Fifteen episodes of infection due to Pseudomonas aeruginosa, including peritonitis and catheter site infections, occurred in nine patients receiving continuous ambulatory peritoneal dialysis over a 27-month period. Eight episodes were associated with catheter loss. Occurrence of P aeruginosa infection was significantly associated with use of povidone-iodine solution to cleanse the catheter site. There was no association with use of povidone-iodine solution to disinfect tubing connections, use of other skin care products or exposure to other environmental sources of P aeruginosa. Cultures of available povidone-iodine products were negative. Local irritation and alteration in skin flora caused by antiseptic solution or low-level contamination of povidone-iodine solution are potential mechanisms of infection.

Methodologies for evaluating the efficacy of antimicrobial agents in the bacterial flora of the hand involve variations of the “glove juice” or “sterile bag” procedures. In this study we demonstrate that the subungual flora contribute significantly to the recovery of bacteria in these procedures. Culturing hands that had the subungual spaces sealed with acrylic polish demonstrated a decay type of curve for the number of bacteria recovered in successive samplings with 106 colony forming units (CFU) bacteria recovered in the first sampling and 103 in the sixth while the opposite hand on which the subungual spaces were not sealed demonstrated a continuous recovery of 106 CFU.

Serial scrubbings of three minutes with plain detergent, 10% povidone-iodine or 4% chlorhexidine demonstrated a three logarithm reduction for povidone-iodine and four logarithm reduction compared to plain detergent alone for chlorhexidine in hands that had subungual spaces sealed, while the opposite hand showed only a 0.5 to 1.0 logarithm reduction. Povidone-iodine produced a 0.6 logarithm reduction and 4% chlorhexidine or 1.4 logarithm reduction in the density of subungual bacteria.

Studies of nosocomial bloodstream infection (NBI) have come primarily from university, municipal or community hospitals with little specific information from veterans hospitals. The objective of this study was to define trends in NBI at a tertiary care veterans hospital. Patients with NBI were identified by retrospective review of microbiology records and of infection control surveillance records from 1979 to 1987 at the Buffalo Veterans Administration Medical Center (VAMC). Between 1979 and 1985 there was no significant upward or downward trend. Beginning in early 1986 a significant increase in NBI was noted. This increase was because of changes in incidence of gram-positive NBI (primarily Streptococcus aureus and enterococcus) while there was no change in incidence of gram-negative NBI. The most common sources of NBI were similar to previous studies and included urinary tract infection, pneumonia, surgical wound infections and intravenous catheter related infections. The frequency of various organisms causing NBI was likewise similar to other studies except for pneumonia in which Streptococcus pneumoniae most commonly was isolated. The yearly incidence of gentamicin resistance among gram-negative NBI isolates demonstrated no significant trend. In conclusion, there was a significant increase in incidence of NBI at the Buffalo VAMC between 1979 and 1987 that occurred primarily in the last two years studied. The importance of gram-positive organisms as a cause of NBI at the Buffalo VAMC has been clearly documented as a recent phenomenon.

The usefulness of routine annual tuberculin skin testing (purified protein derivative [PPD]) of hospital employees has been questioned. Between 1984 and the end of 1987 the PPD conversion rates of hospital employees at a university and psychiatric hospital in North Florida were compared. The number of employees in both hospitals were almost equal and compliance with the annual testing was more than 95%. In the psychiatric hospital tuberculosis screening of patients was practiced on admission and annually thereafter. Although no unsuspected smear-positive tuberculosis patients were diagnosed in the psychiatric hospital as compared to four in the university hospital, the annual conversion rates of employees were 0.42% and 0.13%, respectively (p>0.001). However, the ratios of these conversion rates to the incidence of tuberculosis in the counties where these hospitals are located respectively were similar (20.0 vs 24.3, p=0.7). The community seems be the major source of the PPD conversion. At the university hospital more than $70,000 was spent on the annual PPD testing to discover 15 converters; nine had no or minimal contact with patients and only two complied with isoniazid (INH) prophylaxis. Annual PPD testing is not cost effective in hospitals with a low annual conversion rate among its employees and with low tuberculosis case rates in the hospital and the surrounding community.

This review of the use of surveillance by infection control practitioners (ICPs) in nosocomial infection control programs has identified key components that have led to and supported its continued application and success. These include:

◾ Surveillance targeting of events (diseases);

◾ Early development of standardized definitions;

◾ Wide acceptance of these criteria;

◾ Advocacy, leadership and education of methodology and;

◾ A high level of effectiveness in program practice.

Group B streptococci are the most frequent cause of life-threatening infections in newborn infants in the United States. The reservoir of this organism is thought to be the mother's vaginal or rectal microflora or both. Prevalence surveys have documented arriage rates ranging from 8% to 31%. Increased rates of vertical transmission are seen in infants born to women with dense vaginal colonization at delivery. In most cases, early onset group B streptococcal disease in neonates is diagnosed during the first day of life.' While the carrier rate may approach 30% in pregnant women, the incidence of serious disease in neonates is 0.2% to 0.4% with a mortality rate approaching 50% in some studies. Risk factors prominently associated with early onset group B streptococcal disease have been identified as premature labor, prolonged rupture of amniotic membranes and intrapartum fever.

This article is the first in a new series entitled “Topics in Long-Term Care.” The author, a full-time academic physician, is based at a chronic disease hospital with a 577-bed nursing home where he is head of infectious diseases and, with the support of a full-time infection control officer (RN), manages the infection control programs as the hospital epidemiologist and chair of the infection control committee. Although most interest centers on the nursing home setting, the title reflects the need to consider all long-term care facilities (LTCFs) with special populations and risk factors for acquiring infection, e.g., rehabilitation units, day care centers and day hospitals for the elderly and institutions for developmental disabled persons of all ages.