Surge Capacity for Response to Bioterrorism in Hospital Clinical Microbiology Laboratories

Abstract

Surge capacity is the ability to rapidly mobilize to meet an increased demand. While large amounts of federal funding have been allocated to public health laboratories, little federal funding has been allocated to hospital microbiology laboratories. There are concerns that hospital laboratories may have inadequate surge capacities to deal with a significant bioterrorism incident. A workflow analysis of a clinical microbiology laboratory that serves an urban medical center was performed to identify barriers to surge capacity in the setting of a bioterrorism event and to identify solutions to these problems. Barriers include a national shortage of trained medical technologists, the inability of clinical laboratories to deal with a dramatic increase in the number of blood cultures, a delay while manufacturers increase production of critical products and then transport and deliver these products to clinical laboratories, and a shortage of class II biological safety cabinets. Federal funding could remedy staffing shortages by making the salaries of medical technologists comparable to those of similarly educated health care professionals and by providing financial incentives for students to enroll in clinical laboratory science programs. Blood culture bottles, and possibly continuous-monitoring blood culture instruments, should be added to the national antibiotic stockpile. Federal support must ensure that companies that manufacture essential laboratory supplies are capable of rapidly scaling up production. Hospitals must provide increased numbers of biological safety cabinets and amounts of space dedicated to clinical microbiology laboratories. Laboratories should undertake limited cross-training of technologists, ensure that adequate packaging supplies are available, and be able to move to a 4-day blood culture protocol.

Surge capacity, the ability to rapidly respond to a sudden and dramatic increase in needs, has historically been addressed in the military as one component of readiness and in the civilian emergency preparedness setting to meet the needs of the population in response to natural disasters. The provision of medical services in the setting of a terrorist event is another scenario in which surge capacity must be adequate.

In the setting of a bioterrorism (BT) event, plans in place have increased the surge capacity of the public health and emergency medical system to deal with the diverse needs of the population. These plans include increasing the ability of the United States to rapidly deliver large quantities of antibiotics, vaccines, antitoxins, and ventilators from the national antibiotic stockpile and manufacturing sufficient smallpox vaccine to vaccinate the U.S. population. Plans to enable first responders and emergency departments to mobilize in a timely manner to deal with a tremendous increase in the demand for their services are also in place. There are significant costs associated with the maintenance of surge capacity.

Although large amounts of federal funding have been allocated to laboratories in order to deal with an expected increase in services in response to a BT event, nearly all of this funding has been distributed to federal, state, and local public health laboratories, leaving little funding for hospital laboratories. Hospital clinical microbiology laboratories have been designated level A (sentinel) laboratories in the national Laboratory Response Network (LRN). This designation means that these laboratories should not attempt to identify potential BT agents; however, they are required to have the ability to rapidly rule out the possibility of such agents and to forward those isolates for which this possibility cannot be ruled out to an LRN laboratory of level B or higher, such as a state public health laboratory (11, 12). Forwarding such isolates to a level B laboratory is necessary because specialized testing is required for the definitive identification of BT agents and commercially available bacterial identification systems that are routinely used by clinical microbiology laboratories do a poor job of identifying these pathogens (22).

In the current competitive health care environment, many hospitals have decreased the number of medical technologists in clinical laboratories. In some cases, this decrease has been made possible as a result of increases in efficiency; in others, it has been made possible at the expense of the quality of laboratory services. Specialization of medical technologists means that many technologists who work in other areas of clinical laboratories (blood bank, clinical chemistry, etc.) have not had experience in clinical microbiology in many years, and proficiency testing has shown that restructured microbiology laboratories employing less-experienced generalists rather than experienced microbiologists have an increased rate of major identification errors (5).

