NICOR Commissioning through Evaluation (CtE) Registries/Audits: Left Atrial Appendage Occlusion (LAAO), Mitral valve repair (Mitraclip), and Patent Foramen Ovale Closure (PFOC). (Previously DARS-NIC-53806-V9J4G-v0.8)

The National Institute for Cardiovascular Outcomes Research (NICOR) hosted at Barts Health NHS Trust has received HES Admitted Patient Care and Civil Registration data for use in an audit of patients who have undergone one of three procedures:

- Percutaneous mitral valve leaflet repair for mitral regurgitation using MitraClip;

- Left Atrial Appendage Occlusion (LAAO); and

- Patent Foramen Ovale Closure (PFOC) procedures.

NICOR wishes to retain this data for archiving purposes & the data controller has confirmed that no further processing of the data will occur under this version of the agreement. The data currently held by the data controller was originally disseminated for the purpose of a key deliverable to provide Newcastle Upon Tyne Hospitals NHS Foundation Trust (NUTH) with a pseudonymised linked clinical PFOC, MitraClip and LAAO raw data with ONS and HES data extract. ONS and HES linked data are required from NHS Digital to allow analysis of future clinical events rates after the initial treatment (principally rates of strokes/cerebrovascular events and death).

Pseudonymised HES APC data and identifiable Civil Registration were provided against the cohort of approx. 1965 patients who have received either of the procedures Mitraclip, LAAO and PFOC . This cohort of patients represents all patients treated in the NHS nationally. NICOR are required by NHS England to retain this linked ONS and HES data for each registry for a period of six years. The current request to NHS Digital is for data retention for five years which will be extended subsequently for the sixth year to comply with the directive from NHSE.

In April 2013, NHS England became responsible for directly commissioning specialised services and published a clinical commissioning policy statement which stated that percutaneous mitral valve leaflet repair for mitral regurgitation using MitraClip, LAAO and PFOC procedures would not be routinely funded as the current evidence base had been of insufficient quality. NHS England developed ‘Commissioning through Evaluation’ (CtE) to support the commissioning of clinical activity around an evaluation programme to add to the current evidence base for designated technologies. The overall aim was to provide a clinical and economic evaluation of the MitraClip and the other two procedures. Funding was provided for a limited number of high-risk patients as determined by an appropriately constituted Multi-disciplinary Team (MDT) process.

Prior to 2013 service provision was unstructured and delivered by providers who attempted to establish services through negotiation with local specialised commissioners. The majority of procedures were commissioned through individual funding requests or by individual Trusts funding treatments by other means. In April 2013 NHS England became responsible for commissioning specialised services. However as the evidence base for percutaneous mitral valve leaflet repair for mitral regurgitation using MitraClip, LAAO and PFOC procedures was of insufficient quality the procedures were not to be routinely funded.

Because of the remit of NICE in terms of developing evidence-based guidance NHSE, the principal funder, commissioned NICE to establish the ‘Commissioning through Evaluation’ (CTE) programme in order to gather further evidence to support commissioning of designated technologies. The overall aim was to develop a model for both clinical and economic evaluation of the three procedures which could be used in future for other similar new health technology procedures.

NICE contracted with the Newcastle upon Tyne Hospitals NHS Foundation Trust (NUTH) which hosted one of the NICE External Assessment Centres (EAC) with the task of establishing this programme. The NUTH EAC housed a team of scientists, economists and engineers, who provide an independent assessment of evidence for innovative medical devices and associated economic analysis. The team were substantive employees of NUTH at the time. With its extensive experience and excellent reputation in conducting national cardiovascular audits and registries, the National Institute for Cardiovascular Outcomes Research (NICOR) was commissioned by NUTH to develop, host and maintain the three cardiovascular registries.

NICOR act as the sole data controller for the CtE datasets because these datasets contain variables over and above those needed for the NUTH purposes, which might be of interest to other applicants for the audit data. The NUTH/NICE would not be expected to have to be part of the approvals process for such applications, should they occur. The national CtE Steering Group and a number of Working Groups were established to define the datasets. The Working Groups deemed that the datasets should be designed to answer as many questions as was felt clinically and administratively relevant for the purposes of evaluating the technology in more detail. The datasets were approved by the Working Groups and the Steering Group.

As well as determining the CtE datasets and defining the audit questions, NICOR is also responsible for managing all aspects of the data collection process, data security, data storage, cleaning/validation of data, data linkage with ONS/HES and reporting. The reporting function includes providing pseudonymised data extracts to NUTH as per its requirements, receiving all variables needed for the purposes of the agreed independent analyses. NICOR’s purpose is to contribute to the aim of developing a model for both clinical and economic evaluation of the three procedures through fulfilment of the responsibilities commissioned by NUTH. Processing of the data being requested is necessary for the performance of a task in the public interest or in the exercise of official authority vested in the data controller. To achieve this purpose and the obligations of the data controller (sub-article 6(1)(e). Included within these responsibilities and ultimately the purpose of this request, is to process NHS Digital data for the archiving purpose (sub-article 9(2)(j)).

