Introduction Chronic Hepatitis B (HBV), a major cause of liver disease, is underdiagnosed in the UK with pockets of high prevalence in areas with large immigrant populations. Identifying individuals through a screening programme dedicated to high risk individuals would be beneficial if that target population could be defined.

Our aims were to interrogate a Primary Care database (SystmOne) establishing the best way of identifying individuals suspected of having had a HBV test to exclude them from the target population.

Methods We used four search terms: age (>18), ethnic code, birth place, language spoken. Read codes (RC) were applied to identify individuals previously tested/diagnosed with HBV. RC were identified by using ‘HBV’ as a broad search term.

Results We identified 4256 individuals aged ≥18. 718 (18%) were excluded because of lack of demographic data; 3210 (75%) qualified as the target population. 81 RC pertaining to HBV were found and generated 413 ‘hits’, only 224 ‘hits’ (54%) satisfied our criteria. 206 individuals (6.4%) had HBV serology recorded, 9/3210 (0.28%) were coded for chronic HBV, a further 2 HBsAg positive patients (0.9%) had no RC. After reviewing individual notes electronically we established the following 8 RC to be the ‘most useful’: XaIq7, XaFuS, X306n, X306i, 43B4, XaMBL, XaG1R, XaPEy which would have identified the latter.

Conclusion We have identified the 8 most useful RC to help with a potential screening programme. Using all available HBV RC would be too time consuming. This study highlights that very few high risk immigrants are currently tested for HBV.

Disclosure of Interest None Declared.