NHS Digital Data Release Register - reformatted

Plymouth University projects

12 data files in total were disseminated unsafely (information about files used safely is missing for TRE/"system access" projects).

Project 1 — DARS-NIC-15504-X5K0H

Type of data: information not disclosed for TRE projects

Opt outs honoured: N

Legal basis: Health and Social Care Act 2012

Purposes: ()

Sensitive: Non Sensitive

When:2016.04 — 2016.08.

Access method: One-Off

Data-controller type:

Sublicensing allowed:


  1. Hospital Episode Statistics Admitted Patient Care


Contributing to the broader objectives of a research project funded by the Department of Health Policy Research Programme (addressing how different kinds of provider organisation manage and provide services, and what differences result in terms of service quality, innovation and patients' experience), this analysis of HES data aims to trace the geographic, socio-demographic and clinical expansion of non-NHS provision of NHS services across orthopaedics and ophthalmology.

Understanding and describing the evolving provider landscape since the 2003 Health and Social Care (Community Health and Standards) Act formalised the legal basis of ‘alternative providers’ is the primary objective, but contemporary patterns of orthopaedic and ophthalmic provision (and how they are currently changing) will also be used to contextualise managerial attitudes and behaviours found in the provider organisations visited as part of the research project.

To that end the research team require detailed HES data on ‘provider-purchaser’ relationships, the conditions being treated and the procedures undertaken. Relating this to detailed information on who is being treated (in terms of age, sex, GP and LSOA of residence – which allows linkage with Index of Multiple Deprivation data on the socio-economic characteristics of local populations) provides a basis for understanding the evolving clinical, geographical and social coverage of different kinds of provider organisations. It also provides the basis for deriving summary information on the changing populations being served by different providers, thereby allowing us to (a) better contextualise both the strategic and managerial decisions being made by different providers as well as provider-level measures of service quality, innovation and patient experience and (b) chart how relationships between providers (in terms of who they treat for what) have evolved since the 2003 Act.

As stated in the research protocol, none of the publications arising from this study will identify the people or organisations involved in the research.

Expected Benefits:

The research has been commissioned by the Department of Health Policy Research Programme because, as stated in the application, “those who commission NHS services will increasingly need to know how to select among the different types of organisation for different tasks; and, for each type, the best ways of managing its contribution and relationship to the NHS” and this requires better knowledge “about provider behaviour within the 'provider landscape'”. This is one of a number of projects that have been commissioned by the DH to address aspects of diverse provider landscapes. The benefits of the research will of course depend on the extent to which NHS commissioners and the DH choose to act upon the research findings, but the overarching goal is to provide evidence (a) on the impact of provider diversification and patient choice; (b) to inform practice and policy about the future organisational development of NHS foundation trusts; and (c) to strengthen NHS commissioning practice in different provider environments. As noted above, the findings will be made available during 2017, but it is difficult to predict when any policy/practice changes will be implemented.


Outputs will include a final report to be submitted (July 2017) to the DH’s Policy Research Programme and papers in open access peer-reviewed journals, but the key target audience is managers and policy makers. Thus during the course of 2017 the team will present all findings to national and regional conferences and organisations aimed at service providers, professional bodies and professional leaders (e.g. NHS Confederation). The team will also present findings through a project website, which will incorporate maps illustrating the penetration of provider competition and diversity. Fulfilling the commitment not to identify people or organisations involved in the research, these maps will not explicitly identify the provider organisations providing services in different localities, but merely the extent to extent of non-NHS provision of NHS services. This website will play a major part in fulfilling the team's commitment to provide feedback to patient and managerial informants upon whom much of the wider project will depend.


Processing of pseudonymised individual-level records will take place in one secure location, with only summary statistics (as described below) being extracted for subsequent analysis. Processing will comprise the following:

(i) Patient-level data will be used to create a series of annual matrices describing the flow of patients between, on the one hand, (a) middle super output areas (MSOA) and (b) GPs and, on the other, all providers of orthopaedic and ophthalmic inpatient services. The principal focus will be on high volume services such as hip and knee replacements, arthroscopies and cataract procedures, although the study will examine all orthopaedics and ophthalmology inpatient activity to capture the evolving provision at private hospitals and Independent Sector Treatment Centre’s (ISTCs). For each of a series of defined procedure categories (e.g. hip replacements) a variety of metrics will be derived to describe the provider environment from geographic (MSOA), general practice and provider perspectives. These metrics would, for instance, identify the dominant provider for each MSOA / General Practice, along with the proportion of patients in each MSOA/GP served by that dominant provider. Provider-orientated metrics would seek to describe the extent to which individual providers share catchment populations with other providers of the same or different type (i.e. NHS trusts / private hospitals / ISTCs). These, and other, summary statistics will be shared with project team members and, once aggregated with small numbers supressed in line with the HES analysis guide, be used in subsequent publications. Individual-level data will not be shared and are not required for subsequent analysis.

(ii) Having created annual matrices describing the flow of patients between MSOAs/GPs and individual providers, the socio-demographic characteristics of the patients involved in each MSOA/GP provider link will be summarised by a single individual in the secure location. The age-sex profile of patients will be extracted from HES data itself, whilst patient-level linkage with Lower Layer Super Output Area (LSOA) level Index of Multiple Deprivation (IMD) data will describe the socio-economic characteristics of the areas in which they live. The goal here is to establish whether different providers, and different provider types, serve different cohorts and communities and whether this changes over time. The focus is solely on the more important MSOA/GP - provider links and summary statistics will not be calculated for individual MSOA/GP - provider links involving less than 5 patients.

(iii) A similar exercise will be undertaken to establish whether different providers and provider types serve cohorts with different clinical needs, and whether this changes over time. In this case a variety of traditional (Charlson Index, Functional Comorbidity Index, etc.) and bespoke measures of comorbidity will be implemented (using diagnosis fields, age, HRG coding and within-HES patient linkage to identify readmissions). Small counts will again be suppressed and only summary data will be shared with other members of the team.

(iv) Finally, HES data will be interrogated to establish whether there are systematic variations in waiting times between providers, provider types and provider environments (e.g. with respect to local indices of provider diversity and concentration).

The goal throughout is to use the richness of the pseudonymised individual-level HES data to derive a series of summary measures which capture (a) key aspects of the populations served by different provider types and (conversely) the provider types which serve different populations, (b) relationships between provider types (in terms of clinical/spatial overlap), and (c) evidence of variations in outcomes between provider types (e.g. waiting times and readmissions both overall and with respect to specific conditions and populations). These summary measures cannot be obtained other than through a bespoke analysis of HES data and will then be used, alongside data collected directly from a sample of providers, to better understand both the evolving provider landscape in general and the decisions and attitudes of managers working within that evolving provider landscape.