Brace yourself – swine flu is on the rise again. The second wave coincides with mounting disquiet among doctors about the way we as a nation are responding to the disease. A recent survey by Pulse, a leading GP periodical, found that 61 per cent of family doctors believe the government should review its policy of blanket provision of Tamiflu to all suspected sufferers. It also reported a growing number of cases that have resulted in death or serious harm after conditions such as meningitis have been misdiagnosed as flu, and wrongly treated.
Swine flu has wrong-footed everyone. For many years it was assumed that when the next pandemic arrived, it would create havoc: thousands would die in Britain alone; the exponential demand for services would be mirrored by the decreasing number of healthy doctors able to deliver them; societies would crumble. This projection was based in part on the 1918 global flu pandemic – which caused an estimated 50 to 100 million deaths – coupled with the high fatality rate among the few hundred cases of “bird flu” (H5N1) over the past few years.
In the UK, huge quantities of antiviral drugs were stockpiled, plans for establishing a national pandemic flu phone line were laid, and organisations both public and private were exhorted to draft detailed contingency plans for when the carnage began. Both chambers loaded, the shotgun was trained on the far horizon, the collective eye of the viral surveillance world squinting down the barrel, seeking out the first sign of an emergent threat.
Along came swine flu (H1N1). To be clear: people have died – 82 in the UK and an estimated 4,041 worldwide, at the time of writing. According to the World Health Organisation (WHO), roughly 60 per cent of the severe or fatal cases have occurred in recognised risk groups, such as people with grave underlying health problems; the remaining 40 per cent have affected fit children and adults. Each death is tragic, but the mortality figures are tiny, viewed against the millions of mild cases globally.
Writing in the British Medical Journal in September, Peter Doshi, a doctoral student at the Massachusetts Institute of Technology, proposed a framework to differentiate between disease patterns. Type 1 infections are widespread and severe – exactly the kind towards which the current pandemic flu plans are geared. But swine flu sits in Doshi’s type 3 – widespread but usually mild. These different beasts present different challenges, but because pandemic planning failed to anticipate anything other than a type 1 scenario, swine flu has triggered a completely inappropriate response.
Marginal effects
Central to doctors’ unease is the National Pandemic Flu Service (NPFS). Telephone-based and web-based assessments of symptoms are made using a computer algorithm. If the computer says “swine flu”, patients are given antiviral medication, usually Tamiflu. The problems are twofold.
First, data from the Health Protection Agency, which conducts confirmatory laboratory tests on a sample of cases, shows that less than 10 per cent of NPFS diagnoses are correct. The other 90 per cent of supposed swine flu sufferers have other illnesses whose symptoms happen to overlap. (Doctors fare only slightly better, diagnosing at best a quarter of cases accurately.) Second, the drugs being doled out are by and large worse than useless, even in the correctly diagnosed cases. Tamiflu makes only a marginal difference to the course of uncomplicated flu and causes side effects in up to 40 per cent of people who take it. On the whole, these are just bothersome – vomiting is the most frequent – but serious, even fatally adverse reactions do occasionally occur.
In late August, WHO advised that antivirals were not necessary for fit patients suffering from uncomplicated swine flu. The Department of Health defends its continued policy of Tamiflu-for-all on “safety first” grounds. Its concern is those exceptional cases of severe disease in fit individuals, where Tamiflu might (no one knows for sure) make a difference if started early. However, the logic – that it is better to treat everyone than risk missing those who might have benefited – belongs to a strategy for dealing with a Doshi type 1 pandemic, where the chance of severe disease is high. For swine flu, damage from indiscriminate use of antivirals outweighs the supposed benefits of catching atypical cases early.
The real challenge of a Doshi type 3 pandemic is identifying the small minority who actually and urgently require help. These could be those rare individuals with severe flu, or they might be patients in the early stages of another serious disease such as meningitis. Doctors are good at doing this (though far from infallible) and many GPs believe that only medically qualified staff should be undertaking flu assessments, but this is not achievable, given our capacity for mass hysteria. During the first wave of swine flu – before the NPFS was launched – the NHS front line was overwhelmed by waves of worried callers in flu hot spots. One of the main reasons for setting up the NPFS was to prevent a meltdown in services, but it should now change its focus from diagnosing swine flu (at which it is hopeless) to identifying patients in trouble.
The trickiest problem is when patients who initially feel mildly unwell start to deteriorate. The Department of Health stresses that patients are advised to consult a doctor if they get worse, but this fails to appreciate the Tamiflu effect. Having been “diagnosed” with swine flu and put on antivirals, patients are then falsely reassured that appropriate treatment is under way. By stopping its blanket use of Tamiflu, the NPFS would greatly increase the likelihood of patients consulting a doctor if they deteriorate.
Lessons in planning
Swine flu may in time be seen as a great learning opportunity. It has exposed a rigidity in pandemic planning that needs urgent correction. WHO has a scale to denote the spread of infection – level six being pandemic. It needs to develop a simple, parallel system to differentiate between Doshi types, one that should trigger responses appropriate to the particular challenge posed.
For now, the Department of Health should stop sticking doggedly to contingencies laid against a very different threat. The public and the NHS need clear identification of the at-risk groups and a message that the danger of swine flu, for everyone else, is almost certain to be minimal. They also need information about the warning signs of a more serious problem. The vast stocks of antivirals should be left on the shelf to go quietly out of date.
Yet it may be too late. Another facet of the Tamiflu effect is that we have educated hordes of people that what they thought felt like just a bad cold (and, most of the time, was just a bad cold) needed treatment with powerful drugs involving mystical rituals with a special authorisation number and a flu friend. Doctors will be dealing with mass hysteria in the face of minor illness for some time to come.
Phil Whitaker is a doctor and novelist. He is currently working on his fifth novel, “Sister Sebastian’s Library”
Computer says flu
What do malaria, meningitis, diabetic coma, leukaemia and appendicitis have in common? They are just a few of the conditions that were originally diagnosed as swine flu during the first wave of the pandemic. Most patients have lived to tell the tale; some have not. The case reports have been appearing in the letters pages of medical journals, and on discussion forums of networking sites for doctors. The GPs reporting them are frequently unsure who in authority should be informed.
In fact, the National Patient Safety Agency (NPSA) has been tasked by the Department of Health with investigating alleged misdiagnoses. In a statement, John Scarpello, NPSA deputy medical director, confirmed that the agency had received “a small number of reports where swine flu may have been misdiagnosed”, but was unwilling to go into detail while the facts had not been established. Given that the reporting system is entirely voluntary, and few clinicians know to contact the NPSA, this “small number” of reports is likely to be the tip of an iceberg.
The National Pandemic Flu Service is a first: never before have patients been diagnosed by computer or unqualified call-centre staff. There is a real need for research to examine this approach, yet none appears to be planned. The Department of Health refers inquiries about patient safety to the NPSA, while the NPSA believes commissioning such a study would be outside its capacity and brief.
Phil Whitaker