Public health officials continue to be on high alert for the arrival of the avian flu in the U.S. , as international flu watchers have already documented the presence of the H5N1 strain in China and Turkey , and more recently, Iraq . With flu season underway, public health labs are working to identify the various strains of flu making their way through the U.S. However, in the event of a pandemic avian flu, clinical laboratorians will be called on to work in concert with state and government labs to monitor flu strains through rapid identification of H5N1 specimens. This month's Strategies looks at recent initiatives by the Centers for Disease Control and Prevention (CDC) to improve communication between the government and the clinical laboratory community in order to enhance the response capabilities to public health threats, as well as the agency's current H5N1 testing guidelines and some concerns about rapid testing tools.
Since February 2004, the CDC has been conducting enhanced H5N1 surveillance in order to maximize the agency's chances of detecting the strain's arrival in the U.S. as quickly as possible, explained Lynette Brammer, MPH, an epidemiologist at the CDC's influenza surveillance branch, during a recent teleconference conducted by the agency, “What Clinical Laboratories Need to Know About Their Role in Influenza Testing.”
During times of routine surveillance, each state has a variety of sources for collecting influenza specimens through hospital and physician office labs. These specimens are then sent to the state's public health labs, where PCR testing is done to determine influenza type and subtype. According to the current protocol, public health labs isolate a subset of these viruses and send it to the CDC, where the surveillance labs do additional antigenic characterization, as well as antiviral resistance testing. This surveillance information is then used for selecting a vaccine strain to be used during next year's flu season.
Even though the CDC is conducting enhanced surveillance for the H5N1 strain, laboratorians shouldn't recommend that the flu strain of every patient presenting flu-like symptoms be determined. Instead, laboratories should adhere to the current case definition: the patient must have traveled to an H5N1-infected country, which includes countries with either confirmed human infections or poultry outbreaks. In addition, the patient must meet a clinical description for an influenza-like illness, meaning fever, cough, and sore throat.
“In order for us not to be overwhelmed by significant numbers of suspected cases, we want to limit the case definitions to those with pneumonia and requiring hospitalization,” explained Niranjan Bhat, MD, Medical Epidemiologist with the CDC's influenza branch, in the teleconference. “Lastly, the risk behavior associated with having an exposure to poultry in the affected country is also very important. All of these criteria are useful to focusing our resources in those few travelers who have had a real chance of being H5 infected.”
During the teleconference, the CDC also announced the formation of the Laboratory Outreach Communication System (LOCS). While still in the preliminary stages, LOCS will be the central source for laboratory professionals to get credible information on routine and emergent issues, like the avian flu threat. The structure of this system has yet to be determined, but CDC officials envision it as a volunteer communication network for the exchange of lab-related information between the CDC and the clinical laboratory community.
“From recent experience with a number of public health threats, including natural disasters such as Katrina, it is clear that public health needs a way to rapidly communicate with clinical laboratories,” said Rex Astles, PhD, Senior Health Scientist, CDC Division of Laboratory Systems. “LOCS will be one way to help meet that need. It is important to acknowledge that the LOCS initiative is rapidly evolving to best meet the needs of clinical laboratories and the organizations that represent them. The LOCS concept isn't firmly crystallized yet, but we intend to work with the organizations to determine needs and how we can best serve them.”
The program's efforts will initially focus on two areas: rapid communications with clinical lab professionals, primarily through their organizations; and when appropriate, LOCS-hosted interactive sessions between the CDC and the clinical laboratory community. Other communication ideas under consideration include webinars and self-subscribing email lists for specific emails. One of the LOCS initiatives will be dissemination of information on the recent FDA approval of H5 Influenza A reagents for RT PCR testing in LRN-designated labs, which provides preliminary results on suspected H5 influenza samples within four hours, compared to the three hours it took previous tests to deliver results. Another LOCS program is a recent public health directive encouraging mothers to be screened for prenatal Hepatitis B infection. “This second transmission will emphasize the need to test and report cases to public health so that the number of cases of infant HBV infection will be minimized,” explained Astles.
Keeping Control of Testing Volumes
Currently, almost all known H5N1 confirmed cases have been severe infections involving the respiratory tract, with the vast majority of patients reporting exposure to poultry, said Bhat. As health officials learn more about the virus's behavior and how it changes, these components will likely change, with testing extended to those with a travel history, but who might only be presenting mild asymptomatic infections or nonrespiratory symptoms. But currently, it's important for clinical laboratorians to stick to the case definition so the testing volume remains manageable.