The ability to deal with an increase in laboratory workload is of concern, as there is a national shortage of qualified medical technologists. According to the Board of Registry of the American Society of Clinical Pathologists, the vacancy rates for medical technologist supervisors in 2000 were the highest reported over the survey's 12-year comparison period, and the upward trends in the vacancy rates from 1996 to 2000 continued for staff medical technologists and medical laboratory technicians (25). Unfortunately, this shortage is not likely to be alleviated by graduates of American programs in clinical laboratory science. Enrollment in these programs has declined dramatically since the 1980s, with the number of medical technology graduates falling from 5,318 in 1983 to 2,491 in 1999 (2). The total number of accredited programs in clinical laboratory science fell from 638 to 273 during this period (2). Salaries of medical technologists, once comparable to those of nurses, have not kept pace. Many technologists have left the field to pursue careers in more highly paying areas, such as biotechnology (17).

MATERIALS AND METHODS

Study institution.

Boston Medical Center, a 547-bed hospital urban medical center, had approximately 93,000 emergency department visits and 644,000 hospital-based clinic visits during 2001. The center provides primary, emergency, obstetric, and tertiary care and uses approximately 44,000 blood culture bottles per year in a series of automated blood culture instruments. The instruments have a total capacity of 864 bottles (432 sets, with two blood cultures constituting one set). Routine blood cultures, cultures of peritoneal fluid, and selected blood bank products are incubated for 5 days prior to being discarded if they test negative. The average number of blood culture sets incubated in the instruments is 301 on the basis of the volume of 22,000 sets per year, though the number being incubated at any given time varies greatly depending upon the hospital census and other factors. The laboratory adheres to current standards for the limited testing of possible agents of BT prior to specimen or isolate transportation to a more specialized laboratory within the national LRN.

Workflow analysis.

The steps in the process from the time a patient presents for medical care until a patient specimen is transported to a laboratory of level B or higher or is cultured for a bacterial pathogen and, if necessary, the isolated pathogen is further evaluated and shipped to a laboratory of level B or higher are shown in Fig. 1, along with the impact that a known BT event would have on the clinical microbiology laboratory. Though a suspected or confirmed BT event due to a virus (e.g., a suspected case of smallpox) would not require culture of the pathogen by the hospital laboratory, the impact on the microbiology laboratory workload would be significant as a result of the need to package and ship multiple specimens, with each patient's specimens requiring individual packaging. For example, for a suspected case of smallpox in which the patient has vesicles or pustules, the specimens that are recommended to be collected and shipped include (3) (i) the top of a vesicle in a 1.5- to 2.0-ml plastic tube; (ii) material from the base of a vesicle or pustule on a microscope slide that is air dried and placed in a plastic slide container prior to shipping; (iii) the unroofed base of a lesion that is applied to an electron microscope grid, allowed to air dry, and then placed in a grid box; (iv) two biopsy samples of vesicles, of which one is to be placed in formalin and the other is to be placed in a 1.5- to 2.0-ml plastic tube; (v) 10 ml of patient blood drawn into a plastic marble-topped tube or a plastic yellow-topped serum separator tube; (vi) posterior tonsillar tissue collected with a swab or brush, with the end of the applicator being broken off into a 1.5- to 2-ml screw-cap tube; and (vii) 5 ml of blood drawn into a plastic purple-topped tube.

FIG. 1.

Analysis of steps in the process of specimen collection and culture during a BT event. DFA, direct fluorescent antibody.

RESULTS

In the setting of an established BT event, it is anticipated that individuals seeking routine and emergency medical care for treatment will include victims of the attack, individuals with symptoms clinically indistinguishable from those due to a BT agent, such as an “influenzalike syndrome,” and anxious people, the “worried well.” Though there are plans for emergency departments to differentiate between individuals who are ill as a result of infection by BT agents and those who are ill due to other causes, the ability to do so on the basis of clinical presentation is, at best, an inexact science. As a result, the number of patient specimens sent to hospital microbiology laboratories is likely to far exceed the true burden of disease due to BT agents. In addition, until the epidemiology of the attack is established, many patients without disease or exposure are likely to seek medical attention even though they have not been in areas that have been attacked. The dramatic increase in specimen volume was well documented during fall 2001 in the setting involving cases of anthrax associated with exposure to letters. As a result of the anthrax attacks, laboratories within the LRN tested more than 125,000 clinical specimens and approximately 1 million environmental specimens (14). Even in areas in which there were no cases of anthrax and no known exposure to contaminated mail, there was a tremendous increase in the volume of specimens. For example, following October 2001, the Illinois Department of Public Health processed more than 1,700 specimens, all of which were negative for Bacillus anthracis (13).

