AC-2023-LON-002171 - [2024] EWHC 132 (Admin)

Lee 1 exhibited claims data that was said to be drawn from the FP17s submitted by the Appellant. The statement explained that none of the paper FP17s submitted by Dr Imani were available, as the forms were only retained for a period of 14 months and then were shredded after they had been scanned by the imaging team in Newcastle. The exhibits to Lee 1 included NHS4, a spreadsheet which set out the claims data in relation to claims made by the Appellant’s practices during the period January 2012 to December 2017. This data included the date when the form was received by BSA, the patient it related to and the payment schedule, the TAD, the TCD, the treatment indicated, and the UDAs claimed. NHS4 was used as the basis for the schedule prepared by Dr Scott for the purposes of his report, which in turn became Schedule C (which, as I have indicated, summarised the data in tabular form in relation to each of the claims that had given rise to a charge). Mr Lee also exhibited NHS5, which included all of the fields relating to Part 3 of the FP17 from for the period April 2014 to December 2017, so that it could be seen whether the treatment had been claimed as incomplete (which NHS4 did not show). The additional information in NHS5 was not available in respect of the period prior to April 2014.

In her witness statement the Appellant complained that it was unfair that she was unable to access the FP17 forms to check their contents. She said that she did not consider the BSA’s data to be reliable.

Following concerns that were raised on Dr Imani’s behalf regarding apparent discrepancies, the parties agreed that Mr Lee would provide a further statement explaining the data production and answering these specific points.

Lee 2 described the way that the FP17 forms were processed after they were received by BSA. They were scanned by the Scanning Operations Department, thereby creating the date of receipt. The forms were placed in a “Pouch File” with its own unique reference number. The individually scanned FP17s and the Pouch File were then transferred electronically to a branch of BSA known as NHS Dental Services and from there to a third party supplier, who was initially RR Donnelley and then Capita (who both used the same process). The third party supplier would then manually key in the information contained in each individual FP17 form from the Pouch File into an electronic file (the “raw data”). The raw data included the TAD, the TCD, whether there was an “incomplete” marking and the Band. The electronic file containing the raw data was then transferred back to NHS Dental Services and uploaded to its system. Until February 2016 the system used to upload the electronic files was a VME mainframe and thereafter the Compass system was employed. Compass produced monthly schedules for performers and providers and fed the data into the Data Warehouse each month.

Mr Lee also explained that software called Business Objects was used to pull the data from the Data Warehouse for the purposes of NHS4, whereas a software system called eDEN was used in preparing NHS5.

Lee 2 also addressed the reliability of the data. Mr Lee said that under the service level agreement, the third party supplier was required to ensure an accuracy rate of 99.9% every month and was required to report to BSA on this on a monthly basis. The monthly reports were no longer available, but neither supplier had ever reported an accuracy of less than 99.9% during the period in question. The quality assurance was done by comparing a 10% sample of the forms in the scanned Pouch Files with the data that had been manually keyed in from the scanned paper FP17 forms to see if they matched.

Mr Lee also responded to the apparent discrepancies highlighted by the Appellant’s solicitors. The largest number of such discrepancies had been identified in relation to the CRD shown, respectively, on NHS4 and NHS5. Mr Lee explained that this was not information taken from the FP17, but related to the date when the form was received by NHS. The Business Objects system took the date of receipt of the form by BSA as the CRD, whereas eDEN took the date of the scanning, which in most instances, but not invariably, was the same. A discrepancy between NHS4 and NHS5 in relation to the data for Patient 11 was also attributable to the different ways in which Business Objects and eDEN pulled the FP17 data when no TDC or date of last visit was shown on the form.

On Day 7 of the hearing (31 May 2022) Mr Lee gave evidence in accordance with Lee 2. (During the evidence NHS4 is also referred to as exhibit 8 and NHS5 as exhibit 10.) Mr Lee was asked to elaborate upon how the accuracy tests were conducted by the third party supplier. He said that the data was re-keyed by a different individual and the two compared. He said that on one occasion a BSA staff member had witnessed the data capturing process, including the quality assurance check. He said that on other occasions the monthly audit reports were taken on a trust basis by BSA. The data was inputted in Columbo and in India.

Mr Lee said that he believed that the schedules that were made available to dental providers and performers, against which they could check the accuracy of the inputted information, included the UDA value for the claim and both the TAD and the TCD. On 22 September 2022 the PCC was provided with an Agreed Fact that the monthly reports issued to dentists during the material period included the patient’s name, the band claimed, the UDAs allocated, the patient charge and comments, but did not include the TAD or the TCD.