated with all hip operations could be explained by the age of patients who undergo hip surgery: 55% of these patients were older than 45 compared with only 20% of patients who had other types ofsurgery. Table 8 shows the age-specific rates of surgical wound infections after hip operations as compared with the rates for all other elective operations. The rates after hip operations were higher in both younger and older patients. (The number of infections in patients aged 75 and older was too small to allow comparison.) The age-standardized rates of wound infection was after hip operations and 3.3 % after all the other types of surgery in this service. Therefore, the higher risk of infection after operations on the hip area was not due primarily to the age factor. The general surgery service performed 1,234 operations during the two-year period. Fifty-six patients (4.5%) had postoperative wound infections; the distribution of the 76 isolates is presented in table 3. Approximately three-fourths of the isolates were enteric, gram-negative organisms, E. coli being the most common species. Table 10 shows the specific rates of infection by site of operation. The highest risk of wound infections was associated with surgery on the large intestine (17.5%); the risk was slightly lower for other types of major abdominal surgery, i.e., that on the small bowel (16070), the biliary system (12.5%), and the stomach ( ). More common procedures, e.g., herniotomy, appendectomy, and anal-region surgery, were associated with a relatively low risk of infection. The incidence of surgical wound infection showed a clear trend toward increase with age ( among patients from birth through 29 years, in those aged 30-59, 6.0% in those 60-74, and 9.8% in those 75 and older; table 11). In the chest and heart surgery service, 925 operations were performed; 74 patients developed infection in the surgical wound, a rate of 8.0%. Table 9. Rates of surgical wound infections, by type of operation, among patients who underwent elective operations in the orthopedic service of Chaim Sheba Medical Center, Tel Hashomer, Israel, No. of patients with No. of wound infections Type of operation operations (%) Total hip replacement 83 6 (7.2)I Thompson; Kenn- 162 nailing 25 3 (12.0) 13 (8.0) Other hip operations 54 4 (7.4) Spinal 52 5 (9.6) Soft tissue (1.6) Knee (2.6) Limb malformations 25 2 (8.0) Osteotomies (4.0) Hand (1.6) Other (3.2) Total 1, (3.7)

7 Surveillance of Hospital-Acquired Infections 655 Table 10. Rates of surgical wound infections, by site of operation, among patients who underwent surgery in the general surgery service of Chaim Sheba Medical Center, Tel Hashomer, Israel, No. of patients with No. of wound infections Site of operation operations (010) Stomach (10.3) Small intestine 25 4 (16.0) Large intestine (17.5) Gall bladder and ducts (12.5) Laparotomy 50 4 (8.0) Thyroid and parathyroid 40 0 Breast 40 0 Hernia (0.3) Appendix (3.9) Anal region (0.9) Biopsies 45 0 Other (5.5) Total 1, (4.5) Table 2 presents the 90 bacterial isolates by species. The gram-negative, enteric bacteria accounted for less than one-half of the pathogens (43070). S. aureus was the single most common pathogen, causing 23% of the infections. Table 12 shows the rates of wound infections after various types of operations. Those operations in which thoracotomy was performed were clearly associated with a higher risk of surgical wound infections. Those procedures without thoracotomy (e.g., pacemaker implantation) had a relatively low risk of infection. After coronary bypass operations, the rate of infections exceeded 22%. In eight out of nine cases the thoracotomy incision became infected. Only one patient developed an infection at the site of incision in the leg, where the vein autograft was taken. Operations on the esophagus and the lungs came next in the order of risk (rates of 18% and 14%, respectively). In heart operations, both open and closed, the rates were lower: 9% and 7070, respectively. The only significant difference among heart operations (P = 0.01) was between coronary bypass operations and all other heart operations. Males had a significantly higher rate of infection than did females: vs (table 13), but if the 40 patients who had coronary bypass surgery are excluded (33 were males), the difference in the sex-specific rates of infection among patients in the chest and heart surgery unit is insignificant. The rate of infection increased with age, being 5.5% in children and in patients The rate decreased to 7.4% among patients aged 60 and older (both sexes). This decline proved to be caused by the fact that implantations of pacemakers generally were done in older people (table 13). The age-specific rates of surgical wound infections after pacemaker implantations are compared with those in the rest of the patients in this unit who underwent surgery. If pacemaker implantations are excluded, the decline in the rate of infections in patients older than 60 disappears. Discussion This surveillance program was developed and carried out primarily as an experimental model. It covered only six of more than 40 wards of the Chaim Sheba Medical Center, and the data presented here are not to be taken as representative of the hospital as a whole. However, a rate of nosocomial infections of 12.2% is higher than the rates usually reported by general hospitals, which are in the range of % [1-4], although occasionally a rate higher than 10% has been described [5]. Strict adherence to the preset criteria resulted in the inclusion of cases that, most probably, were not of infectious etiology, e.g., the cases of conjunctivitis in newborns or those of phlebitis at the iv site not associated with bacteremia or formation of pus. Exclusion of these types of infections would have resulted in a rate of for hospital-acquired infections (1,290 cases for 15,207 admissions), a rate that approaches the range usually reported by general hospitals. A more detailed analysis of the data may be used in setting priorities for future action, Table 11. Rates of surgical wound infections, by age, among patients who underwent surgery in the general surgery service of Chaim Sheba Medical Center, Tel Hashomer, Israel, Age group No. of patients with (years) No. of operations wound infections (0J0) (3.4) (4.6) (6.0) (9.8) Unknown 52 0 Total 1, (4.5)

