Nosocomial infections from contaminated endoscopes: a flawed automated endoscope washer. An investigation using molecular epidemiology.

Approximately 1 year after purchase of one manufacturer's automated endoscope washing machine, we began to detect heavy contamination of upper gastrointestinal (UGI) endoscopes cultured after cleaning and disinfection in the washer. During the first 6 months of 1988, 77% of surveillance cultures (20-mL flush through the biopsy channel) were positive for gram-negative bacilli (median concentration, 10(5) cfu/mL), most frequently Pseudomonas aeruginosa serotype 10. During the first 19 months of use of the washer, nosocomial post-UGI endoscopy colonization or infections with P. aeruginosa increased 36%. Investigations show that endoscope contamination derives from a flaw in the design of the EW-10 washer: the detergent holding tank, inlet water hose, and air vents cannot be reliably disinfected and contain heavy biofilms that recontaminate the machine after it has been disinfected, as specified by the manufacturer, with glutaraldehyde. Only by rinsing machine-cleansed endoscopes with 70% alcohol followed by forced air drying has reliable disinfection been achieved. Since adaptation of terminal alcohol treatment and drying, post-UGI endoscopy colonization or infection by P. aeruginosa has declined threefold (p less than 0.001). Testing in other centers using the manufacturer's EW-10 or EW-20 washer has shown similar contamination. In three centers, including our own, postendoscopy infections by machine-associated type 10 P. aeruginosa have been confirmed by demonstrating concordance between isolates from contaminated machines or endoscopes and from infected patients by immunoblot of whole cell lysates and by pulsed-field electrophoresis of DraI endonuclease-digested genomic DNA. This problem reaffirms the vulnerability to microbial contamination of water-containing apparatus and equipment in patient care and points up the critical importance of engineering design to prevent contamination.

Treatment of life-threatening primary pulmonary hemosiderosis with cyclophosphamide.

This report describes a five-year-old boy with severe pulmonary hemorrhage caused by primary pulmonary hemosiderosis with cow's milk sensitivity. After failing to respond to corticosteroids and azathioprine, he dramatically improved after being given cyclophosphamide. He worsened after discontinuation of cyclophosphamide on two occasions and improved significantly with its reinstitution. Cyclophosphamide was continued for 14 months without further bleeding or adverse effects. The patient has remained in remission for nearly five years. Cyclophosphamide may be a life-saving alternative therapy for refractory pulmonary hemorrhage with pulmonary hemosiderosis.

Comparative study of cefazolin, cefamandole, and vancomycin for surgical prophylaxis in cardiac and vascular operations. A double-blind randomized trial.

Three-hundred twenty-one adults undergoing cardiac or major vascular operations were randomized to receive intravenous cefazolin, cefamandole, or vancomycin for prophylaxis against surgical infection in a double-blind trial. All three regimens provided therapeutic blood levels throughout operation in patients studied undergoing cardiopulmonary bypass. The prevalence of surgical wound infection was lowest with vancomycin (4 infections [3.7%] versus 14 [12.3%] and 13 [11.5%] in the cefazolin and cefamandole groups, respectively; p = 0.05); there were no thoracic wound infections in cardiac operations in the vancomycin group (p = 0.04). The mean duration of postoperative hospitalization was lowest in the vancomycin group (10.1 days; p < 0.01) and highest in the cefazolin group (12.9 days). Prophylaxis with vancomycin or cefamandole, compared with cefazolin, did not prevent nosocomial cutaneous colonization by methicillin-resistant coagulase-negative staphylococci; colonization or infection with vancomycin-resistant staphylococci or enterococci was not detected. Adverse effects attributable to the prophylactic regimen were infrequent in all three groups. Eight patients given vancomycin became hypotensive during administration of a dose, despite infusion during a 1-hour period; however, slowing the rate of administration and pretreating with diphenhydramine allowed vancomycin to be resumed and prophylaxis completed uneventfully in five of the patients. We conclude that administration of vancomycin (approximately 15 mg/kg), immediately preoperatively, provides therapeutic blood levels for surgical prophylaxis throughout most cardiac and vascular operations, resulting in protection against postoperative infection superior to that obtained with cefazolin or cefamandole. Vancomycin deserves consideration for inclusion in the prophylactic regimen (1) for prosthetic valve replacement and prosthetic vascular graft implantation, to reduce the risk of implant infection by methicillin-resistant coagulase-negative staphylococci and enterococci; (2) for any cardiovascular operation if the patient has recently received broad-spectrum antimicrobial therapy; and (3) for all cardiovascular operations in centers with a high prevalence of surgical infection with methicillin-resistant staphylococci or enterococci. Guidelines for dosing and administration of vancomycin for cardiovascular surgical prophylaxis are provided.

