NHS Digital Data Release Register - reformatted

Methods Analytics Ltd

Project 1 — DARS-NIC-09519-D5G0R

Opt outs honoured: N, Y

Sensitive: Sensitive, and Non Sensitive

When: 2016/04 (or before) — 2018/05. breached contract — audit report.

Repeats: Ongoing, One-Off

Legal basis: Health and Social Care Act 2012, Section 42(4) of the Statistics and Registration Service Act (2007) as amended by section 287 of the Health and Social Care Act (2012)

Categories: Anonymised - ICO code compliant, Identifiable

Datasets:

  • Hospital Episode Statistics Outpatients
  • Hospital Episode Statistics Admitted Patient Care
  • Hospital Episode Statistics Accident and Emergency
  • Standard Monthly Extract : SUS PbR A&E
  • Standard Monthly Extract : SUS PbR APC Episodes
  • Standard Monthly Extract : SUS PbR APC Spells
  • Standard Monthly Extract : SUS PbR OP
  • Hospital Episode Statistics Critical Care
  • Bridge file: Hospital Episode Statistics to Diagnostic Imaging Dataset
  • Diagnostic Imaging Dataset
  • Bridge file: Hospital Episode Statistics to Mental Health Minimum Data Set
  • Mental Health and Learning Disabilities Data Set
  • Bridge file: Hospital Episode Statistics to Mortality Data from the Office of National Statistics
  • Office for National Statistics Mortality Data
  • Mental Health Services Data Set
  • Secondary Uses Service Payment By Results Spells
  • Secondary Uses Service Payment By Results Outpatients

Benefits:

Benefits relating to each of the purpose statements is listed below: 1) Stethoscope Free (formerly Acute Trust Quality Dashboard) – free to the NHS and the public is information tool showing aggregated indicator data across the domains of the NHS operating framework. This has significant usage across the NHS with hundreds of visits each period and users can download a free pdf report, with approximately 30 free pdfs downloaded by users each period. The free public Stethoscope website was used as input for the Keogh mortality reviews and is visited by Monitor, CQC and NTDA among many others. Methods are aware that the free pdf download is used to inform Trust boards, having been asked for permission by Trust secretaries. Stethoscope Subscriber – a password protected secure service offered with an annual subscription to cover the costs of data hosting and processing, licensing for Qlikview, development of the tool and hosting user groups. This offers users much more frequently than publically available sources updated indicator data with the ability to ‘drill’ into the data and filter by different options to provide insight and understanding of the quality of care. Users would be assigned access to the tool by an ‘administrator’ in their organisation and examples of users include Trust Chief Executives, Medical and Nursing Directors, Specialty Managers, Clinicians and Information Departments. CCG, Local Area Team and commissioning region subscribers may grant access to the tool for use by Quality Managers, Public Health analysts, Commissioning Managers and Executives. It is important for Methods Analytics to work with their customers to ensure they can interpret the data and use it to take appropriate actions to safeguard against excess mortality and reduce mortality and improve the quality of care where possible. Many indicators are available dealing with Quality and Safety issues NHS Organisations face to allow decision makers to take actions based on up to date information. Methods have CCG, provider and NHS England regions as subscribers with over 100,000 page views per year and 100% contract renewal from subscribers indicating the value of the system to NHS users. The Stethoscope subscriber system has been used to support Quality Surveillance Groups, Quality Summits, and board to board oversight meetings. 2) Methods Analytics work with Trusts and CCG’s, and wider programmes such as the DH GIRFT team to provide ad hoc reporting matching their requirements, using HES/SUS and SUS PbR data as appropriate to derive insight into a specific topic or issue. A real life example is a review of urgent care within an NHS Trust: Methods Analytics used HES data to build a picture of issues around urgent care including where patients are flowing from, how referral patterns are changing over time and conversion rates that was used by the organization to initiate a transformation programme and improve urgent care timeliness and outcomes. Similar projects focusing on mortality have resulted in large and lasting reductions in hospital mortality. Projects include a large amount of clinical engagement to ensure that data in the reports is used in the best way possible to make changes to services that benefit patients in any organisation working with Methods Analytics. The Analytics team includes clinicians and consultants to provide the right expertise when discussing any insight with Methods Analytics’ NHS clients. Methods are also working with the DH GIRFT programme, generating report across 12 specialty areas under Lord Carters NHS efficiency programme. These detailed data rich reports are shaped by national lead clinicians for each specialty and they then visit every provider in England to discuss their data with them in order to improve the quality and efficiency of care. This programme is currently being rolled out. 3) The NSCC dashboards and PET tool was developed in partnership with the Royal College of Surgeons and NHS England Rightcare programme to support the work of the National Surgical Commissioning Centre http://www.rcseng.ac.uk/healthcare-bodies/nscc. They developed commissioning guides for CCGs on specific interventions which all have NICE accreditation. As part of this work, Methods Analytics developed the PET tool to allow commissioners to access data to support the guidelines. The commissioning Guides are approved by the National Institution for Clinical Excellence and together with the data tools are used by commissioners across England to improve services for patients and monitor those improvements. RCSE has made the guidelines publically available and also the data tools in line with the requirements of the governments transparency agenda. Therefore there are no ‘customers’ as the tool is available to all. This tool has been live since 2012 and Methods Analytics has been recontracted, funded by NHS England, to maintain and enhance these tools. The tool receives of the order of 350 hits per month from NHS and wider public. 4) SWORD is a tool developed with the ALS and AUGIS to provide to their consultant surgeon members detailed, clinically valid metric that report activity, quality and outcome metrics for surgical pathways. The tool is now live in a development state for user validation and testing, with consultant surgeons starting to request, and being provided with, access. Wider roll out is ongoing, including developing relationships with other surgical specialties. By allowing surgeons to see how their quality of care varies from other surgeons performing the same operations they can work to improve the levels of care they are able to offer and improve the safety for patients they are operating on, in order to get a full understanding it is important they are able to identify themselves in the tool. Surgeons can also use the data could also be used for revalidation purposes therefore ensuring patient safety by providing evidence a surgeon is up to date and fit for practice.

