Summary. Evidence on the extent of the part played by infected hens' eggs in causing salmonella food-poisoning is inconclusive. The role of freshly cooked shell eggs is currently much exaggerated. Prevention should be sought through improved catering practices and kitchen hygiene, and attempts to eradicate salmonellas from laying flocks are likely to be ineffective.

Despite increased resources devoted to food‐poisoning prevention, reported incidence of food poisoning continues to rise. Improvements in prevention strategies might therefore be necessary and there may also be opportunities for reducing the burden of regulatory control. Suggests that, of the two components of prevention, surveillance and control, control activities are returned to the private sector, allowing public sector agencies to concentrate on surveillance, the precursor to effective control. The Offices, Shops and Railways Premises Act, Hoists and Lifts Regulations model is offered as a means of devolving regulatory control.

Describes a study to measure the quality of service provided by food‐poisoning surveillance agencies in England and Wales in terms of the requirements of a representative consumer ‐ the egg producing industry ‐ adopting “egg associated” outbreak investigation reports as the reference output. Defines and makes use of four primary performance indicators: accessibility of information; completeness of evidence supplied in food‐poisoning outbreak investigation reports as to the sources of infection in “egg‐associated” outbreaks; timeliness of information published; and utility of information and advice aimed at preventing or controlling food poisoning. Finds that quality expectations in each parameter measured are not met. Examines reasons why surveillance agencies have not delivered the quality demanded. Makes use of detailed case studies to illustrate inadequacies of current practice. Attributes failure to deliver “accessibility” to a lack of recognition on the status or nature of “consumers”, combined with a self‐maintenance motivation of the part of the surveillance agencies. Finds that failures to deliver “completeness” and “utility” may result from the same defects which give rise to the lack of “accessibility” in that, failing to recognize the consumers of a public service for what they are, the agencies feel no need to provide them with the data they require. The research indicates that self‐maintenance by scientific epidemiologists may introduce biases which when combined with a politically inspired need to transfer responsibility for food‐poisoning outbreaks, skew the conduct of investigations and their conclusions. Contends that this is compounded by serious and multiple inadequacies in the conduct of investigations, arising at least in part from the lack of training and relative inexperience of investigators, the whole conditioned by interdisciplinary rivalry between the professional groups staffing the different agencies. Finds that in addition failures to exploit or develop epidemiological technologies has affected the ability of investigators to resolve the uncertainties identified. Makes recommendations directed at improving the performance of the surveillance agencies which, if adopted will substantially enhance food poisoning control efforts.

The Editors of Letters in Applied Microbiology will, at their discretion, publish invited and submitted ‘Opinions’ on subjects in the general area of Applied Microbiology. They will not be subjected to the normal refereeing procedures and reprints will not be provided. The ‘Opinions’ will not necessarily represent the views of the Society for Applied Bacteriology or of the Editors. The Editors may invite or readers may submit ‘Responses’ to published ‘Opinions’, provided that they are not merely polemics, and these will also be published at the discretion of the Editors. ‘Responses’ will be treated precisely like ‘Opinions’. They should clearly indicate, in their first sentence, the ‘Opinion’ to which they are a response.

business matters only to ensure that doctors are fully informed of all the facts about particular issues before they consider speaking publicly about them. These specific policies have not replaced, and are not in any conffict with, paragraph 330 and have been agreed with all our consultants and staff through their negotiating committee in consultation with the BMA. The health of any organisation depends in large measure on its ability to be honest with itself, and we commend our policy of frankness and accountability to others.