In Science, it is important to use words in very precise ways. This leads to specialised vocabularies which may be difficult for laymen to follow but a godsend to pseudoscientists who can either use real terms incorrectly or simply make up their own. However, nothing is more dangerous than for scientists to misuse words in their publications and then spread this misinformation to the press.
And that is exactly what has happened in a paper recently published in PLoS Medicine by a group of scientists working with Dr. Marc Lipsitch.
There are several ways to determine if someone is infected with the new H1N1 virus. Real time PCR is the most commonly used, but virus can also be cultured and sequenced. [Flu A tests fail about 50% of the time and are not considered reliable]. Seroprevalence studies could be done to determine whether people have been infected with H1N1 in the past by looking for antibodies to it, but no results from such studies have been released.
Influenza-like illness (ILI) is a dumping ground for a wide variety symptoms and diseases. Although some people with ILI actually have influenza, most do not. This is well known and has been experimentally verified, as I discussed previously – Estimates of case fatality rates based on influenza-like illness are wrong. I had written that blog in response to statements to the press by Dr. Lipsitch back in September. Apparently, those comments in September relate to the paper which has just now appeared in PLoS Medicine. As I pointed out back in September, an empirical study done in Sweden found that only 5% of ILI cases actually had pandemic H1N1. Most were infected with rhinoviruses or had other diseases. So, it should be clear to anyone that ILI does not equal an H1N1 case.
Should be, but apparently isn’t.
From the Lipsitch study:
We estimate the severity of pH1N1 infection from data from spring–summer 2009 wave of infections in the United States. The New York City and Milwaukee health departments pursued differing surveillance strategies that provided high-quality data on complementary aspects of pH1N1 infection severity, with Milwaukee documenting medically attended cases and hospitalizations, and New York documenting hospitalizations, ICU/ventilation use, and fatalities. These are the numerators of the ratios of interest. The denominator for these ratios is the number of symptomatic pH1N1 cases in a population, which cannot be assessed directly. We use two different approaches to estimate this quantity. In the first (Approach 1), we use self-reported rates of patients seeking medical attention for ILI from several CDC investigations to estimate the number of symptomatic cases from the number of medically attended cases, which are estimated from data from Milwaukee. In the second (Approach 2), we use self-reported incidence of ILI in New York City, and making the assumption that these ILI cases represent the true denominator of symptomatic cases, we directly estimate the ratio between hospitalizations, ICU admissions/mechanical ventilation, and deaths (adjusting for ascertainment) in New York City.
Given that we already know that in one empirical study 95% of ILI cases were not due to influenza, it is obvious that the Lipsitch study has made a fatally incorrect assumption. The fact that the rest of the study involves complicated statistics is irrelevant (although quite impressive to gullible reporters). If your assumptions are wrong, your model cannot produce valid results.
Some may forgive Dr. Lipsitch his extensive (and quite wrong) assumptions on the grounds that that is the best that we can do. However, this is simply not true. We could test vast numbers of people with high-throughput Real Time PCR, if the CDC wished to (or perhaps, knew how to). We could have done seroprevalence tests months ago, if the CDC wished to (or perhaps knew how to). The decision not to gather actual, real data was a choice on the part of the CDC, not a limitation in technology, as Dr. Lipsitch implies.
We should not blame Dr. Lipsitch for the CDC’s failings. However, he can be blamed for grossly overinterpreting his own data:
From, The Age, December 9, 2009
‘I think it is very likely to be the mildest pandemic on record,” said Marc Lipsitch, a professor of epidemiology at the Harvard School of Public Health, who led a federally funded analysis with researchers at the Centres for Disease Control and Prevention (CDC).
Such a broad statement about a new virus based on such a shaky foundation (actually, no foundation at all) would be inappropriate at any time. Making this statement in the middle of a pandemic is irresponsible.
So far over 500 children are estimated to have died in the US from this virus. Doomed pregnant women on ventilators have had emergency C-sections to save their babies. Teachers and health care workers are dying.
And flu season is just beginning.
I don’t claim to know the future. I do know that those who cannot tell the difference between ILI and an actual case of H1N1 have no business talking to reporters about what is going to happen.
No business at all.
Presanis et al. (2009) The Severity of Pandemic H1N1 Influenza in the United States, from April to July 2009: A Bayesian Analysis. PLoS Medicine.