OBJECTIVES: Hepatitis delta (HDV) is a rare condition, affecting a subset of HBV+ patients. EMA have recently approved the first treatment for this condition. Prevalence of HDV is unknown, data are scant in the literature, biggest risk factor for HBV is country of birth. Objective of this research was to generate a scalable model framework to provide prevalence estimates of HDV for any given country, to guide local & national access discussions.

METHODS: Model was created combing 3 data sets (HBV prevalence, HDV coinfection & population data by country of birth) to predict prevalence of HDV. Data were matched using country or region (where country not available).

1. No prevalence estimates available - “nearest neighbor” methods applied to select HBV & HDV values.
2. Where countries were grouped into regions, two methods were applied (ONS vs. WHO definitions).

RESULTS: Using UK as a pilot, data were available for 60 individual countries of birth at national level, but only available at aggregated groupings at sub-national level. Of 60 named countries, HBV data was available for 55 / 60, HDV data was available for 38 / 60. When applied to the UK, HDV prevalence was calculated at: ~15,000 to ~18,000 (depending on data gap methodology). ONS method produced more granular, higher result compared with WHO method (18k vs 15k).

CONCLUSIONS: Model is sensitive to variation in availability of migration data in terms of prevalence estimates. Model appears to cope with scaling for additional countries and subnational estimates. Consideration should be given to differences in migration population data at the national vs. the sub-national level. The model framework can easily be extended and updated to incorporate new data when available, providing a framework for consistent prevalence estimates.