While telephone calls to the laboratory from physicians and other care providers are likely to be dramatically increased, this increase is not likely to be isolated to any particular step in the process. It will result in an increase in the amount of time that laboratory employees spend on the phone rather than performing laboratory work.

Although a BT event will result in collaboration with law enforcement authorities and culture isolates have the potential to be used in legal proceedings, current recommendations for BT procedures do not include chain-of-custody tracking for hospital clinical microbiology laboratories, though such tracking is under discussion. The documentation and tracking of specimens from patients with illnesses thought to be due to BT agents from the point of collection until the test results are finalized would add additional labor to an overburdened laboratory. Requests for chain-of-custody information for patient specimens sent to hospital clinical microbiology laboratories are uncommon. If recommendations change to include chain-of-custody tracking for clinical specimens once BT is suspected, the number of requests for chain-of-custody information would likely be many times higher than the number of patients with true disease due to a BT agent.

A barrier that hospital laboratories would have to deal with is the dramatic increase in the number of patient specimens in the presence of a limited number of class II biological safety cabinets (BSC), especially BSC located in rooms that are under negative pressure. The use of BSC is essential for work involving unidentified category A, category B, and category C critical biological agents, many of which are well-documented causes of laboratory-associated infections. Work with some of these agents requires facilities with a higher containment capacity than class II BSC can provide, and if these agents are suspected, patient specimens may need to be sent directly to a higher-level LRN laboratory.

One especially biohazardous bacterial category A agent is Francisella tularensis, which, despite its infrequency of isolation, has been a common cause of laboratory-associated infection. In a study of 3,921 cases of laboratory-associated infections, this agent ranked second in the United States as a cause of laboratory-associated infections, behind only Brucella species (which are classified as category B agents) (20). Other critical biological agents for which there are documented causes of laboratory-associated infections include the category A agents B. anthracis (4, 20), Yersinia pestis (20), agents causing viral hemorrhagic fever (10, 16), and variola major virus (smallpox) (1). Other causes of laboratory-associated infections include Coxiella burnetii, found to be very common in Pike's study (20), Burkholderia mallei (23), Rickettsia prowazekii (though all reports predate 1968 [6]), a number of food- or waterborne pathogens, and several arboviruses (24) (including Venezuelan, eastern equine, and western equine encephalitis viruses). One safety concern is that many of the viral agents of interest are able to grow in tissue culture cell lines that are commonly used in clinical virology laboratories. If viral cultures were mistakenly performed on a clinical specimen that contained one of these viruses, the result might be laboratory-associated infections.

In the setting of a BT incident with one of the above-mentioned pathogens, the lack of adequate numbers of BSC presents great difficulties. Of course, positive cultures suspected of growing biohazardous agents will be manipulated in BSC. However, it will not be possible to know whether a positive broth bacterial culture, which is more likely to generate an infectious aerosol than is a culture on solid medium, contains a potentially biohazardous agent or merely an organism of low pathogenicity. Although there is a lesser risk of infection from bacterial growth on solid media, the recovery of a possible biohazardous agent on solid media should prompt the further workup of the agent within a BSC. As a result, in the setting of a BT event, all positive broth cultures, such as blood cultures and many of their subcultures to solid media, will be manipulated in BSC, as will selected cultures grown on solid media. This will dramatically increase the need for BSC and present a bottleneck in the laboratory workflow.

