The aim of this work was to analyse community pharmacy dispensing data at scale, to provide insight to NHS Wales on areas of high cost and potential fraud. This work was also as an opportunity to develop Audit Wales’s expertise in fraud analytics techniques.

Fraud and error present a significant challenge to public finances in Wales. We have previously estimated that fraud and error cost anywhere between £100 million and £1 billion each year to Welsh public services.

Our Data Analytics team has been working with NHS CFS Wales colleagues on a pilot project looking at community pharmacy dispensing data, given community pharmacy is an area of considerable expenditure and with known fraud risks.

Our pilot covered Swansea Bay University Health Board and Cwm Taf Morgannwg University Health Board. We focused our analysis on three areas of known risks around fraud and cost:

• Higher cost formulations - Different formulations (e.g. liquids, tablets, capsules, creams, branded and non-branded etc) of the same active substance can vary widely in price.
• Expensive items: Items with a net ingredient cost of £100 or more.
• Special orders (Specials): Items requiring special preparation by a registered manufacturer.

Although subject to various caveats, we estimate that during the three years covered by the pilot, approximately £700,000 could have been saved by the NHS in Wales by reducing costs associated with Specials.

And whilst we did not find any immediate evidence of fraud, we flagged numerous outliers to health boards. From these outliers, we identified overpayments totalling £22,000, which is now being recovered.

We are today publishing a letter that summarises our work and includes questions for the NHS in Wales to ask itself as it acts on our findings.  

We will build upon the learning from the pilot and look to undertake new fraud analytic projects, including one using a data matching approach to explore whether patients are accurately registered in GP lists.