As part of the first two phases of the SENIC Project (Study on the Efficacy of Nosocomial Infection Control), information was collected from the heads of the infection surveillance and control programs (ISCPs) in U.S. hospitals. The data were analyzed to describe these respondents and to determine whether differences among them were related to their areas of professional training or to characteristics of the hospitals where they were located. The findings indicate that the ISCP heads constitute a very heterogeneous group, with substantial differences in age, professional training (40% are pathologists), characteristics of their medical practices, memberships in professional organizations related to infection control, time spent in ISCP activities, approach to epidemiologic problems, and opinions on the preventability of nosocomial infections and the seriousness of infection problems in their hospitals. These differences are related strongly to the ISCP heads' professional training, size of hospital, and, to a lesser extent, medical school affiliation, but there is little evidence that the differences are related to regional or urban-rural location or type of ownership of the hospitals. The average ISCP head estimates that about half of all nosocomial infections are preventable, but these estimates vary inversely with tenure in the position and the tendency to approach a clinical problem epidemiologically.

One measure used in the management of staphylococcal disease outbreaks in newborn nurseries is to obtain nasal cultures from nursery personnel and then treat nasal carriers with topical antibiotic ointments. Because recent infection control guidelines often omit reference to this measure, we surveyed current practices in larger hospitals in the U.S.A. Seventy-one percent of respondents indicated that they would obtain nasal cultures from personnel during a staphylococcal disease outbreak, and 40% indicated that they would prescribe topical antibiotic ointments for personnel with positive nasal cultures before the results of bacteriophage typing became known. Because little has been written about the efficacy of topical intranasal antibiotics within the past decade, we compared bacitracin ointment with a vancomycin ointment for treatment of the staphylococcal nasal carrier state. Both ointments reduced nasal carriage in the majority of instances. However, similar reduction was also observed in an untreated control group. The need for different approaches to the problem of nasal carriers is discussed.

A methicillin-resistant strain of Staphylococcus aureus (MRSA, phage type 84/85) was introduced into City of Memphis Hospital by a burn patient who had recently been treated for MRSA bacteremia in another institution 500 miles distant. Despite prompt recognition of the problem and institution of isolation procedures, six other patients developed secondary colonization during the ensuing six months, and five of these experienced clinically significant infections with MRSA. Three of the patients originally infected with MRSA, as well as two additional patients, subsequently developed colonization with staphylococcal strains of phage type 84/85 that were resistant to both methicillin and gentamicin (MRGRSA). Spread of the staphylococcal strains was most likely accomplished primarily via passive transfer from person to person. The hydrotherapy unit, which became contaminated with both MRSA and MRGRSA, may have played a secondary role. As illustrated by this outbreak, patients carrying potentially dangerous bacterial strains should be identified and informed of the problems posed by such carriage. It may be imprudent to admit such patients to hospitals that are free of the potential pathogen.

The outbreak described here exemplifies a number of potential problems associated with control of nosocomial staphylococcal infections: (a) interhospital spread of methicillin-resistant strains; (b) secondary patient-to-patient intrahospital spread; and (c) emergence of even more resistant strains, possibly associated with selective pressures exerted by widespread use of broad-spectrum antimicrobial agents.

The sporicidal activities and shelf life stabilities of three alkaline glutaraldehyde-containing sterilizing solutions were evaluated. The time required to reduce a population of spores by one log is the D-value. The test methods employed were the D-value and survival rates. Sporicidin, a new preparation containing a phenolic buffer, was both more stable and more active against test spores than two other test agents, Cidex and Cidex-7.

In three years we encountered two patients with hospital-acquired myiasis, a rarely reported nosocomial problem. Both patients were elderly and had lengthy thoracic surgery in August in the same operating room. Larvae removed from the nares of one patient and from the chest incision of the other were of the same species, Phaenicia serricata. There was no evidence of tissue destruction or invasion in either case. Investigation revealed several factors that contributed to the presence of flies in the operating room. After a presumed environmental access site was closed and insecticide spraying was augmented, no additional cases occurred. This experience illustrates an unusual problem that may confront those responsible for infection control programs.

