Telling cautionary tales of outbreaks, oversights
Patient safety threatened in a variety of ways
Inadequate processing of donor tissue has been linked to the death of a transplant patient, prompting health officials to review the wide variety of treatment methods used in tissue centers, an investigator reported recently in Salt Lake City at the annual conference of the Society for Healthcare Epidemiology of America (SHEA).
The Atlanta-based Centers for Disease Control and Prevention (CDC) is collaborating with the American Association of Tissue Banks of McLean, VA, to do a survey of practices, procedures, and culture methods, says Marion A. Kainer, MD, a CDC epidemic intelligence service (EIS) officer. "There is enormous variation in the processing and culturing methods used by tissue processors," she said.
U.S. tissue banks distribute about 750,000 allografts annually for transplantation. Currently, some tissue banks report using aseptic methods for processing musculoskeletal tissue allografts, and some use post-processing cultures as quality-control indicators. However, the end product may not be sterile, as underscored by four recently reported cases of allograft-associated septic arthritis, Kainer said.1
In particular, the CDC investigation was triggered by the death of a 23-year-old man from Clostridium sordellii sepsis after he received a knee cartilage allograft provided by tissue bank "A."
The investigation revealed that the tissue bank processed 30 tissues from the donor linked to the fatal infection. Of those, 10 were implanted into nine patients, one was discarded, and 19 were cultured at the CDC. In addition to the aforementioned case, one other transplant patient acquired infection that developed into septic arthritis.
That patient recovered after ampicillin/sulbactam therapy. The tissue processing methods used at the facility included suspension of tissue in a nonsporicidal antimicrobial solution at 37 degrees C for 22 hours. Under the protocol used, no pre-processing cultures were taken. All 12 post-processing aerobic and anaerobic cultures of companion tissue were negative at tissue bank A. However, the CDC isolated C. sordellii from at least two of 19 processed, nonimplanted tissues.
Existing processing methods at the tissue bank were inadequate indicators of tissue safety, investigators concluded. Effective sterilization methods that do not functionally alter musculoskeletal tissue are needed to prevent allograft-related infections. The historical problem is that sterilizing tissue may weaken it, undermining the therapeutic benefit to the recipient. "In the past, using the traditional sterilization methods, [the choice] has been having [tissue] sterility or strength," Kainer told Hospital Infection Control. "Therefore, many tissue processors have elected to process tissues aseptically, [but] they were not aware that these tissues could result in bacterial infections.
"There are now, however, new technologies available, including a low-temperature chemical sterilization method, which is sporicidal and does not affect the [tissue] biomechanical properties. So this is now no longer a case of having to decide between sterility and strength. You can actually have both," she added.
A deadly drop from multidose vial
In another outbreak reported at SHEA, the use of multidose vials of albuterol led to fatal Burk-holderia cepacia infections, reported Peter Axelrod, MD, epidemiologist at Temple University Hospital in Philadelphia.2 "One patient, we determined, died as a direct result of B. cepacia infection, and for two [patients], it contributed to death," he said.
Between July 1, 2000, and April 30, 2001, the mean background rate of B. cepacia pneumonia was 0.6 cases a month. The rate for B. cepacia bacteremia was 0.1 cases a month. During the May 1 to July 31, 2001, epidemic period, however, the mean pneumonia rate was 5.3 cases a month and the bactericidal rate was one a month. During the epidemic period, 55 patients had B. cepacia respiratory colonization or pneumonia.
The cases pointed back to the intensive care unit, where respiratory therapists distributed albuterol drops into the nebulizers on ventilated patients. They carried small vials of albuterol in their pockets, using a medicine dropper to dispense the solution. "You’re not supposed to touch the little well that you drop the albuterol into," Axelrod said. "However, it is really easy to touch it, and when we talked to the therapists, they freely admitted that they often touched it. You’re touching a system that is contaminated with patient bacteria."
A switch to patient-dedicated vials stopped the outbreak, he noted. The hospital tried to avoid using multidose vials, but physicians wanted to get high doses of albuterol to the patients. Since the typical solution is too diluted for filling the "well" in respiratory equipment, a concentrated solution was purchased in multidose vials. "The intervention was fairly simple," Axelrod told SHEA attendees. "We just told the respiratory therapists [that] if they use the multidose vial, leave it in the room of the patient and when you go the next patient get a new bottle."
In what was more of a quality breakdown than a clinical outbreak, three patient deaths due to intracranial hemorrhage were traced to a series of quality-control lapses in an unidentified Philadelphia hospital laboratory, reported Soju Chang, MD, a CDC EIS officer.3
Prothrombin time (PT) and international normalized ratio (INR) are laboratory values for monitoring blood coagulation in patients receiving warfarin, an anticoagulant drug with more than 8 million prescriptions in the United States in 1999. On July 25, 2001, personnel at the unidentified Philadelphia hospital recognized falsely low INR reporting from its laboratory after a patient noted ease of bruising.
To find the root cause of the laboratory error and its impact on affected patients, investigators observed laboratory staff and processes. Next, they compared characteristics of patients tested during the period when errors were reported (June 4 to July 25, 2001; N = 843) with a pre-error period (April 15 to June 3, 2001; N = 835).
"We evaluated a subgroup of patients from each period having at least one high (> 3.0) INR, a level exceeding the usual therapeutic range, for possible associations between exposure to the error and adverse events, " Chang told SHEA attendees.
Investigators determined that insufficient quality control and incorrect programming of a laboratory information system generated 2,384 falsely low INRs in the error period. Patients exposed to the error were more likely to have bleeding complications, three of which resulted in the aforementioned deaths.
"Preventable errors due to inadequate laboratory quality control went unrecognized for seven weeks leading to adverse events and deaths," Change said. Laboratories should verify automated LIS calculations, and clinicians should suspect laboratory errors when clinical presentations are discordant with PT/INR results. Despite a chance to obtain another reagent, an old form remained in use and that contributed to the confusion. There also were problems with labeling, and physicians were apparently not reading the package insert for the new reagent. "The physicians were reminded to treat patients and not just laboratory results," he said.
1. Kainer M, Castor M, Ledell K, et al. Clostridium sordellii infection associated with cartilage allograft transplantation: The need for tissue safety. Abstract 95. Presented at the annual conference for the Society for Health Care Epidemiology of America. Salt Lake City; April 6-9, 2002.
2. Crespo A, Axelrod P, St. John K, et al. An epidemic of Burkholderia cepacia pneumonia linked to specific practices in the handling of albuterol for nebulizers. Abstract 267. Presented at the Society for Health Care Epidemiology of America. Salt Lake City; April 6-9, 2002.
3. Chang S, Selenic D, Bell M, et al. Adverse events and deaths associated with laboratory errors at a hospital — Philadelphia, 2001. Abstract 93. Presented at the Society for Health Care Epidemiology of America. Salt Lake City; April 6-9, 2002.