The incidence of nosocomial infections and antimicrobial resistance rates of nosocomial pathogens vary considerably among countries and even among intensive care units (ICUs) within one hospital. Such differences might be partly due to the selection pressure exerted by certain antibiotics, since intensive care patients are given more antimicrobials than any other group of patients. We therefore compared resistance rates of four important nosocomial pathogens (Staphylococcus aureus, E. coli, Klebsiella pneumoniae, Pseudomonas aeruginosa) isolated from patients in general wards and ICUs. There were few trends toward higher resistance of ICU isolates, and most differences were found with Klebsiella pneumoniae.

We also tried to relate antibiotic use in ICUs and frequency of antibiotic resistance of five selected nosocomial pathogens. The ampicillin and cephalosporin resistance of E. coli and Klebsiella pneumoniae arose along with an increase in usage of both drugs. Decreasing prescription of cotrimoxazole was not reflected by decrease in resistance of Staphylococcus aureus and Staphylococcus epidermidis. Increasing prescriptions of tetracyclines were followed by an increasing resistance of E. coli, but not of Staphylococci. The oxacillin resistance of Staphylococcus epidermidis almost paralleled the consumption, the opposite was true for Staphylococcus aureus. There seemed to be a rather close relationship between the incidence of resistant Staphylococcus aureus, Staphylococcus epidermidis and Pseudomonas aeruginosa strains and the quantities of gentamicin, tobramycin and azlocillin prescribed.

The increase or decrease of prescriptions of certain antimicrobials increased or decreased their resistance rate to the respective drugs of only certain bacterial strains in one ICU, but not in the other. The findings in our hospital cannot necessarily be applied to other hospitals. Restriction of antimicrobial usage however decreased resistance rates in most situations.

Apart from the special mechanisms for spread of IV device and respiratory equipment-related infections, the mechanisms for spread of infections in the intensive care unit (ICU) are no different from other areas of the hospital. Control of transmission must emphasize the necessity of blocking the contact and airborne routes of spread.

Hand disinfection with an alcoholic preparation and the use of patient-bound gowns for all direct patient contact are efficient measures against contact spread of infections.

Airborne spread is best reduced by nursing patients in one-patient rooms with doors closed. A substantial part of the beds should be in such rooms. However, they are of little value if basic barrier nursing techniques are neglected. An ante-room to the one-patient room is practical for hand disinfection, gowning, etc., and if ventilated, offers added protection against airborne spread.

Barriers against transmission of microorganisms must be created between patient rooms. Since basic barrier nursing techniques are the determining factors for infection control, the need for adequate numbers of well-trained and well-educated staff cannot be overestimated.

The prevalence and pattern of nosocomial infections in a new hospital in the United Arab Emirates is presented. During an 18-month period in which there were 6,544 discharges, 379 nosocomial infections occurred in 310 patients, representing an attack rate of 4.7%. The commonest site of infection was the urinary tract which accounted for 42.2% of the total. This was followed by surgical wound infections, cutaneous infections, bloodstream infections, lower and upper respiratory infections, infected burns and gastrointestinal infections. The most common organisms were Escherichia coli (22.7% of isolates), Pseudomonas species (17.5%) and Staphylococcus aureus (16.7%). Escherichia coli was the most common cause of urinary tract infections and bacteremia; Staphylococcus aureus was the most common cause of surgical wound and cutaneous infections. There was only one Serratia marcescens infection. We observed a high incidence of Pseudomonas infections, particularly of the urinary tract, and Staphylococcal surgical wound and other cutaneous infections, but otherwise our data are similar to nosocomial infection data from US hospitals.