Viral illnesses in Strong Memorial Hospital were monitored over a 17-month period. Using criteria based primarily on the incubation periods for a number of common virus infections, the infections we found were classified as hospital- or community-acquired. Hospital-acquired viral infections occurred on most hospital services; the majority of infections occurred on the pediatric and psychiatric services. Infections due to herpesviruses were seen more frequently in a group of patients aged 14 years or older, while infections in patients aged three years or younger were more likely to be due to respiratory syncytial virus, influenzavirus, adenovirus, or parainfluenza virus. Patients with nosocomial infections due to viruses were hospitalized an average of 9.3 days longer than uninfected controls; thus nosocomial viral infections result in increased costs of hospitalization.

Protracted hospital-based epidemics of urinary tract infection and bacteremia due to multiply resistant gram-negative bacilli have become an increasingly common and serious problem. Failure to control such outbreaks stems partly from inability to eradicate a key reservoir, the catheterized bladder. Since eradication of bacteriuria in noncatheterized patients can be achieved with single doses of antimicrobials and correlates with urinary rather than with serum antibiotic concentrations, drugs to which an organism appears resistant by discdiffusion testing, if excreted in the urine in high concentrations, might also prove useful in eliminating catheter-associated bacteriuria. Alternatively, urinary antiseptics, for which antimicrobial sensitivity testing is not usually done, might be effective. To test this hypothesis we determined the minimum inhibitory concentrations (MICs) of 45 multiply resistant Proteus, Serratia, Klebsiella, and Pseudomonas strains isolated in 13 recent epidemics of nosocomial urinary tract infections against 10 selected antimicrobials and urinary antiseptics, and compared these MICs with expected urinary concentrations of each drug. For each genus tested, MICs for at least two antimicrobials or urinary antiseptics were well below easily achievable urinary drug concentrations. Zone size criteria often predicted which drugs had MICs below achievable urinary levels. Little difference was found between MICs determined in Mueller-Hinton broth and in urine. During an epidemic, simultaneous treatment of all patients with bacteriuria by administration of a urinary antiseptic or an antibiotic that achieves high concentrations in urine, in conjunction with brief catheter removal, might prove useful in controlling any further infection.

Five cases of nosocomially acquired herpetic whitlow are presented; three cases developed following a common exposure. All infections were acquired despite the use of surgical gloves as recommended by existing guidelines on control of nosocomial infection. In two cases the mistaken diagnosis of pyogenic paronychia led to surgical intervention, with excessive morbidity and disability. Among the five cases, morbidity ranged from several days of unemployment to permanent sequelae necessitating a career change. Revised guidelines for infection control should include the use of gloves on both hands whenever patient care involves contact with a debilitated patient with herpes simplex labialis.

This article reviews the most likely mechanisms of transmission of the commonly encountered respiratory viruses (influenza, respiratory syncytial virus, parainfluenza, rhinovirus), herpesviruses, and hepatitis viruses, and presents the guidelines used currently for prevention and control that are in use at Strong Memorial Hospital.

Two epidemics of pseudobacteremia are reported. The first, due to Staphyloccus aureus, was caused by a physician who had active staphylococcal skin infection and nasal colonization. Because the blood culture system in use at the time was open, and used screw cap bottles, we assume that the physician contaminated the bottles at the time of inoculation. The second outbreak, caused by Aerococcus viridans, was traced to contamination of the blood culture bottle tops as they were received from the manufacturer. We assume that there was inadequate disinfection of the bottle tops by the physicians prior to their use.

During a three-week period a pseudoepidemic of fungal infections occurred involving patients from two different services. Trichosporon cutaneum and a Penicillium species were isolated from bronchial washings and sputa obtained after fiberoptic bronchoscopy on eight clinically uninfected patients. Investigation revealed contamination of cocaine solutions used for topical anesthesia during bronchoscopy. Contamination is thought to have occurred during preparation of the solutions by pharmacy personnel. Revision of techniques used in formulating and dispensing the solutions resulted in cessation of the pseudoepidemic.

