Having the flu is a draining experience. The symptoms overtake the body quickly and in no discernible order. Appetite disappears; during the day you feel exhausted, but at night you have trouble sleeping. Your stomach aches, badly. Your head is pounding and your throat is scratchy. Many sufferers feel like they could vomit at any moment.
For healthy adults, these symptoms are manageable; often a simple regimen of bed rest, paracetamol, warmth and plenty of water will diminish the condition after a few days. Children and the elderly, though, are far more vulnerable. In the period between October 2017 and January 2018 in the UK, 155 people died of the condition, according to Public Health England. It was the highest death toll in four years.
“The unpredictability of the virus remains our largest challenge,” says Dr Emily Toth Martin, an assistant professor in epidemiology at the University of Michigan School of Public Medicine. "We don’t know what strain to be ready for each year, and each [one] has some different patterns for severity and who will be affected. We also can’t predict exactly when influenza season will start and end each year, so it is difficult to mobilise resources when flu season is earlier than normal.”
Often, the only clues that medical authorities in Europe and North America can use to prepare themselves have to be sourced from the other side of the world. “We did have a sense from the data coming out of the southern hemisphere, and particularly Australia and New Zealand, that we were expecting a predominant H3N2 season this year, which was the case,” says Dr Jesse Papenburg, a paediatric infectious disease specialist and microbiologist at Montreal Children’s Hospital. A subtype of influenza A, “we were expecting that the vaccine efficacy would not be very good, which again, seemed to be the case”.
In the absence of an effective vaccine, clinicians can also fall back on diagnostics to help stall the spread of the virus, the logic being that those people who have confirmed cases of influenza can receive effective treatment more quickly, thereby helping to reduce patient load. Until recently, however, clinicians haven’t had access to influenza diagnostic tests that can provide results in a timely manner, often relying on results sent back from dedicated labs. And during a particularly bad outbreak, these resources can be stretched to their absolute limit.
– Dr Emily Toth Martin
“Public health [authorities] during the 2009 pandemic, at one point, did discourage testing for patients that were not severe enough to be hospitalised… it was just taking up too many resources in different jurisdictions,” says Papenburg.
Even in a smoothly run hospital that uses traditional polymerase chain reaction (PRC) testing for influenza detection, it can take several hours to return assays. “I can tell you that in our institution, for instance, the lab-based testing [has] a median turnaround time of about eight hours during influenza season,” explains Papenburg, including time lost to transportation and batching. “Five of those hours are really just the analytical time of how long it takes to take the specimen, extract the nucleic acid and then do all the thermal cycling for the PCR.”
Virus hunting
In recent years, however, several pointof- care tests have become available to clinicians. Some of the most common are rapid influenza diagnostic tests (RIDTs), which are cost-effective and can be used outside the laboratory. These immunoassays work by attempting to detect influenza antigens in respiratory secretion samples. They’re of a similar nature, says Papenburg, to home pregnancy tests. “You see a positive control line and [if] your second line for the specimen appears positive, you know that there’s influenza A or B in the specimen,” he explains. “And that’s read visually by the person who’s doing the test.”
When they emerged in the 1990s, these tests were so simple to operate that they were granted waivers by the US FDA to be used by clinicians outside of the laboratory. People quickly realised, however, that their sensitivity was fairly poor. “Rapid tests are fast, easy to perform and often don’t require expensive equipment,” says Martin. “However, these tests are not as accurate as PCR-based tests, and have been shown to miss out a lot of patients who have influenza.”
More reliable successors to RIDTs, called digital immunoassays (DIAs), have since come onto the market, in time to comply with new US FDA regulations set to come into force later this year that ban the use of tests that fail to achieve a false-positive rate of lower than 20%. These types of diagnostics contain “a similar mechanism to the traditional rapid influenza diagnostic tests”, says Papenburg, but saw their performance improve thanks to “the fact that there is a machine or an instrument that reads the presence of the antigen”. This, in effect, removed a lot of the subjectivity inherent in the use of older RIDTs.