Medium production facilities for bacterial cultures are no longer commonly available in hospital laboratories. In the event of a sudden increase in the need for microbiology supplies, manufacturers are limited in their ability to rapidly increase the production of media and reagents. In addition to the need to increase production, media must pass quality control and sterility tests prior to their release for use in clinical laboratories, further delaying their availability. Once media and reagents have passed appropriate tests, there will be difficulty in obtaining them from suppliers during a period of high demand. This lack of appropriate industrial surge capacity might present major difficulties both in the response to a BT event and in the ability to continue to provide standard laboratory services. Even during steady-state demand, it is common for some microbiology supplies to be on “back order” due to quality control or sterility problems at the level of the manufacturer. Microbiology laboratories may be able to change to a different vendor in such a setting, but this would not be feasible during a period of dramatically increased demand, when there would be national shortages from all vendors.

One area in particular that will be problematic is that of blood cultures. The great majority of the blood culture market is supplied by three companies, each of which manufactures a continuously monitoring blood culture instrument. Each instrument is able to incubate only blood culture bottles made by the manufacturer of that instrument. These instruments have a defined maximum capacity, e.g., the ability to incubate 240 blood culture bottles.

At Boston Medical Center, based upon the average number of bottles typically in incubation and the capacity of the instrument, an increase in the number of blood cultures by approximately 2 (one set) per hour or a daily increase in the number of patients by 12, each of whom has two sets of blood cultures drawn, would be sufficient to cause the hospital's blood culture instrument to reach its capacity. If the capacity of the instruments within the laboratory is exceeded, medical technologists will have to manually inspect each blood culture bottle that is incubated outside the instrument. In addition, each bottle will need either manual subcultures, a Gram stain, or an acridine orange stain of the blood culture broth smear (9), adding greatly to the labor requirement of an already-burdened laboratory. Each time blood culture broth is removed from a bottle, the procedure must be performed in a BSC. Depending on the volume of work, there is the potential that hundreds of bottles would require manual subcultures, dramatically increasing the use of BSC and interfering with the ability of a technician to perform other laboratory procedures that require their use, such as work with mycobacteria, fungi, and viruses. Continuously monitoring blood culture devices have been shown to detect positive blood cultures more rapidly than manual and semiautomated blood culture systems (18, 19, 26). Thus, if the number of blood culture bottles being cultured exceeds the available capacity of the instruments in the laboratory, there is likely to be not only an increased workload but also a further delay in the time to detection of bacteremia.

Specific regulations for the packaging and shipping of biohazardous materials exist, including the Dangerous Goods Regulations of the International Air Transport Association, part 72 of title 42 (Public Health Service) of the U.S. Code of Federal Regulations (CFR) on the interstate transportation of etiologic agents, parts 171 to 178 of title 49 (Department of Transportation) of the CFR on hazardous materials regulations, and part 1910.1030 of title 29 (Occupational Health and Safety Administration) of the CFR on occupational exposure to blood-borne pathogens (3). When a hospital microbiology laboratory is unable to rule out the possibility that a bacterial isolate is a BT agent, packaging and shipping the isolate to an LRN laboratory with greater expertise will be required. Packaging and labeling of the isolated organism in compliance with regulations is a time-consuming task. If many such isolates are shipped, there will be a significant impact on the workload of the clinical microbiology laboratory. In addition, as laboratories do not typically stockpile large quantities of the materials used for packaging, it is likely that some laboratories will run out of the necessary packaging materials.

