How do you monitor swine flu?

Yesterday, reporters descended on Whitehall to hear about the methods of influenza monitoring. Sir Liam Donaldson, the Chief Medical Officer led the numbers feast alongside Justin McCracken, chief executive of the Health Protection Agency, and Ian Dalton, the new flu czar.

Sir Liam explained what data are being tracked: total number of cases, antiviral usage, hospitalisations and deaths (see pyramid, below).

Hospitalisations

The number of hospitalised patients as of 8am on Wednesday was 335. 107 of those were under the age of 16, while only 39 were over the age of 65. 43 were in critical care.

When is the flu not the flu?

So, now that we’re not testing every case of suspected swine flu, how do you know what the total number of cases really is? Measuring the number of cases seen by GPs (and NHS direct) is just the tip of the iceberg (see below), but is seen as the best way we have.

GP consultation estimates are made by the RCGP’s influenza monitoring unit. Around 100 general practices throughout England provide data on consultation rates for patients with flu-like illness. As the diagram below (courtesy of DoH/RCGP) shows, consultation rates over the last few weeks have shot up from the baseline level to a rate normally seen in seasonal flu. The most recent national average is 51.9 consultations per 100,000 of the population per week, which we are told is equivalent to about 27,000 consultations a week. There’s a lot of regional variation though, with the rate in London and the West Midlands being well over 100.

But does 27,000 consultations by people with flu-like illnesses in a week mean 27,000 new swine flu sufferers? There seems to have been little said so far during the pandemic about other causes of flu-like symptoms. With seasonal flu not around at the moment, you could be forgiven for assuming that almost all people with the right symptoms have swine flu. But that doesn’t seem to be the case: only 28% of those with these symptoms actually have a positive swab for H1N1 according to McCracken. So over 70% of people being told they have swine flu probably don’t have anything more exciting than a cold.

Consultation rates: skating on thin ice?

Is it possible that in this particular iceberg you wouldn’t have to dig very deep to find the bottom? The graph above, which compares consultation rates year on year, shows that GPs are seeing more patients with flu than at any point in 2007/08. But does that mean that swine flu is already worse than seasonal flu or is causing more illness? Probably not. Consultation rates depend on many factors, not just the underlying prevalence. Public health messages (from the media and government) can increase consultation rates, as we’ve seen in the surge in cervical screening uptake following Jade Goody’s death. New treatments can too (like sildenafil, although the surge in demand wasn’t quite as dramatic as expected). With swine flu both of these factors are present. In a seasonal flu outbreak flu gets barely a mention from the media and the vast majority of people with flu-like symptoms don’t see their doctor as there is very little they can do. Contrast that with the current outbreak, where every person with flu-like symptoms is, for the moment at least, being advised to call their GP so they can get their dose of antivirals. It’s not surprising that consultation rates are high, but how does the actual prevalence of H1N1 compare with that of a seasonal influenza outbreak?

Sneeze and Click update

When is the national flu pandemic service (sneeze and click) going to launch? Sir Liam wouldn’t give a date yesterday, but did say that it depends on the number of cases across the country and how much pressure existing services are under. To measure that we now have a new toy called the NHS pressure rating. It’s basically a map of England. At the moment it’s green, but as the pressure mounts, parts will begin to turn yellow, then red. Details on what factors will contribute to the rating were a bit sketchy, so if anyone has any ideas do chip in.
Tom Nolan is the clinical community editor of doc2doc, the BMJ’s professional networking community.