Fifteen rural Wisconsin acute care community hospitals with an average approved bed size of 55 and an average daily census of 28 patients participated in a nosocomial infection control project. Each hospital Infection Control Practitioner (ICP) was trained and conducted prospective nosocomial infection surveillance on all patients admitted to the hospital for 6 consecutive months between May 1,1984 and April 30, 1985. Two hundred twenty nosocomial infections were reported among 13,420 discharged patients for an incidence rate of 1.64 infections per 100 discharged patients. One hundred sixty-four patients had one nosocomial infection. Twenty-three patients had two or more. Infection rates were highest among gynecology— 4.9% and general surgery patients— 4.0%, and lowest among newborns—0.3% and pediatric patients—0%. 39.7% of the infections were of the urinary tract, 27.9% of surgical wounds, 16% pneumonia, and 1.4% primary bacteremia. The other infections were in seven additional sites. Risk factors associated with acquisition of infections included old age, urinary catheterization, and/or a surgical procedure. The overall nosocomial surgical wound infection incidence for inpatient procedures was 1.9%, with incidences of 0.4% for hernia repair, 1.3% for cholecystectomy, 3.3% for appendectomy, 4.0% for total abdominal hysterectomy, and 3.9% for cesarean sections; The incidence of nosocomial infections was 2.7 infections per 100 discharged patients age 65 years or over and 0.9 infections per 100 discharged patients less than 65 years. Two hundred thirty-six microorganisms were cultured from 175 of the infections. Staphylococcus aureus, coagulase-negative Staphylococcus, and Enterococcus were the most common gram-positive organisms. E. coli, Pseudomonas, Proteus, and Klebsiella were the most common gram-negative organisms cultured. The overall nosocomial infection incidence was much lower than reported rates for other groups of hospitals. These data have implications for the review and potential modification of complex guidelines and requirements for infection control in smaller rural hospitals.

It is widely held that bubbling humidifiers do not produce microaerosols, although prior studies have resulted in conflicting evidence. We have studied this phenomenon in a clean room using an airborne particle counter and samplers for airborne bacteria. At gas flow rates between 10 and 80 L/min, a Cascade 1 humidifier produced between 460 and 999 water droplets/L humidified gas. Total water volume aerosolized was approximately 10-8 ml/L humidified gas. Seventy-three percent of the particles had diameters between 1 and 5 microns. With the reservoir containing 6.4 × 106P. aeruginosa/ml, it produced between 2 and 9 P. aeruginosa/Lhumidified gas. Most of the bacteria were in particles of a size likely to be deposited in the lung. This bacterial carryover was between 20 and 150 times the amount predicted by multiplication of the water volume aerosolized times the concentration of bacteria in the humidifier reservoir. An Air Life humidifier produced fewer particles which were also of a size likely to be deposited in the lung and, when the reservoir contained P. aeruginosa, it aerosolized bacteria. Wick-type humidifiers did not produce detectable aerosol or bacterial carry-over. Although the clinical significance of these findings has not been established, they provide a rationale for the CDC recommendations for procedures designed to keep bubbling humidifier reservoir water uncontaminated.

The frequency of handwashing in two intensive care units (ICUs) was observed. Handwashing after direct contact with patients or their support equipment was recorded. The ratio of beds to sinks was 1:1 in the medical ICU and 4:1 in the surgical ICU. Surveillance of physicians, nurses, and other personnel demonstrated a greater frequency of handwashing by nurses (63%) compared with physicians (19%) and other personnel (25%). The nurses in the unit with one sink per bed had a significantly greater number of handwashes (76%) than those in the unit with fewer sinks (51%).

The authors modified the Centers for Disease Control's guideline for disease-specific isolation precautions to a hospital computerized information system. Entering a suspected diagnosis selected from the isolation option on computer terminals generated: 1) a printout listing the isolation instructions, infective material(s), and persons who should avoid exposure; 2) an order for the appropriate supplies; 3) a patient charge based on the supplies required; and 4) an option for stopping, changing, or listing the orders. In order to implement this system, both extensive in-service training for nurses and efforts to change ordering practices of physicians were necessary.

Prevalence surveys before and after computerization were used to evaluate the new system. Combined surveys showed that isolation was ordered for only 21% of patients when indicated. Failure to isolate was identified as a significant problem. As a consequence, continuous surveillance and consultation of all infected patients were instituted, resulting in isolation orders for 81% when indicated. The computerized disease-specific system has resulted in better and more accurate use of isolation, probably due to in service education and surveillance efforts.

Given the wide range of available health care personnel handwashing agents, selection of an appropriate product may be difficult. This decision may be made on the basis of user preference, cost, or other factors unrelated to product effectiveness. Four criteria—efficacy, safety, cost, and acceptability—are appropriate for systematic evaluation of handwashing products. These criteria are applied to para-chloro-meta-xylenol (PCMX), a compound used with increasing frequency in health care personnel handwashing agents. Published data regarding the biochemical properties, efficacy, and safety of PCMX are summarized. We conclude that the substance appears to be safe and efficacious. However, the activity of PCMX is highly formula-dependent and many of the studies available in scientific literature have been conducted in Europe using a variety of testing conditions and formulations different from those currently available in the US. Clinical studies of marketed formulations are beginning to appear in the literature. Such studies will provide the data needed for adequate product evaluation.

As we have become experienced in dealing with AIDS, we have learned that early concern over employees who refuse to care for patients with AIDS was merely the tip of the iceberg from the infection control standpoint. During the past 5 years, additional questions and concerns have been raised regarding employee health and AIDS. An interesting body of literature has emerged recently that discusses many of these current employee health issues. Some of these questions have been answered in major communications in The New England Journal of Medicine, the Journal of the American Medical Association, and The Lancet. Many other topics have been covered in short communications in letters to the editor in these same journals and in the CDC's Morbidity and Mortality Weekly Report (MMWR). When taken together, this writing presents a very clear picture of what is happening with AIDS in the health care arena. It is obvious that most of the concerns regarding AIDS in health care settings can be answered by taking a close look at the available epidemiologic data. It is important to remember that the decision-making process related to AIDS is easier if we try to separate the emotional issues surrounding the disease from the scientific data available.

Medicine and infectious disease textbooks have traditionally advised physicians to obtain urine cultures routinely before the initiation of therapy for presumed urinary tract infection. In practice, however, many physicians treat patients empirically, without doing a culture, and a recent cost-effectiveness analysis has supported this approach, at least in selected patients. What, then, are trie pros and cons of diagnostic urine cultures, and in which situations does current evidence favor their use?