Outputs:

Outputs for the data will be as follows and are related back to the 4 purposes: 1) Stethoscope – Live. Tool developed as Ruby on Rails bespoke web tool combined with Qlikview 11 dashboards https://stethoscope.methods.co.uk. 2) Ad hoc reports and bespoke tools – ongoing. Created using MS Office Suite , Excel 2013, Word 2013, Qlikview and Tableau and also rendered as .PDF format for final reports on highly aggregate data. These include multiple reports on Mortality and Emergency care are for CCGs, providers and NHS England regions to support understanding of causes of failure and direct service improvement initiatives. Methods are also supporting the DH GIRFT programme, a wide ranging programme to improve secondary care quality and outcomes, with bespoke analysis and reporting for every provider in England. 3) Royal College of Surgeons – Live dashboards and tool developed in Qlikview11 called the “Procedures Explorer Tool” (http://rcs.methods.co.uk/pet.html). 4) SWORD – live (development phase with restricted users for validation). http://www.augis.org/sword/ All outputs are suppressed in line with the HES Analysis Guide.

Processing:

For all purposes above the data is made available to Methods Analytics through HSCIC’s secure ftp, after which it is imported into directly into Methods Analytics SQL data warehouse that is hosted as discrete physical servers by Redcentric. The process is handled by a single Data Base Administrator, , as per the HSCIC guidance who has the password for the secure ftp.. This process means the individual will set off a set of automatic instructions to import the data into SQL via an ‘SSIS package’. The package itself handles the data import process. Redcentric provide rack space, power, internet connectivity (controlled by physical firewalls) and high level server management support (such as server system software patching). They do not have any access to data within the servers. Methods Analytics will install, maintain and operate all non-operating system software and have sole access to the servers. The installed software will be MS SQL Server 2014,and R server statistics package. Methods Analytics users will have to complete a Data Centre Access request form which is signed off by their line manager before being granted access the server. using 2 factor authentication, encrypted, VPN. The VPN supports the use of both SafeNet Software based tokens and hardware based tokens – each of these types requires a pin-code in order to generate a one-time password for the VPN. Each token is only usable on one device. The use of tokens restricts each Methods Analytics user to a single computer, with a drive encrypted using Microsoft Bitlocker. With the unique token issued to each member of Methods Analytics staff who need access, this identifies them discretely and incorporates user level access control limiting access to tables and data at a per user level. Data on the servers is encrypted using XTS AES 256-bit encryption at rest. All processing will be undertaken within the server environment. No episode level data will leave the secure environment. Once the record level data has been processed, only anonymous, aggregated data (with small numbers suppressed in line with the HES analysis guide), is transferred out of the secure Redcentric environment, as described in individual sections below. No record level data is provided to any third party, and all record level or aggregated data (small numbers unsuppressed data) is held at RedCentric. The development of Qlikview tools will be undertaken in the secure Redcentric environment. 1) HES/SUS PbR Data is processed into indicators within the Redcentric environment and the resulting aggregated data (with small numbers unsuppressed) dataset is transferred to a Qlikview server which is also within Redcentric over a secure internal unidirectional VPN tunnel. Suppression of small numbers as per the HES Analytics Guide is applied by the Qlikview product as the application is used, and thus ensures that no small number unsuppressed data is available to the user. Stethoscope’s functionality built in the web for providing indicator Alerts and MyView uses a different data model to Qlikview. The data is still processed in Redcentric and suppressed in line with HES Analysis Guide. Whilst the Analysis Guide does permit small numbers at certain geographical levels, Methods Analytics apply small number suppression to any low numbers in the data table regardless of the level of aggregation (i.e. Regional, Provider etc) before it is transferred to an Amazon Web Service instance (in the EU Ireland region), where the data is restored to a SQL database which serves the web product. To be clear, only aggregated small number suppressed data (anonymous data therefore) is held or processed within Ireland. 2). HES/SUS PbR Data is processed into indicators and counts within the Redcentric environment and undergo a process to create an anonymous, small number suppressed tabulation. Those tabulations, which are suppressed in line with the HES Analysis Guide, are transferred via encrypted VPN to encrypted PCs/laptops in order to build reports using a suite of business intelligence software consisting of MS Office, MS PowerBI, Qlikview or Tableau. Reports are also rendered as Adobe PDF documents before being distributed. 3) HES/SUS PbR Data is processed into indicators and counts within the Redcentric and the resulting aggregated (small numbers unsuppressed) dataset is transferred to a Qlikview server which is also within Redcentric over a secure unidirectional VPN tunnel. Suppression of small numbers as per the HES Analysis Guide is applied at the application layer (Qlikview) and ensures that small numbers are not available to the user. 4) HES/SUS PbR Data is processed into indicators and counts within the server environment and the resulting pseudonymised dataset is transferred to a Qlikview server which is also within Redcentric over a secure unidirectional VPN tunnel. Suppression of small numbers as per the HES Analysis Guide is applied at the application layer (Qlikview) and ensures that no numbers <=5 are available to the user at this time, except where a consultant views their own data, in which case no suppression is applied.

