From JSTOR.org, June 2007
We examined a cluster of 5 hemodialysis patients who contracted gram‐negative bacteremia. A nurse who used an artificial fingernail to open a vial of heparin that was mixed to make a flush solution had a culture of an artificial fingernail specimen positive for Serratia marcescens. The typing of the S. marcescens strains isolated from the 5 patients and the nurse showed them to be identical. This finding provides strong support for policies prohibiting artificial nails for healthcare workers in all hemodialysis units.
The hands and fingernails of healthcare workers have been associated with outbreaks of nosocomial infections.1 Artificial fingernails have been proven to harbor microorganisms to an extent greater than natural nails and have been specifically implicated as a causal factor in transmitting bacteria and fungi to patients.1‐6 Because of this increased risk of contamination via artificial fingernails, current Centers for Disease Control and Prevention guidelines state that healthcare workers should not be allowed to wear artificial nails if they have direct contact with patients at high risk of acquiring infection.1 Patients who receive hemodialysis with tunneled catheters are particularly vulnerable to episodes of catheter‐related bacteremia.7,8 We report here an outbreak of gram‐negative bacteremia in 5 dialysis patients secondary to contamination of an intravenous infusion by a nurse wearing artificial fingernails.
On the day of the outbreak (day 0), 31 patients underwent dialysis in 2 sessions, including 12 who received dialysis via tunneled catheters. During the afternoon dialysis session on day 0, 3 patients became symptomatic with fever and rigors. Two of these patients were later found to have bacteremia due to Serratia marcescens and Achromobacter xylosoxidans, and the third had bacteremia due to A. xylosoxidans alone. Over the following 4 days, 2 additional patients—both of whom had undergone dialysis during the morning session on day 0—became febrile; both were found to have bacteremia due to these same 2 gram‐negative organisms.
All 61 patients who were receiving dialysis in this unit were observed for symptoms of infection. The 12 patients who underwent dialysis via tunneled catheters on the day of the outbreak all had blood samples for culture obtained. Environmental samples for culture were obtained from materials used in the hemodialysis procedure and from items used in the dialysis unit by staff members and/or patients.
All staff members were interviewed about their activities on the day of the outbreak, and information was obtained about patient assignments, preparation of medication and flush solutions, the use of barriers and hand hygiene, and sharing of lotions and other materials. Using a premoistened swab, culture samples were taken from the hands and fingernails of staff members.
Of the 31 patients who underwent dialysis on day 0, none of the 19 who received dialysis via grafts or fistulas became bacteremic. Of the 12 patients who received dialysis via a tunneled catheter, 5 patients had blood cultures positive for S. marcescens and/or A. xylosoxidans. Thirty other patients underwent dialysis either the day before or the day after the outbreak, but none became bacteremic. A total of 52 culture samples were obtained from dialysis fluids and the environment, and none of the cultures were positive for either S. marcescens or A. xylosoxidans.
During interviews, staff members reported that on day 0, there had been a shortage of the single‐use vials of heparin flush solution used to clear the lines of patients with tunneled catheters. In the absence of single‐use vials of heparin, 2 nurses had each opened a multidose heparin vial and mixed it into a 1,000 mL bag of saline, so that small aliquots could be withdrawn and used throughout the day to flush the dialysis catheters.
The heparin flush solution mixed by nurse 1 was used for 6 patients during the day, none of whom became bacteremic. The heparin flush solution mixed by nurse 2 was used for 6 other patients, 5 of whom developed bacteremia. On questioning, nurse 2 admitted that she had not washed her hands prior to mixing the heparin flush solution and that she had used an ungloved fingernail to detach the metal lid from the vial of heparin.
Of the cultures of hand and nail samples from 21 healthcare workers from the dialysis unit, all were negative for S. marcescens and A. xylosoxidans, except 1: the culture of the samples from the artificial fingernails of nurse 2 was positive for S. marcescens.
The S. marcescens strains isolated from patients in this outbreak and from nurse 2 were typed by the National Institutes of Health Clinical Center Microbiology Laboratory, using random amplification of polymorphic DNA (RAPD) analysis and repetitive sequence‐based polymerase chain reaction (repPCR) DNA fingerprinting. The epidemiologic profiles by RAPD analysis and repPCR typing of the S. marcescens strains isolated from the 5 patients and from nurse 2 were identical.
In this report, we have linked an outbreak of bacteremia in 5 hemodialysis patients to a nurse wearing artificial fingernails. This nurse introduced bacteria by using her fingernail to open a vial of heparin that was added to a bag of saline and left at room temperature to be used throughout the day as a flush solution for tunneled catheters.
The epidemiologic evidence supporting this link comprises the following findings. (1) Only patients who had tunneled catheters and received heparin flush solution became bacteremic, and no patients who underwent dialysis using permanent vascular access developed bacteremia. (2) Only patients who received the heparin flush solution mixed by the nurse wearing artificial nails became bacteremic—specifically, 5 of 6 patients who received the flush solution mixed by the nurse wearing artificial nails developed bacteremia, whereas 0 of 6 patients who received the flush solution mixed by the other nurse developed bacteremia. (3) Patients who underwent dialysis in the afternoon session became symptomatic immediately, whereas those who received dialysis in the morning session became symptomatic over several days. This is consistent with the hypothesis that patients in the afternoon session were infected with more organisms, as a result of the growth of bacteria in the infusion bag throughout the day.
Although the original bags of heparin flush solution were not available to culture, the S. marcescens isolates recovered from each of the bacteremic patients were found to be identical by 2 different methods—RAPD and repPCR DNA fingerprinting—and this same strain was isolated from underneath the fingernails of the nurse wearing artificial nails. While we did not isolate A. xylosoxidans from any source, it is logical to assume that this organism was also introduced by the nurse wearing artificial nails in a manner similar to that by which S. marcescens was introduced. A. xylosoxidans thrives in a moist environment and has been shown to be associated with nosocomial infections, including bacteremia.9
Poor hand hygiene certainly contributed to the genesis of this outbreak. The nurse wearing artificial nails stated that she did not wash her hands prior to opening the vial of heparin, nor did she wear gloves. She did, in fact, pry off the metal lid of the heparin container using an artificial fingernail.
Artificial fingernails are known to harbor more pathogens than natural nails. In 1 study, 86% of healthcare workers wearing artificial nails had a pathogen identified on the nails, compared to 35% of healthcare workers with natural nails.10 Published studies have demonstrated a causal link between artificial nails and outbreaks of Klebsiella pneumonia and Pseudomonas aeruginosa infection in different neonatal intensive care units,4‐6 an outbreak of postoperative wound infections due to S. marcescens,2 and an outbreak of osteomyelitis due to Candida albicans in 3 patients following laminectomy.3
National hand hygiene guidelines state that healthcare workers should “not wear artificial fingernails or extenders when having direct contact with patients at high‐risk (e.g., those in intensive‐care units or operating rooms).”1(p33) On the basis of the present report of an outbreak of gram‐negative bacteremia causally linked to a healthcare worker wearing artificial nails, and the well established risk of bacteremia in patients undergoing hemodialysis, policies prohibiting artificial nails should be strictly enforced in all hemodialysis units.