Update: Delayed Onset Pseudomonas fluorescens Bloodstream Infections After Exposure to Contaminated Heparin Flush --- Michigan and South Dakota, 2005--2006
In March 2005, CDC reported a multistate outbreak of Pseudomonas fluorescens bloodstream infections associated with use of syringes preloaded with heparin intravenous catheter flush (1). The heparin flush became contaminated during preparation by IV Flush, LLC (Rowlett, Texas). Thirty-six patients in four states were identified who had been exposed to the contaminated flush and subsequently experienced P. fluorescens bloodstream infection during December 2004--February 2005 (1). Based on a recommendation by the Food and Drug Administration (FDA), IV Flush voluntarily recalled the preloaded syringes in late January (1); on January 31 and February 4, 2005, FDA issued nationwide alerts recommending that consumers and institutions stop using and return the preloaded syringes to IV Flush or the distributor (Pinnacle Medical Supply, Rowlett, Texas). Approximately 3 months after the product was recalled, patients in Michigan and South Dakota were identified with P. fluorescens bloodstream infections. As of April 2006, a total of 15 patients in Michigan and 13 in South Dakota had been identified with delayed onset P. fluorescens bloodstream infections, with occurrences ranging from 84 to 421 days after their last potential exposure to the contaminated flush. The patients all had indwelling central venous catheters and received treatment during October 2005--February 2006 at clinics known to have used the contaminated flush. This report describes the investigation of these cases, which determined that these were delayed onset cases of P. fluorescens bloodstream infection from a past exposure to contaminated flush, and provides recommendations for ongoing surveillance for delayed P. fluorescens bloodstream infections among similarly exposed patients.
In October 2005, the Michigan Department of Community Health (MDCH) was notified by a hospital infection-control practitioner of a case of P. fluorescens bloodstream infection in a woman aged 51 years with breast cancer. She was receiving chemotherapy through an implantable venous port (a type of indwelling central venous catheter) and was being treated at the only Michigan clinic that had used the product manufactured by IV Flush. The patient's bloodstream infection was identified 233 days after her last potential exposure to the contaminated flush. After consultation with CDC, MDCH became aware of additional cases of delayed onset P. fluorescens bloodstream infection at the only clinic in South Dakota that had used the implicated flush. After being contacted by MDCH, the South Dakota clinic and the South Dakota Department of Health provided case information.
In this report, a case is defined as illness in a Michigan or South Dakota resident with 1) culture-confirmed P. fluorescens* bloodstream infection diagnosed during February 4, 2005--March 31, 2006, and 2) who had received treatment at a clinic known to have used the contaminated flush before it was recalled. MDCH requested the Michigan and South Dakota clinics that had used the contaminated flush to review all microbiology records and report all cases of P. fluorescens bloodstream infection diagnosed after the product was recalled in January 2005. Medical records of all patients with a diagnosis of P. fluorescens bloodstream infection were reviewed, with a focus on determining last potential exposure to the contaminated heparin flush.
Local and state laboratories recovered P. fluorescens isolates from blood samples provided by the clinics and hospital emergency departments. CDC laboratories tested catheters that had been removed surgically from the patients and shipped to CDC with assistance from the health departments; in some instances, P. fluorescens bloodstream infection had already been diagnosed after a positive blood culture, and in others, blood cultures were negative for patients who had known exposures to the contaminated flush and later had onset of bloodstream infection symptoms. Catheter sections were prerinsed to remove nonadherent cells and cultured for the presence of P. fluorescens biofilms. CDC compared blood and catheter isolates by using pulsed-field gel electrophoresis (PFGE) and used scanning electron microscopy to confirm the presence of P. fluorescens biofilms.
A total of 28 patients from Michigan and South Dakota had illness consistent with the case definition, with diagnosis dates ranging from April 29, 2005, to March 10, 2006. Median age was 58 years (range: 24--79 years). Twenty (71%) patients were female. All had cancer and were outpatients with implantable venous ports. Of the 28 patients, 27 (96%) experienced chills <8 hours after receiving a port flushing (i.e., with uncontaminated flush); other signs and symptoms included fever in 14 (50%) patients and nausea or vomiting in 10 (36%). Twenty-two (79%) patients were treated with oral antibiotics; some received antibiotics when bloodstream infection was suspected because of clinical symptoms (especially if they had been to an emergency department) or when they had a positive blood culture. Because some patients had negative blood cultures and did not receive diagnoses until catheter removal and culture, they received antibiotics later. All 28 had their ports (i.e., catheters) removed surgically. No deaths were reported. The mean time from last potential exposure to the contaminated flush until diagnostic specimen collection was 237 days (range: 84--421 days).
P. fluorescens was recovered from port catheters in 17 (81%) of 21 patients,§ from blood cultures in 15 (79%) of 19 patients, and from both blood and catheter cultures in four (33%) of 12 patients. PFGE was performed on 19 available P. fluorescens isolates from blood or catheter cultures. PFGE patterns indicated that all 19 isolates were genetically indistinguishable from or closely related to isolates from the March 2005 investigation (1). Catheter specimens were sent to CDC for examination by scanning electron microscopy, which indicated the presence of biofilms with bacilli adhering to catheter lumens.
