Consultation on changes to NHS Digital Prescribing and Medicines Statistical Publications

We asked for your support in enhancing our role from custodians of national health and care data, to one where we are also data collaborators and innovators and as part of this, work with partner organisations to remove duplication across the system so we are all delivering maximum value.

As a result we are developing our partnership with the NHS Business Services Authority to migrate existing primary care medicines publications to align with the NHSBSA data releases whilst expanding and increasing the usability of prescribing data in our remaining publications as new data sources become available.

Together, we are working in partnership to deliver a patient level Primary Care medicines dataset through NHS Digital’s Data Access Request Service (DARS) by summer 2020. This will:

• deliver and facilitate analysis of more granular data, data linkage and statistical analysis of linked data;
• make anonymous outputs available as open data;
• identify new opportunities in the data through linkage to wider health data to improve patient safety and outcomes;
• ensure coherence, consistency and value across our outputs; and
• reduce duplication of existing aggregate outputs.

Thirteen responses were received from a range of organisations at all levels of the NHS and industry. The responses received in the main supported our proposals to migrate publications to the NHSBSA as the data controllers to enable timelier access to data and ensure data was available from a single source of truth. Responses also included suggestions to improve usability by providing access to the data at a more granular level.

Concerns were raised however on the continuity of the quality and level of data that would still be available following the change in ownership of publications. 

You highlighted the following key points in response to our proposals:

• Publications should continue to meet the standards required of National Statistics following any change in ownership;
• The level of the data should, as a minimum, remain the same or become more granular where available to support analysis by Primary Care Networks and patient level LSOAs;
• Any changes in data formats should be communicated with users well in advance of any changes;
• Any future suppression of data should not impact on current availability and usability of the data.

Therefore, we are taking the following steps to support these proposals as detailed in the We Did section.

Migrate the Prescription Cost Analysis (PCA) publication to the NHSBSA

The PCA proved to be one of the most popular of our publications with respondents, supporting a range of uses. All respondents were supportive of the move of this publication to the NHSBSA for 2020. Users were concerned that any changes to the data structures and format need to be communicated to users in a transparent and timely manner.

Users would also like to see:

• Comparable statistics at a national level between the countries of the UK;
• Patient numbers and age breakdowns;
• Data split by prescribing and dispensing views;
• Granular data down to STP, CCG, Primary Care Network and Practice level.

NHS Digital will work with the NHSBSA to ensure a comparable data set is released as a minimum in early 2020 and adequate supporting documentation is published. The NHSBSA will also look to introduce summary patient level statistics to this publication in the future alongside their current monthly release of PCA data. This will include high level thematic insight pieces on specific topics or areas of interest aligned to the NHS Long Term Plan, for example, medicines used in Mental Health.

Migrate the Practice Level Prescribing in England release to the NHSBSA

The responses from users highlighted the issues of multiple publications with users split between using the PLP from either NHS Digital or the NHSBSA version. One response even stated they used the data as published by a third party.

Users were supportive of the migration to a single dataset published by the NHSBSA as long as there was no loss of data production and changes in structure and format were clearly communicated. Commercial users of the data noted that any suppression applied to this dataset would impact their business models of providing medicines and device reporting in NHS primary care, potentially making the dataset unusable.

Users need to be clear that the NHSBSA are the data controllers of the data, so any suppression protocols they apply would also affect the dataset that is passed to NHS Digital for publication. The NHSBSA are still developing and consulting on their disclosure control protocol and users need to be mindful that NHS Digital currently apply disclosure control to their publications where required and NHS BSA do not intend to do anything significantly different.

Users would also like to see expansion of this detailed dataset to include patient numbers and age breakdowns to improve polypharmacy analysis. Disclosure control may be needed for this information to enable the data to be released.

NHS Digital intend to publish a final dataset in February 2020 of data for December 2019 in its current format subject to feedback following more detailed work with users to ensure a seamless transition. Links and search facilities via the NHS Digital website will be provided to the NHSBSA dataset. Documentation on the differences in the datasets will be clarified and made available to enable users to transition to the NHSBSA data release as soon as possible. NHS Digital will work with the NHS BSA to ensure any supporting meta-data is searchable so users can quickly and easily locate the data.

 Migrate the Prescribing on Diabetes publication to the NHSBSA

The publication focuses on a specific set of medicines relating to the treatment of diabetes and as a result has a specialised user base. Users welcomed the move to timelier data becoming available with an increase in frequency to quarterly data and were supportive of the move if the raw dataset continued to be available.

As a result, the NHS Digital diabetes publication in November 2019 will move to using the latest NHSBSA data as the source and will be lighter in terms of its analysis. It will focus on 4 or 5 key points and then signpost users to a variety of supporting information on Diabetes, such as the National Diabetes Audit and QOF registers. NHS Digital will work with the NHSBSA to produce a quarterly Diabetes release in 2020 that will continue to produce National and CCG level data and expand to include some patient demographic data.

 Migrate the General Pharmaceutical Services to the NHSBSA

Only one response out of the 13 said they used the General Pharmaceutical Services publication and stated that they also used additional data extracts direct from the NHSBSA and were concerned about variations in the numbers they received.

NHS Digital will therefore produce a final General Pharmaceutical Services publication in November 2019 of some high-level figures with raw data files and signpost users to the variety of different data sources. We will work with the NHSBSA and key stakeholders and users to determine the key information that is required to support policy development of pharmaceutical services and develop how best to present this information. The aim is that the NHSBSA will publish a redesigned annual publication, focused on user needs and aligned to the data available to the NHS as management information to ensure a single version of the statistics is available. This may expand in future to potentially be a single publication covering all pharmacy, ophthalmic and dentistry contractors in primary care.

NICE Technology Appraisals in the NHS in England (Innovation Scorecard) to reduce frequency

Developments to the Innovation Scorecard are overseen by a Strategic Working Group and Technical Working Group with representatives from key stakeholder organisations such as the Office for Life Sciences, NICE, NHS England and industry representatives. Work is on-going with users to improve usability and functionality in association with delivering the requirements from the Accelerated Access Review and the 2019 Voluntary Scheme for Branded Medicines Pricing.

Users supported the move to a bi-annual release if this would result in an improved product for users and the breadth and depth of information was at least maintained if not enhanced. As a result, the Innovation Scorecard will be reduced in frequency from 4 to 2 publications a year. The aim will be to publish in April and October each year. This will enable the scope of the product to be clearly defined with stakeholders and developed into the publication cycle so the Innovation Scorecard can transition into a product that meets the needs of end users.

Prescribing Costs in Hospitals and the Community to move to new data sources

Users responded stating that this publication is useful and unhelpful in equal measure. Users recognise it is the only source of information on overall cost of medicines in the NHS whilst being aware of its limitations due to the data sources available. Users welcomed a move to data that more accurately reflects the true cost of medicines to the NHS to determine the true scale of spend and growth.

The publication in November 2019 will start the transition of the report to new data sources that are becoming available and to focus on actual costs to the NHS. The primary care prescribing component will be sourced from the NHS BSA’s new data warehouse which brings several data improvements, the main one being that prescribing by Dentists will also be included to ensure the whole of primary care prescribing is captured in the costs.

The secondary care prescribing component will use actual cost data from Rx-Info that is currently being used for monitoring purposes within the NHS. The HPAI data set from IQVIA will continue to be used for costs at list price to provide contextual information on the indicative discounts currently being achieved by the NHS.

The publication will initially focus on National level data due to the differences in regional reporting between the data sources. We will work with data suppliers to align this over time so that future publications become richer and more detailed in the information provided on the cost of medicine use in the NHS.