electronic medium. This medium is well suited to being searched, updated, and copied. We are currently exploring this option locally. 5 Any electronic method of dissemination will require careful management and will in itself only be a further tool to aid decision making.Contributors: AH and DK designed, initiated, and coordinated the study. DP assisted with the design, interpretation, and direction of the study. FP collected the data and assisted with their analysis and interpretation. The paper was written jointly by AH, DK, DP, and FP. AH and DK are guarantors for this study.Funding: Cambridge and Huntingdon Health Authority. Conflict of interest: None. Subjects, methods, and resultsWe contacted the staff in three main settings-the health authority headquarters, an accident and emergency department, and various departments in another hospital-and asked them to complete a form that contained boxes for the respondent's name, the 26 letters of the alphabet, and the digits 0-9. They were told that examples of handwriting were needed to test computer software for optical character recognition and were asked to write as neatly as possible. All 92 staff present in the three settings were asked to participate, and none refused. We analysed their responses with Teleform, a software package that allows handwritten replies on standard forms to be scanned and translated into text for computer analysis.3 Any unrecognised characters are highlighted, and an error score is generated.For the analysis, the staff were divided into three groups: doctors, nurses plus other medical professions, and administrative staff. We collated the results with the spss statistical program. As the error scores were not normally distributed, we used median values when comparing each group and used the Kruskal-Wallis or Mann-Whitney U test to test any observed differences for significance. In order to control for possible confounding we examined the effects of sex, setting, and age separately.The table shows the median legibility error score for each professional group. Numeric legibility was similar for all groups and not considered further. For letters there was a significant difference between the groups (P = 0.006). The doctors had a higher median score compared with the other two groups individually (P = 0.01 for nurses plus other medical professions, P = 0.005 for administrative staff) or combined (P = 0.001). Analysis of female respondents alone revealed a similar pattern, with the doctors having a higher median error score than the other two groups (P = 0.032 for nurses plus other medical professions, P = 0.09 for administrative staff, P = 0.036 for the groups combined).The doctors had a slightly higher median age (37.5 years) than did the other two groups (33.0 years and 31.5 years respectively), but this difference was not significant (P = 0.78), nor was there any significant effect of age on legibility for all respondents or for doctors alone. The doctors in each of the three main settingshealth authority headquarters, accident and ...

IntroductionPrimary healthcare records are used for studies within large data repositories. One of the limitations of using these routinely collected data for epilepsy research is the possibility of including incorrectly recorded diagnoses. To our knowledge, the accuracy of UK GP diagnosis codes for epilepsy has only partially been validated. Objectives and ApproachWe aimed to validate the accuracy of case ascertainment algorithms in identifying people with epilepsy in routinely collected Welsh healthcare data. A reference population of 150 people with definite epilepsy and 150 people without epilepsy was ascertained from hospital records and linked to records held within the Secure Anonymised Information Linkage (SAIL) databank in Wales. We used three different algorithms to identify the reference population: a) individuals with an epilepsy diagnosis code and two consecutive AED prescription codes; b) individuals with an epilepsy diagnosis code only; c) individuals with two consecutive AED prescription codes only. ResultsWe applied the algorithms to all patients and to adults and children separately. For all patients, combining diagnosis and AED prescription codes had a sensitivity of 84% (95% ci 77–90) and specificity of 98% (95–100) in identifying people with epilepsy; diagnosis codes alone had a sensitivity of 86% (80–91) and a specificity of 97% (92–99); and AED prescription codes alone achieved a sensitivity of 92% (70–83) and a specificity of 73% (65–80). Using AED codes only was more accurate in children, achieving a sensitivity of 88% (75–95) and specificity of 98% (88–100). This can be explained by the widespread use of AEDs for indications other than epilepsy in adults, which is not the case for children. Conclusion/ImplicationsGP epilepsy diagnosis and AED prescription codes can be used to identify people with epilepsy using anonymised healthcare records in Wales. In children using AED prescription codes alone is an accurate way to identify epilepsy cases. These results are generalizable to other studies that use UK primary care records.

No abstract