On-site investigations in Michigan and South Dakota confirmed that the clinics were no longer using and had returned the recalled flush. Cultures of the new flush being used in Michigan and South Dakota did not recover any organisms. Surveillance for additional cases of delayed onset P. fluorescens bloodstream infection continues at the Michigan and South Dakota clinics described in this report. IV Flush, the manufacturer of the contaminated product, has ceased operation.
Reported by: J McHale, Sioux Falls; J Clayton, South Dakota Dept of Health. C Kim, MD, Univ of Michigan School of Public Health; M Wilkins, DVM, E Wells, MD, J Rudrik, PhD, Michigan Dept of Community Health. MJ Arduino, DrPH, J Noble-Wang, PhD, B Jensen, MMSc, J Carr, A Srinivasan, MD, Div of Healthcare Quality Promotion, National Center for Preparedness, Detection, and Control of Infectious Diseases (proposed); M Gershman, MD, EIS Officer, CDC.
This report describes the first known cases of substantially delayed bloodstream infections (i.e., 84--421 days) after exposure to a contaminated intravenous solution. PFGE analysis linking P. fluorescens isolates from the current investigation both to one another and to those of the original investigation (March 2005) confirmed that these were delayed onset cases from past exposure to the contaminated heparin flush.
P. fluorescens is an aerobic, gram-negative bacterial rod that grows best at temperatures of approximately 77ºF--86ºF (25ºC--30°C) and grows poorly at the standard hospital microbiology incubation temperature of approximately 97ºF (36°C) (1,2). Therefore, identification of P. fluorescens can be difficult, depending on laboratory capabilities. The bacteria typically thrive in moist environments (including water, soil, and foods) and are not a frequent cause of human infections. However, P. fluorescens has been reported to cause occasional cases of transfusion-associated septicemia in blood recipients, including fatal reactions (3), and catheter-related bacteremia in patients with cancer (4).
At least four factors contributed to the delayed onset of symptoms and diagnosis of bloodstream infection in the patients described. First, P. fluorescens colonized catheter lumens by forming biofilms (as confirmed by electron microscopy at CDC); previous electron microscopy studies have indicated that nearly all indwelling vascular catheters become colonized by microorganisms embedded in a biofilm layer (5). Biofilm is composed of these microorganisms and a structural matrix of extracellular polymers, primarily polysaccharides, produced by the microorganisms. P. fluorescens in the contaminated heparin flush either colonized preexisting catheter lumen biofilms or initiated new biofilm formation; heparin might have stimulated biofilm formation, which has been reported in vitro with another bacterial species (6). Although P. fluorescens might not have entered patient bloodstreams in sufficient quantities to cause symptoms on initial exposure to the contaminated flush, biofilm formation enabled the bacteria to persist in patient port catheters. The bacteria might have proliferated in the biofilm, from which they were disrupted by subsequent, uncontaminated flushes and released into the bloodstream, finally causing symptoms. Second, 12 patients who were no longer receiving chemotherapy received catheter flushes infrequently and consequently did not have frequent flush-related bloodstream infection symptoms. Third, patients still receiving chemotherapy initially assumed bloodstream infection symptoms were chemotherapy side effects because some symptoms are similar. Finally, isolating P. fluorescens from clinical specimens initially was difficult because of its growth requirements.
The data in this report are subject to at least four limitations. First, cases might have been missed because of false-negative blood cultures at local laboratories. Second, information on South Dakota cases might be inaccurate or incomplete because investigators in Michigan performed South Dakota chart reviews by proxy. Third, recall bias might have resulted in incomplete or inaccurate information from patients regarding symptoms (e.g., type, onset date, or timing of onset relative to flushes). Finally, the case-finding methods varied between the South Dakota and Michigan clinics.
The data in this report indicate that patients with implantable venous ports who receive heparin flushes contaminated with P. fluorescens are at risk for bloodstream infection up to 14 months after last receiving the contaminated flush. Subsequent episodic catheter flushing with uncontaminated flush might physically disrupt biofilms causing symptomatic, intermittent bacteremia. Health-care providers should conduct ongoing surveillance and be aware of possible bloodstream infection in patients with indwelling catheters who have received contaminated injections, even several months after exposure. Catheter removal in such instances is strongly recommended, especially among immunocompromised patients, because antibiotic therapy alone might not eradicate P. fluorescens from catheter biofilms. Providers should alert laboratories when bacterial species with atypical growth requirements are suspected clinically.
This report is based, in part, on data contributed by C Abuel, T Peer, PJ Powers, Osteopathic Medical Oncology and Hematology (clinic), Michigan.
* Including two patients with isolates identified as Pseudomonas fluorescens-putida group.
Confirmation of P. fluorescens presence through positive blood or catheter culture in a symptomatic patient (e.g., a patient with at least one of the following signs or symptoms: chills, fever, nausea, or vomiting).
§ The remaining seven catheters either were discarded by surgeons or hospital laboratorians before they could be transported to CDC or were available but not sent.
Disclaimer All MMWR HTML versions of articles are electronic conversions from ASCII text into HTML. This conversion may have resulted in character translation or format errors in the HTML version. Users should not rely on this HTML document, but are referred to the electronic PDF version and/or the original MMWR paper copy for the official text, figures, and tables. An original paper copy of this issue can be obtained from the Superintendent of Documents, U.S. Government Printing Office (GPO), Washington, DC 20402-9371; telephone: (202) 512-1800. Contact GPO for current prices.**Questions or messages regarding errors in formatting should be addressed to firstname.lastname@example.org.
Date last reviewed: 9/7/2006