Use of NHS Direct calls for surveillance of influenza--a second year's experience.

A second year's call data to NHS Direct are presented to evaluate their usefulness for influenza surveillance. During the winter of 2000-01, age-group-specific data relating to the 'cold/flu' algorithm were collected from six NHS Direct sites (population coverage: 16 million). The total number of calls was collected from all 23 NHS Direct sites on a daily basis. Despite the winter of 2000-01 having been a season of low activity for influenza in the United Kingdom, NHS Direct data demonstrated a peak in the 'cold/flu' calls as a proportion of the total calls (3.1% [672 'cold/flu' calls] during week 06/01). This coincided with the peak recorded by routine influenza surveillance systems. There was also an earlier peak in the proportion of 'cold/flu' calls (3.3% during weeks 52/00 [789 'cold/flu' calls] and 01/01 [749 'cold/flu' calls]) which may have been due to other respiratory infections, the lack of specificity of the definition of NHS Direct 'cold/flu' calls and an increase in 'out-of-hours' calls to NHS Direct at the time. Despite limitations, the timeliness of NHS Direct data, the total population coverage of the service and the ability to provide local information on 'cold/flu' calls make the call data suitable for further surveillance during the winter of 2001-02. It is hoped that as NHS Direct reaches a 'steady state' in terms of population coverage and uniformity of clinical support systems, it will be possible to begin to construct 'baselines' for the respiratory disease related call data.