Surface antimicrobial activity of heparin-bonded and antiseptic-impregnated vascular catheters.

Most Swan-Ganz pulmonary artery catheters have heparin bonded to the surface with benzalkonium chloride, a cationic surfactant, to reduce thrombosis. Since benzalkonium is bactericidal, the antimicrobial activity of heparin-bonded pulmonary artery catheters was investigated in an in vitro assay. Each catheter exhibited activity against a wide variety of potential microbial pathogens, including Candida albicans. The magnitude of activity against individual organisms correlated strongly with their in vitro susceptibility to benzalkonium chloride (r = .94, P < .002). A chlorhexidine-silver sulfadiazine-impregnated catheter exhibited even greater activity than the heparin-bonded catheters (P = .01). When exposed to serum for 24 h, heparin-bonded catheters lost > or = 50% of their antimicrobial activity, whereas the activity of the chlorhexidine-silver sulfadiazine-impregnated catheter was minimally affected. The fortuitous surface antimicrobial activity of heparin-bonded catheters may account for the low incidence of catheter-related bacteremia (mean, 1.0%) compared with Swan-Ganz catheters of the same materials but not coated with benzalkonium-heparin (mean, 2.8%).

Prevention of central venous catheter-related bloodstream infection by use of an antiseptic-impregnated catheter. A randomized, controlled trial.

BACKGROUND: Bloodstream infection related to short-term use of noncuffed central venous catheters is a common and serious problem. Technologic innovations to reduce the risk for these infections are needed. OBJECTIVE: To determine 1) the efficacy of a novel antiseptic catheter in preventing central venous catheter-related infection, 2) patient tolerance of this catheter, and 3) the sources of bloodstream infection originating from noncuffed, multilumen central venous catheters. DESIGN: Randomized, controlled clinical trial. SETTING: Medical-surgical intensive care unit of a 450-bed university hospital. PARTICIPANTS: 158 adults scheduled to receive a central venous catheter; 403 catheters were studied. INTERVENTION: Participants received either a standard triple-lumen polyurethane catheter or a catheter that was indistinguishable from the standard catheter and was impregnated with chlorhexidine and silver sulfadiazine. MEASUREMENTS: Catheters were studied for colonization and catheter-related bloodstream infection at removal; local and systemic effects of catheters were assessed. The origin of each catheter-associated bloodstream infection was sought by culturing all potential sources (skin, catheter segments, hubs, and infusate) and confirmed by restriction-fragment DNA subtyping. RESULTS: Antiseptic catheters were less likely to be colonized at removal than control catheters (13.5 compared with 24.1 colonized catheters per 100 catheters; relative risk, 0.56 [95% CI, 0.36 to 0.89]; P = 0.005) and were nearly fivefold less likely to produce bloodstream infection (1.0 compared with 4.7 infections per 100 catheters; 1.6 compared with 7.6 infections per 1000 catheter-days; relative risk, 0.21 [CI, 0.03 to 0.95]; P = 0.03). In the control group, 8 catheter-related bloodstream infections were caused by Staphylococcus aureus, gram-negative bacilli, enterococci, or Candida species; no infections with these organisms occurred in the antiseptic catheter group (P = 0.003). No adverse effects from the antiseptic catheter were seen, and none of the 122 isolates obtained from infected catheters in either group showed in vitro resistance to chlorhexidine-silver sulfadiazine. Cost-benefit analysis indicated that the antiseptic catheter should prove cost-beneficial if an institution's rate of catheter-related bacteremia with noncuffed central venous catheters is at least 3 infections per 1000 catheter-days). CONCLUSIONS: The chlorhexidine-silver sulfadiazine catheter is well tolerated, reduces the incidence of catheter-related infection, extends the time that noncuffed central venous catheters can be safely left in place for the short term, and should allow cost savings.

Nosocomial Pseudomonas pickettii bacteremias traced to narcotic tampering. A case for selective drug screening of health care personnel.

Three patients in a university hospital developed nosocomial infusion-related Pseudomonas pickettii bacteremia. Investigation identified six additional patients who had received intravenous fluid contaminated by P pickettii but did not become ill. All nine patients had had surgery, and each of these patients but only nine of 19 operated-on control patients had received intravenous fentanyl citrate in the operating room; the mean dose given to the nine case patients was far greater than that given to control patients. Fentanyl in 20 (40%) of 50 predrawn 30-mL syringes was shown to be contaminated by P pickettii. Contamination was caused by theft of fentanyl from predrawn synringes and replacement by distilled water contaminated by P pickettii. Narcotic theft by health care personnel may cause patients to suffer pain needlessly and can also result in dire unanticipated consequences, such as nosocomial bacteremia. Whereas drug testing in the workplace is highly controversial, we believe that testing of health care personnel is indicated when drug abuse or theft is suspected.