Modelling Healthcare-Evidence Responsive Behaviours (HERBs) in Doctors: A proof of concept study

No data is to be disseminated under this version of the agreement.

The project aims to model how responsive different graduate groups of consultants are to the publication of new guidelines regarding the use of drug-eluting stents in percutaneous coronary intervention. It acts as a first step towards modelling actual clinical behaviour of consultants.

Drug-eluting stents were introduced in 2003 as an alternative to bare metal stents. Evidence emerged in 2006 regarding the potential harm of drug-eluting stents in some high risk patients. A second generation of drug-eluting stents were subsequently developed. Guidelines released in 2010 supported the use of drug-eluting stents in most patients, and guidelines released in 2014 recommended their use in nearly all patients.

University of York/Generl Medical Council aim to investigate how the use of drug-eluting stents changed in relation to the emergence of evidence against and subsequently in favour of their use. As such, the request is for data on all finished consultant episodes relating to percutaneous transluminal balloon angioplasty and insertion of stent into coronary artery' from 2006/2007 to 2016/2017 inclusive. 4 character information is needed on the procedure carried out to identify the type of stenting procedure performed. As it is not possible differentiate between drug-eluting stents and others using OPCS 4.2 codes, the sample is restricted to 2006 onwards when OPCS4.3 codings were introduced. Other data requested are thought to be potential confounding variables and as such need to be controlled for in the statistical modelling procedure.

This HES data extract will be linked to data held by the General Medical Council on the characteristics of doctors (the List of Registered Medical Practitioners), and then psuedoanonymised. From this combined dataset it will be possible to determine which consultants graduated in the UK, and which graduated outside of the UK. Modelling of this combined data set will allow us to answer the research question.

The request is funded by the University of York Research Priming Fund, 2018/19 round. The outputs of this work will feed into larger programmes of work.

The General Medical Council (GMC) will act as data controllers and data processors for this application. The GMC will receive the HES extract. The University of York is paying the GMC to link the HES extract to data held by the GMC. The data will then be psuedonanynomised and placed in the Health Informatics Centre safehaven hosted by the University of Dundee. As such, University of York researchers will not have access to identifiable consultant data. See section 5b for further details.

No data is to be disseminated under this version of the data sharing agreement.

The University of York, in collaboration with the GMC, have determined the purpose and means of processing the data. The GMC will receive the initial HES data extract. This extract will be identifiable record level data, containing consultant code (i.e. GMC number) for each episode. Only the GMC will have access to this data, which will be stored in the secure servers of the GMC. Only basic patient level data will be requested. Access to the data at the GMC will be determined by job role. Only data analysts that work on the UK Medical Education Database (UKMED) will have access to the data. This is currently restricted to 4 people. The GMC will link the HES data extract to the List of Registered Medical Practitioners using the consultants GMC number. The linked data will then be pseudoanonymised by the GMC. If necessary, the data will be blunted to ensure non-identifiability.

The pseudoanonymised data will then be placed in the Health Informatics Centre (HIC) Safehaven hosted by the University of Dundee. The research team at the University of York will then be granted access to the pseudoanonymised data set remotely (see https://www.dundee.ac.uk/hic/hicsafehaven/) via the VMHorizon Secure Remote link. The University of York will not hold the data.

Statistical modelling will then be carried out in order to answer the research objectives outlined in the previous section. Only summary statistics and coefficients from statistical modelling can be extracted from the safehaven. No individual data are permitted to leave the safehaven.

The data processing will be performed under Article 6, section (e) of the GDPR. The Medical Act 1983 provides the legal basis for this flow of data. The GMC does not having specific data sharing powers in relation to research but relies on its broad statutory functions under the Medical Act 1983. These functions give the GMC the ability to use personal data for research purposes where it is proportionate to do so. The data which the GMC obtains for research are not used for other purposes within the organisation (for example, as part of fitness to practise investigations into individual practitioners). The data is only released to researchers in anonymised or pseudonymised research extracts, through a Safe Haven which prevents unauthorised identification or publication.

The GMC has the following functions and powers under the Medical Act:

Section 1 establishes the GMC’s overarching objective:

(1A) The over-arching objective of the General Council in exercising their functions is the protection of the public.

