EDITOR—We challenge the method that Donaldson and Keatinge used to calculate mortality due to influenza.1 Data covering the period 1970-99 were used, but the authors state that the regression analysis started on 1 January 1990. On the basis of the 10 year analysis, they estimate an average of 1265 per million excess winter deaths from all causes, equivalent to 67 000 nationally in England and Wales. They say that 2.4% of this excess (equivalent to a national average of 1620 deaths per year) is attributable to influenza.. This estimate contrasts with those obtained by other groups: Tillett et al estimated an annual average of 12 000 deaths (1968-9 to1977-8)2, Nicholson 13 800 (1975-6 to1989-90)3, and Fleming 12 500 (1989-90 to 1998-9).4 The estimate of 12 500 is equivalent to 19% of the 67 000 total excess winter deaths, rather than 2.4%.

Donaldson and Keatinge estimated total deaths attributable …