The work aimed to gather more data on the safety and effectiveness of three complex cardiology procedures

which are not currently routinely funded by the NHS, but nonetheless show significant promise as a

future treatment. The data gathered from centres undertaking these procedures was analysed to inform

future commissioning intentions.

This is not an experimental research project, but an evaluation of service delivery. There is no specific hypothesis being tested. NICOR have defined 11 specific questions to frame the CtE evaluation programme:

1) Can UK clinical teams undertaking MitraClip reproduce the reduction in mitral regurgitation seen in the early clinical trials?

2) Is reduction in mitral regurgitation mediated by MitraClip associated with improvements in quality of life?

3) Does MitraClip improve survival rates?

4) Does MitraClip reduce the frequency of subsequent hospital admissions?

5) Are the early benefits in reduction in mitral regurgitation maintained in the medium-term? Is there a need for repeat treatment over time (either by a repeat percutaneous procedure or surgery)?

6) What proportion of patients referred to a specialist Mitraclip service as defined in the CtE documents were assessed by the MDT as suitable for the intervention? What proportion of the patients considered suitable for the procedures received it and what proportion of them benefited?

7) What are the short and medium term risk of complications from MitraClip use?

8) Are clinical outcomes with MitraClip associated with particular patient characteristics (clinical or demographic)?

9) What are the full procedural costs of using MitraClip to the NHS?

10) What costs savings might occur in the NHS as a result of MitraClip™ therapy?

11) Is MitraClip cost-effective from the perspective of the NHS?

It is intended that the data reported from the registries will be used to part inform an update of IPG 309 (The NICE Interventional Procedures Guidance on MitraClip), which will be reviewed by NICE, in light of the CtE registry findings.

Separately, NHS England Commissioning Policies for all 3 cardiac procedures will also be reviewed, with a view to future patient access to these treatments.

As the original data has been collected for evaluation (Commissioning through Evaluation) of new health technologies the original data needs to be retained for a minimum period of six years, as per the statutory requirements. The data will be retained by NICOR only, as NUTH has destroyed all NHSD data relating to CtE data sharing agreement (as per the data destruction certificate that has been submitted to NHSD). If during this period safety issues, relating to the devices, have been identified then a data sharing agreement for data processing would be implemented with NHSD. Otherwise this request is for data retention only.

NICOR are the sole data controller for NHS Digital data currently held. Redcentric PLC are listed as a Data Processor as they provide storage of the data, however no data is processed outside of storing by Redcentric.

Through a specifically designed database, NICOR received patient identifiable data collected by NHS Hospital Trusts about patients who have had a Mitraclip, LAAO and PFOC together with data about the treatment itself and outcomes both to hospital discharge and during predetermined follow-up periods. From the agreed dataset, those fields needed to answer the specific evaluation questions were provided to the NUTH. No patient identifiable data was shared with NUTH. The datasets for the three registries are available on the NICOR website.

NICOR then sent the following identifiers to NHS Digital for NHS Digital to provide NICOR with linked Civil Registration mortality and HES data:

- Member Number

- Name (both forename & surname)

- Postcode

- NHS Number

- Gender

- Date of Birth

This data was then linked to HES Admitted Patient Care data. HES APC data was used to evaluate frequency and duration of re-admissions and pre-admissions for patients recorded in the above-mentioned registries. The OPCS4 and ICD10 codes were requested for each linked patient. These were used to search for any episodes of stroke/cardiovascular event, post-cardiac CtE procedure etc. and also linked to HES Civil Registration data to find out if any of the cohort had died.

Civil Registration data was used, firstly, to identify patients who had died out of hospital, or in between follow-up appointments, where deaths may not have been known by the associated treatment centre (therefore missing from the registry follow-up data). This will validate the mortality rate found using registry data alone. Secondly, to confirm registered cause of death and determine whether this is of neurological or cardiac origin. This will allow the study to determine how many of the deaths in these patients may be attributed to failure of the procedure, or another cause. The registry data linked with NHS Digital data will only be archived. There is no intention to match the archived data under this DSA to data from any other source (publically or otherwise available) nor is there any requirement /intention to attempt to re-identify any individuals.

The Civil Registration/HES linked data (returned to NICOR): member number is the only identifiable data returned. The senior developer then re-linked the member number to the identifiers. Once the initial data checks and data verification had taken place, where senior developers verified the data once NHS Digital data had been re-linked with procedure data, all identifiers were stripped from the data. All data was securely stored on a departmental shared drive, which was restricted to staff working on the project. The nominated staff (substantive employees of NICOR) have varying levels of permissions to view and access the data depending on their position and involvement.

The Civil Registration/HES linked data transferred from NICOR to NUTH contained no patient identifiers and was stored securely within the NUTH network which has access protected in accordance with the NUTH ‘Network Security & Access Control Policy’. Access to the data will be restricted to authorised staff working on the project. This data at NITH has now been destroyed.