Clinicians also have the option of commissioning molecular pointof- care tests. Designed to detect the presence of viral nucleic acid in samples, until recently they were confined to laboratory settings. “There’s really been kind of a revolution in recent years, in terms of taking molecular diagnostic tools, making them simple enough to be used outside the lab, and now available for use at the point of care,” says Papenburg.
Out with the old
What’s more, there is evidence to suggest that the sensitivity of DIAs and molecular POC tests significantly outperforms that of the older RIDTs. In a systematic review and metaanalysis of 162 studies published in the Annals of Internal Medicine in 2017 looking into the effectiveness of RIDTs, including newer digital immunoassays, versus molecular POC tests, researchers – Papenburg among them – concluded that older RIDTs were easily outclassed in several segments.
“We showed that, first of all the older tests really don’t perform very well,” he explains, with sensitivities hovering just under 35% for adult cases. “It’s because adults have seen more influenza viruses over their lifetimes, compared with children, and have some level of cross-reactive immunity against these different strains, and therefore shed virus in smaller amounts and for a shorter period of time.”
The digital immunoassays, meanwhile, were found to have a minimum sensitivity of 80% for the detection of influenza A and 76.8% for influenza B. The molecular tests also attained 91.6 and 95.4% sensitivity for the detection of influenza A and B respectively. Even so, there remain several significant impediments to their practical adoption across Europe and North America.
“The main drawbacks of these newer molecular diagnostic tests, and even the digital immunoassays, are twofold,” explains Papenburg. “One is that you now need an instrument to be able to run the test, whereas the older rapid tests, as lousy as they were in terms of sensitivity, were free-standing. Now you have to buy or lease an instrument for the digital immunoassays and the same thing for the point-of-care molecular diagnostic tests. And each test itself is a greater cost compared with the older rapid tests.”
Martin shares these concerns. “Too many platforms also rely on expensive machines dedicated to a single testing platform,” she says. “These machines are often prohibitively expensive for a community medical practice.”
– Dr Jesse Papenburg
Indeed, some molecular-based tests can cost up to $50 each, twice as much as many of the older RIDTs. They can also take up to half an hour to produce a result, which probably isn’t quick enough for the dozens of sneezing patients quietly waiting to be seen in their GP's waiting room.
“If molecular POC tests can become fast enough to use in the time of a normal doctor’s visit, it would allow the accurate diagnosis of flu in the doctor’s office,” says Martin. “Antivirals need to be started as soon as possible to be most effective and accurate rapid testing would help with this.”
Papenburg certainly believes that it won’t be long before those conditions are met. “I think things are going to continue to improve, now that we’ve attained really high sensitivities for some of these assays and that they’re simple enough to be used at the point of care,” he says. “People are going to be looking at making them even faster, so that we’re not waiting half an hour for a result.”
Added effectiveness
The provision of more accurate, costeffective and timely point of care tests could also boost the efficiency of antiviral prescriptions. “We know that, with antivirals, the sooner they get prescribed in patients with influenza, the greater the beneficial effect that they have, not only in outpatients, but also in hospitalised patients,” explains Papenburg. “There is evidence to show that antivirals, when prescribed early in patients sick enough to be hospitalised, can reduce severe outcomes in ICU admissions and even [reduce] mortality.”
In time, this could also push down unnecessary antibiotic use in outpatient clinics and emergency departments. “That should reduce the risk that [a] patient has a bacterial infection causing their acute febrile respiratory illness,” he says. This could potentially “reduce unnecessary antibiotic use, thereby reducing exposure to these antibiotics that we know can cause side effects [like] C. difficile associated diarrhoea and the emergence of resistance, not only in the individual but population levels.”
It is a hypothesis that, admits Papenburg, remains unproven. Even so, he remains hopeful that increased use of point-of-care diagnostics tests will provide an answer soon “It’s a question that still remains somewhat up in the air,” says the microbiologist. “We need to do more research on the clinical impact if their use in different settings, such as outpatient clinics and emergency departments.”