DISCUSSION

The national antibiotic stockpile includes millions of doses of antibiotics, the majority of which are maintained in vendor-managed warehouses. Federal contracts with pharmaceutical companies and with suppliers ensure that the antibiotics are maintained and rotated in a systematic manner that takes into account the expiration dates of the products. Blood culture bottles, which are essential in the setting of a BT incident due to a bacterial agent and have a much longer shelf life than do solid media, should be included in the national antibiotic stockpile to help prevent the near certainty of a shortage in the setting of a BT event. Since the great majority of blood culture bottles in the United States are no longer manually examined, consideration should be given to stockpiling instruments that are compatible with the blood culture bottles, though this might be prohibitively expensive. The alternative, stockpiling of manual blood culture systems, would be of less help due to the need for manual examination of the bottles by technologists in the setting of a technologist shortage. In addition, federal support is needed to ensure that industrial surge capacity is adequate for critical items. While it is prohibitively expensive for manufacturers to maintain underutilized production facilities, it is essential that those companies that manufacture critical laboratory supplies are capable of rapidly scaling up production when more supplies are needed.

An additional intervention is to use local hospital data in order to validate a decrease in the time of blood culture incubation from 5 to 4 days in the event of a surge in blood culture demand. A 4-day incubation period has been shown to be acceptable for each of the three major blood culture instruments (7, 8, 15), though a shortened incubation time would not be acceptable if the BT agent was either F. tularensis or Brucella species. For example, although blood cultures have been flagged as positive in <11 h by a continuously monitoring blood culture system for a patient with fatal F. tularensis bacteremia (21), the standard recommendation when F. tularensis is suspected is to hold cultures for a minimum of 14 days and to perform blind subcultures onto an enriched medium every 3 to 5 days, even with continuously monitoring blood culture systems (9).

The shortage of qualified medical technologists is not merely a labor issue; it is a public health and biodefense issue. In the long term, surge capacity can be increased by the provision of federal funding to bring the salaries of medical technologists into line with those of similarly educated health care professionals. Plans could include the provision of financial incentives for students to enroll in college programs in clinical laboratory science and the funding of initiatives to increase the ability of accredited programs to increase enrollment. In the short term, laboratories should cross-train a number of medical technologists working in other laboratory sections so that they are able to perform some of the basic support tasks required in clinical microbiology, such as packing biohazardous materials for shipping. In this way, in the event of a sudden and dramatic need for additional medical technologists in the clinical microbiology laboratory, trained personnel will be immediately available to meet some of the increase in workload. Hospitals will have to hire additional medical technologists, since the process of training removes one technologist from productive work.

In order to provide increased numbers of BSC, hospitals could dedicate additional space to clinical microbiology laboratories and modify existing airflow in rooms in which the BSC will be placed. This expense cannot readily be borne by hospitals, many of which are operating with deficits, in the current competitive environment. Public funds might well be needed in order to prepare hospitals to meet the needs that are likely to be encountered during a BT event.

Clinical microbiology laboratories need to ensure that they have an adequate supply of packaging materials to meet the anticipated shipping needs in case of a BT event.

If chain-of-custody tracking is established as the norm for level A (sentinel) laboratories in the setting of a suspected or established BT event, the large number of requests for chain-of-custody information that would occur requires that an institutional plan be established in advance. In many laboratories, the clinical chemistry laboratory has the greatest expertise in chain-of-custody tracking as a result of its handling of specimens submitted for drug testing. Other areas of the hospital, such as the emergency department, also have experience with chain-of-custody tracking. This expertise should be tapped in planning for the large number of anticipated specimens and, if possible, in staffing the microbiology laboratory to ensure that proper chain-of-custody procedures are followed.

In summary, there are a number of challenges that must be met in order to have adequate surge capacity in hospital and other LRN level A (sentinel) clinical microbiology laboratories. Solutions require working with industry and earmarking federal funds to ensure that culture media and diagnostic reagents can rapidly be made available, either by incorporating them into the national antibiotic stockpile or by increasing industrial surge capacity; increasing the pool of qualified medical technologists to alleviate the national shortage; instituting limited cross-training of medical technologists who work in laboratories other than microbiology laboratories; and ensuring that hospitals have an adequate number of BSC. This objective will be accomplished only if there is both recognition of the problem of limited surge capacity in clinical microbiology laboratories and the political will to solve the problem.