To ask the Secretary of State for Health (1) pursuant to the answer of 26 February 2009, Official Report, column 1028W, on malnutrition, what the statistical margin of error would be if the General Practice Research Database sample were used to project a population-wide figure; (262545)

The General Practice Research Database (GPRD) is a collection of data from general practices selected to be as geographically demographically representative as possible of the UK. The age and gender distribution is similar to that as detailed by the Office for National Statistics on the basis of census data. GPRD practices are provided with recording guidelines against which data quality is monitored.

The GPRD contains records of patients recorded as suffering from malnutrition under a range of different terms, known as ‘Read Terms'. There are many thousands of Read codes in total, and 24 separate codes relating to malnutrition. This range of different codes is designed to allow the general practitioner (GP) to record the condition at different levels of specificity of disease. The Read code used is at the discretion of the GP and is based on their diagnosis of the patient's condition

The data as previously given are a reasonable population based estimate of the level of malnutrition, as recorded within primary care data, based upon the GPRD sample size.

The analysis undertaken to count the number of such cases, in the GPRD, is not done routinely. It was done in response to a specific request. The analysis could be repeated at any time in the future.