Abstract
Experience from 25 years of clinical monitoring of influenza and related illnesses has been considered in relation to the prediction of epidemics of influenza. The monitoring has taken place in a network of sentinel practices in England and Wales who report new episodes of illness weekly as they occur. The practice data are aggregated at the Birmingham Research Unit of the Royal College of General Practitioners and weekly rates per 100,000 population are published in age-specific groups. Major epidemics of respiratory infection are heralded by an increase in the weekly rates for influenza and influenza-like illness of 30 per 100,000 and by increase in the rate for all acute respiratory illness of 150 per 100,000. Increases of this magnitude have been consistently associated with major epidemics of respiratory illness which in turn cause substantial increases in deaths. When epidemics occur around the turn of the year, respiratory syncytial virus has to be considered as a possible cause but otherwise epidemics of this magnitude are invariably due to infection by Influenza A virus.