The York Health Economics Consortium (YHEC) is not involved in any data processing - they only received aggregated (in line with the HES Analyses Guide) outputs of the analyses from NUTH such as mortality rates and hospital readmission rates, to be used as inputs to their economic modelling work for NHS England.

All organisations party to this agreement must comply with the Data Sharing Framework Contract requirements, including those regarding the use (and purposes of that use) by “Personnel” (as defined within the Data Sharing Framework Contract namely: employees, agents and contractors of the Data Recipient who may have access to that data).

NHS Digital reminds all organisations party to this agreement of the need to comply with the Data Sharing Framework Contract requirements, including those regarding the use (and purposes of that use) by “Personnel” (as defined within the Data Sharing Framework Contract namely: employees, agents and contractors of the Data Recipient who may have access to that data).

Aggregate data with small numbers suppressed in line with the HES Analysis Guide which has been derived from the data in this application was contained within reports published in 2018, for presentation to NICE and NHS England. These reports may also be made available to other stakeholders. The reported data may also be visualised in various formats on-line, as relevant to the stakeholders. No record level data or patient identifiable data will be released in any report. The proposed publication date for these outputs is early 2018. (Results from authorised work may also be subsequently published in peer-reviewed journals).

The NICE/CtE report for Percutaneous patent foramen ovale closure (PFOC) was published in July 2018, and can be found here: https://www.england.nhs.uk/wp-content/uploads/2018/07/nice-pfoc-cte-project-report.pdf

The primary purpose of the Civil Registration/HES linked data was to inform the CtE programme. All published reports are made available on the NICOR website (https://www.nicor.org.uk/) or the NHSE Commissioning through Evaluation (CtE) website. This only happens once the funder has reviewed and approved the report for publication. All outputs will contain only data that is aggregated with small numbers suppressed in line with the HES Analysis Guide.

No further outputs are planned under this application, which is purely for the archiving of the linked data.

The review of the initial NHS experience with these treatments aligned to an updated review of the world literature will enable NHS England Specialised Commissioning to update its commissioning policies regarding the use of the MitraClip, LAAO and PFOC procedures.

It is important to retain the linked data (Civil Registration/HES linked data to registry data) at NICOR for the next six years. A further extension of the agreement will be requested at the end of the five-year period of this agreement. The retention of the linked data is vital and would benefit the programme if any safety issues relating to the treatment procedures that patients received are identified the appropriate investigation can take place in the interest of public safety. Further processing of the data will not occur under this version of the agreement and should additional processing be required it will be requested under an amendment in the future.

There were a number of benefits realised once the work had been completed:

1) The ability to look at cardiovascular admissions which may be related to, and affected by, the medical management of a patient’s condition. This will provide a much more detailed and complete picture of readmissions, and help determine the full impact of the original treatment on readmission rates and mortality outcomes.

2) The ability to evaluate readmission rates and reasons for readmission will be extremely beneficial to commissioners, healthcare economists and regulatory bodies such as NICE in terms of assessing the long term effects on patients undergoing the various cardiac procedures. It will allow assessment of the effect of different variables on these outcomes and the assessment of the total burden of care, thereby aiding an analysis of cost effectiveness.

3) To provide additional insight into outcomes (especially adverse reactions such as stroke) which Barts Health (NICOR)/NUTH can then include in any reports used to inform quality improvement work/evaluation of new health technology. Linkage to the HES dataset would allow further exploration of the geographic, socio-economic and organisational data of patients in much more detail. This may lead to a better understanding of commissioning patterns within the UK. In addition, the HES dataset collects information on augmented care and the patient care pathway.

4) To identify relevant safety and efficacy outcomes that were not known to the clinicians undertaking the procedures in the tertiary centres because the relevant episodes of care took place in other hospitals.

Should these treatments become routinely commissioned, these types of outputs will be useful for future reports derived from data collected by NICOR.

At present, for LAAO and PFOC, two reports have been submitted to NHS England using registry data alone, describing patient outcomes such as mortality and post-procedure neurological events, which have been reported by the centres following up their patients. The MitraClip report was also to NHS England in February 2018 and described patient outcomes such as mortality and post-procedure events requiring hospital readmission.

It has been vital to use the Civil Registration and HES data, linked to each registry for realising the following benefits:

(i) to confirm that all deaths and post-procedure hospital readmissions (for neurological / cardiac events) have been captured in the registry. NICOR do not have 100% follow up of CtE cases - so do not know if there is missing information for many patients post-procedure.

(ii) where additional events are identified in Civil Registration and HES, the combination of registry-reported outcomes and Civil Registration / HES recorded outcomes will validate the mortality rates and efficacy measures (stroke prevention rates and cardiac outcomes) of these 3 procedures and provide more robust data to NHS England. These data will directly inform NHS England's commissioning policy decision-making, which will determine patient access to these procedures in future. Where there have been discrepancies between the registry and HES/Civil Registration recorded data NICOR have endeavoured to take additional steps to validate the registry data (particularly where there has been perceived 'safety issue'.