(1B) The pursuit by the General Council of their over-arching objective involves the pursuit of the following objectives—

(a) to protect, promote and maintain the health, safety and well-being of the public,

(b) to promote and maintain public confidence in the medical profession, and

(c) to promote and maintain proper professional standards and conduct for members of that profession.

Section 5 establishes functions related to undergraduate medical training:

(1) The General Council shall have the general function of promoting high standards of medical education and co-ordinating all stages of medical education.

(2) For the purpose of discharging that function the General Council shall -

(a) determine the extent of the knowledge and skill which is to be required for the granting of primary United Kingdom qualifications and secure that the instruction given in or under the direction of bodies or combinations of bodies in the United Kingdom to persons studying for such qualifications is sufficient to equip them with knowledge and skill of that extent;

(b) determine the standard of proficiency which is to be required from candidates at qualifying examinations and secure the maintenance of that standard;

Section 34H establishes functions relating to postgraduate medical training:

(1) The General Council shall—

(a) establish standards of, and requirements relating to, postgraduate medical education and training, including those necessary for the award of a CCT in general practice and in each recognised specialty; (b) secure the maintenance of the standards and requirements established under paragraph (a); and (c) develop and promote postgraduate medical education and training in the United Kingdom.

(2) In exercising their functions under this Part, the main objectives of the General Council, in addition to the over-arching objective, are—

(a) to ensure that the needs of persons undertaking postgraduate medical education and training in each of England, Wales, Scotland and Northern Ireland are met by the standards the General Council establish under subsection (1)(a) and to have proper regard to the differing considerations applying to different groups of such persons; and (b) to ensure that the needs of employers and those engaging the services of general practitioners and specialists within the UK health services are met by the standards the General Council establish under subsection (1)(a).

Schedule 1 part II allows the GMC to carry out broad activities conducive to its functions, and permits collaboration with related bodies:

9. It shall be within the capacity of the General Council as a corporation to do such things and enter into such transactions as are in their opinion incidental or conducive to the performance of their functions under this Act, including the borrowing of money.

9A (1) In exercising their functions, the General Council shall…

…(b) co-operate, in so far as is appropriate and reasonably practicable, with public bodies or other persons concerned with -

(i) the employment (whether or not under a contract of service) of provisionally or fully registered medical practitioners,

(ii) the education or training of medical practitioners or other health care professionals,

(iii) the regulation of, or the co-ordination of the regulation of, other health or social care professionals,

(iv) the regulation of health services, and

(v) the provision, supervision or management of health services.

The research proposal aims to address important questions about the links between medical training and clinical practice, and therefore supports the GMC’s statutory functions and can be used in the context of policy development.

No data will be disseminated under this version of the agreement.

This work is expected to produce one high-impact, open-access, peer-reviewed publication. Target journals include the BMJ or BMC Medicine.Target date for submission is Q3 2020. It is anticipated that publication of this work will be accompanied by a press release and social media activity.

University of York/GMC also anticipate significant findings will be presented at a number of local, national and international conferences and presentations.

Alongside coefficients from statistical models, outputs will contain only aggregate level data with small numbers suppressed in line with HES analysis guide.

Significant research findings will be disseminated directly to relevant stakeholders (e.g. the GMC) to inform policy decisions going forwards. Findings will also be used to develop future funding proposals.

No data will be disseminated under this version of the agreement.

The UK still heavily relies on overseas trained doctors and the impact of ‘Brexit’ will almost certainly impact the numbers of EEA trained doctors practising, which have already dropped by 6% over the last 5 years. Whilst there are plans to expand UK medical schools this won’t influence consultant numbers for over a decade. Thus, there are implications for workforce recruitment as well as medical regulation and education. Currently it is not known if clinical behaviour differs between different groups of doctors. This study will begin to investigate this. If differences are found, this study will lay foundations for a programme of research aimed at reducing unwarranted variation in medical practice, a topic of interest to the NHS via the 'Getting It Right First Time' (GIRFT) Programme.

Yielded Benefits is not a requirement for new applications.