Between May 16 and July 1,1973, four definite cases and one possible case of clinical salmonellosis occurred in a 175-bed community hospital; there were no deaths. Three of the four patients with definite salmonellosis had had cholecystectomies done by the same general surgeon (A); the fourth was an intensive care unit nurse who cared for one of the ill patients during the diarrheal phase of illness before salmonellosis was diagnosed. Epidemiologic investigation implicated the plastic tubing of an intermittent-suction machine located in the recovery room as the environmental reservoir of the organism, and having a nasogastric tube in place postoperatively was the critical host factor related to illness. The salmonella organisms isolated from the suction machine tubing were identical in serologic reaction, biochemical test results, and bacteriophage susceptibility pattern to those recovered from the four patients with confirmed salmonellosis. Antimicrobial susceptibility patterns were similar but not identical.

Selected features of infection control programs among the 163 general hospitals in Tennessee were surveyed in 1976 and 1979. Each hospital but one had a designated infection control practitioner. Three-fourths of the hospitals had fewer than 200 beds and most were in rural areas. The practitioners in these small hospitals worked in an isolated professional milieu: few (4%) had attended a basic training course or were members of a national (11%) or local (16%) infection control association. They also had significantly less access to standard infection control resource publications than did practitioners in large hospitals. Use of aqueous quaternary ammonium compounds for disinfection was reported by 37% of all hospitals in 1979; 68% of hospitals routinely performed bacteriologic cultures of personnel or the environment. In contrast, only 3% of hospitals did not have a policy specifying the use of sterile closed-system drainage of indwelling bladder catheters. Although these practices varied somewhat by hospital size, the differences were not statistically significant. Modest improvement in each parameter was noted since 1976. Pathology was the most common medical specialty (34%) among chairman of infection control committees; internal medicine and pediatrics accounted for only 13%. The practice of routine microbiologic monitoring was significantly more common among hospitals with chairmen who were pathologists. The implications of these findings for national priorities in hospital infection control are discussed.

The incidence of infections among patients undergoing renal transplantation has been reported as high as 83% during the first post-transplant year. In an effort to reduce the occurrence of such infections, we evaluated the role of perioperative prophylactic antibiotics in these patients. Thirty-seven patients undergoing transplantation were assigned randomly to treatment groups (N=20, cefamandole, 1 gm IM every 6 hrs for 8 doses/tobramycin 1 mg/kg IM for 1 dose, starting 2 hrs prior to surgery) and control groups (N=17, no antibiotics). A uniform infection surveillance system was used. Eighty-one percent (30/37) of patients became infected; 70% (14/20) of treated patients and 94% (16/17) of controls (p=0.14). Since antibiotic prophylaxis might be expected to exert the greatest influence in the immediate postoperative period, we examined patient outcome data for both the first seven post-transplant days and for the entire hospitalization. Of the 38% (14/37) of patients who developed infections during the first posttransplant week, a statistically greater proportion came from the control group than from the treated group (10/17 vs. 4/20, p=0.04). These data suggest that this regimen is effective in reducing the occurrence of infections during the first week following renal transplantation.

Between July 2 and July 16, 1979, three patients were reported to have gram-positive cocci as seen on Gram stain of excised, macerated heart valve tissue. No organisms were isolated in culture; however, our usual procedure for handling such valves is not adequate for isolation of strict anaerobes. The first patient was treated for presumed endocarditis; this resulted in three additional weeks of hospitalization.

Review of the remaining 34 patients who had heart valve replacement surgery between February 1 and July 20, 1979 revealed three additional cases.

Careful analysis revealed that nonviable, gram-positive cocci present in commercially prepared trypticase soy broth were being added to the tissue just prior to mortar and pestle grinding in the microbiology laboratory. The contaminated tubes were noted to have a visible sediment when carefully examined.

The procedure for handling the broth has been revised; a simple screening procedure is used now to identify potentially contaminated broth. No further instances of false positive results on Gram-staining of heart value tissue have occurred since this screening procedure was introduced.

This article details the appropriate microbiologic support that is critical to the successful investigation of nosocomial infection problems. The infection control team must have ready access to microbiologic data, and the laboratory should retain epidemiologically relevant bacterial isolates. Investigation of epidemics is facilitated by precise identification of bacteria and careful antibiotic susceptibility testing. In some situations, biotyping, serotyping, phage typing, bacteriocin typing, and other specialized techniques may be required. Plasmid analysis may be useful in the investigation of nosocomial infection problems caused by antibiotic-resistant bacteria.