8 656 Egoz and Michaeli Table 12. Rates of surgical wound infections, by type of operation, among patients who underwent surgery in the chest and heart surgery service of the Chaim Sheba Medical Center, Tel Hashomer, Israel, No. of patients with wound infections Type of operation No. of operations (070) Open heart (9.1) Coronary bypass 40 9 (22.5) Closed heart (7.2) Pacemaker (2.1) Lung ISO 21 (14.0) Esophagus (by thoracotomy) 22 4 (18.2) Other 98 3 (3.1) Total (8.0) at least in the departments that participated in the program. Postoperative wound infections clearly are the most frequent and serious infections acquired in our surgical services. Our rates for wound infections exceeded those reported by the National Nosocomial Infections Study in the United States [1, 6] but did not exceed those reported by Cruse from a thorough surveillance program in the surgical services of a general hospital in Canada [7]. Several groups with high risk of acquiring surgical wound infections were identified: older patients, those undergoing extensive and lengthy operations, and those requiring emergency surgery. Programs for prevention of surgical wound infections, such as those for the use of antibiotic prophylaxis, should be directed primarily to those groups. There is also a place for educational and administrative action. The importance of nosocomial infections and the frequency with which they occur must be stressed among hospital personnel and administrators in order to emphasize the need for hygienic practices, both environmental and personal, and for strict adherence to the too-often neglected principles of asepsis - even simple handwashing. Urinary tract infections are also a common problem calling for action. The clinical significance of these infections in the departments for newborn and premature babies is not clear, but no doubt exists about the importance of urinary tract infections in the surgical services that participated in the program. The problem of phlebitis resulting from indwelling venous catheters requires further study to identify risk factors, such as those associated with specific types of iv equipment and the particular sites of infusion. The surveillance program by itself could not answer such questions and was not intended to do so. Often specific studies must be performed to answer specific questions; however, surveillance programs are important for delineating areas of higher priority and presenting a balanced picture of the patterns and trends of the problem of nosocomial infections. Therefore, it was our intention that surveillance be expanded to as many wards as possible and becomes a routine activity of the hospital, serving as a basis for prevention of nosocomial infection, as a tool of evaluation of preventive action, and as a component in the monitoring of the quality of care given in our hospital. We feel a tool was developed that can meet these goals. References 1. Center for Disease Control. National Nosocomial Infections Study Report. Annual Summary Center for Disease Control, Atlanta, Ga., issued February Table 13. Rates of surgical wound infection, by age, sex, and type of operation, among patients who underwent surgery in the chest and heart surgery service of Chaim Sheba Medical Center, Tel Hashomer, Israel, No. of infections/no. of operations (0J0) for indicated age group (years) Procedure Total Unknown All operations 74/925 (8.0) 8/146 (5.5) 7/102 (6.9) (12.6) 23/311 (7.4) 0/81 Males 52/535 (9.7) 3/68 (4.4) 6/56 (10.7) 25/161 (15.5) 18/198 (9.1) 0/52 Females 22/390 (5.6) 5/78 (6.4) 1/46 (2.2) 11/124 (8.9) 5/113 (4.4) 0129 Pacemaker operations (2.1) 0/4 1/15 (6.7) 4/169 (2.4) 0/46 All other operations 69/691 (10.0) (6.1) (13.0) 19/142 (13.4) 0/35

9 Surveillance of Hospital-Acquired Infections Britt, M. R., Burke, J. P., Nordquist, A. G., Wilfert, J. N., Smith, C. B. Infection control in small hospitals. J.A.M.A. 236: , Wenzel, R. P., Osterman, C. A., Hunting, K. J., Gwaltney, J. M., Jr. Hospital-acquired infections. I. Surveillance in a university hospital. Am. J. Epidemiol. 103: , Simchen, E., Michel, J., Epstein, L. M., Sacks, T. G. Infections in the surgical departments of a teaching hospital in Jerusalem. Isr. J. Med. Sci. 12: , Stark, F. R., Seufert, M., Collins, T.C., Franklin, F. Total surveillance program of infections: an analysis of two new programs in army teaching hospitals. Milit. Med. 141:33-35, Center for Disease Control. Trends in surgical wound infection rates - United States. Morbidity and Mortality Weekly Report 29:27-33, Cruse, P. J. E. Incidence of wound infection on the surgical services. Surg. Clin. North Am. 55: , 1975.