“To the extent that they can, clinical laboratories should try to apply the case definition, because as you imagine, if there's a cluster of suspected cases, like in Los Angeles, or if the first human cases are confirmed in the United States, clinical laboratories may be flooded with specimens from suspected patients, and these may included the ‘worried well' as well as other people with other respiratory disease,” said Bhat. Therefore, it's important for physicians as well as clinical laboratorians to employ a case definition in order to protect laboratory workers, prioritize specimens and testing, and also educate both the public as well as clinicians about when H5 specific testing is appropriate. “You can use the case definition to help educate the clinician, institute biosafety measures, and prioritize your testing should you receive a significant number at once,” he added.
If a patient meets a case definition, CDC officials are encouraging providers to send the specimen to their state lab, although they recognize that providers may prefer to use their affiliated clinical laboratories. “Whether a provider sends a specimen directly to a state public health lab or a positive result is sent from a clinical lab, the first line for a response should be by the state lab or the state epidemiologist, as they are in the best position to mobilize resources,” said Bhat.
Testing Kit Concerns
With clinical laboratories playing a major role in influenza diagnosis, many facilities are relying on the variety of rapid antigen detection tests, many of which deliver results within a half hour. There are currently 13 FDA-cleared test kits capable of detecting the influenza A virus, both the influenza A or B without identifying the virus type, as well as both influenza A or B and the type, according to Pete Schult, PhD, Director of Wisconsin State Laboratory of Hygiene's Communicable Disease Division, who also spoke at the CDC teleconference. Since many of these kits are CLIA-waived, they can be performed outside the laboratory. “The potential impact of these tests in clinical practice is obvious,” he said. “Less obvious perhaps is the impact these diagnostic tests may have on public health response both to seasonal influenza, specifically our annual influenza epidemic, and to emergence of novel strains which have the potential to cause the next pandemic.”
The advantages of these test kits—easy to perform, rapid turn around time, inexpensive—make them a powerful response tool. However, Schult pointed out some concerns relating to sensitivities and predictive values. While preliminary reports have indicated specificities are generally between 90–95%, the median sensitivity of some kits is 70–75% when compared to cell culture immunofluoresence. In addition, there have been preliminary reports of poor sensitivity to novel strains—like H5N1—compared to PCR testing. There appears to be little solution to the sensitivity issue at this time, because it's dependent on manufacturers making better kits. However, it does make specimen quality even more important, added Schult. “We can control specimen quality,” he explained. “The type of specimen we are using, the timing of collection, and the technique by which the specimen is collected, as well as the patients who we are going to use these tests on.”
In addition to sensitivity concerns, there are also problems with the test kits' predictive value—the probability of the presence or absence of disease given the results of the test. There are two types of predictive value—predictive value positive (PVP) and negative (PVN). These test kits have very poor PVP, or a high false positive rate, during periods of low prevalence in a typical flu season. Unfortunately, it is during these times when there is a high demand for the test. This could cause considerable problems for public health response if during a period of very low prevalence just following emergence of a novel subtype, which is when pandemic concerns would be heightened, according to Schult. PVP will improve, however, when these kits are used during periods of high flu prevalence.
In terms of PVN, relatively poor sensitivity will pose a problem the closer it gets to the peak of influenza activity. However, because of these reasons, PVN is good early in the flu season or off-season. Because of these predictive value issues, use of rapid test detection can be optimized by confirming early, late, and out-of-season positive results, as well as peak-season negatives. Prevalence indicators should also be considered, so that clinicians and laboratorians can decide when to test, when to qualify a result, and when to confirm a result. Of primary importance is what the patient is presenting, said Schult. “We need to educate our clinicians on predictive values and the limitations of test results,” he added. “We need to emphasize that these rapid test results need to be interpreted in the context of patient/clinical presentation and the current influenza epidemiological pattern.”
For more information:
Julie McDowell is the Editor of Strategies. She can be reached by email.
In the last issue of Strategies , “Advent of Blood Substitutes Raises Testing Interference Questions,” the original article misstated that HemoTech is scheduled for Phase III clinical trials this year, pending approval by the FDA. Officials with HemoTech's manufacturer, HemoBioTech Inc. ( Dallas , Texas ) are speaking with FDA officials about beginning Phase I clinical trials this year in the U.S. In addition, the product will likely not be commercially available until 2009, not 2008, as originally stated in the article. An updated version of the article is available online: http://online.aacc.org/AACC/publications/strategies/012706.htm to current issue.