Objectives:

The data will be used to support the NHS (specifically (NHS England, CCGs, CSUs, providers of NHS funded care and professional bodies) through the delivery of tools and bespoke analysis. Existing work with NHS partners includes ‘GIRFT’ (Getting It Right First Time) for the Department of Health and National RightCare for NHS England. Methods Analytics have also been commissioned for other projects by the NHS both directly and through frameworks such as OJEU and G-Cloud. The data is also used for Stethoscope - a quality variation tool which provides national benchmarking of HES based indicators. Methods Analytics will also be using the data to help provide insight to commissioners and providers through tools and dashboards that will allow users to explore indicators of quality/safety and Patient Experience in their organisations. Users will be provided with indicators showing national benchmarking and statistical techniques are applied to the data to show variation and difference in services (for example against 2 and 3 Standard Deviations from national mean). Absolutely no raw data will be shared outside of the Methods Analytics team. There are five uses of data requested (and each is discussed further within the processing, outputs and benefits section). The specific uses are :- 1) 1) For Stethoscope - a quality variation tool which provides national benchmarking of HES based indicators that is made available free to the public at an organisation roll up level, and more granular information to subscribing NHS organisations (NHS England, CCGs, CSUs and providers of NHS funded care). Consultant code will also provide a further level of drilldown in the Stethoscope product to provide Trusts to explore and understand the variation in care between their own consultants across Method Analytics’ indicator set. Access to the more granular tool is provided securely to named subscribers only, with individual surgeons able to compare themselves to a national cohort of surgeons. Access controls restrict access to consultant identifiable data so that only authorised staff at an individual Trust can only see data for their own employees, and such data is suppressed in line with the HES analysis guide. No access to servers containing HES data is possible through Stethoscope as the Stethoscope servers are not linked in any way to the secure environment. Only aggregated data (suppressed in line with the HES Analysis Guide) is surfaced through Stethoscope. 2) For bespoke tools and analysis for individual NHS clients (NHS England, CCG and providers of NHS funded care), CSUs and professional bodies. The majority of these reports contain data items from Stethoscope but are reported as dashboards for individual organisations. They also contain bespoke metrics generated from HES data presented as aggregated (small number suppressed in line with HES Analysis Guide) tabulated data and/or charts and graphics, and can have accompanying narrative interpretation. 3) For creating and hosting dashboards and an explorer tool developed with the surgical associations working group under a NICE accredited methodology. This is work for the National Surgical Commissioning Centre, hosted by the Royal College of Surgeons of England and part of the NHS England Rightcare programme. These tools show activity rates and simple outcomes for CCG populations and care providers using HES/SUS PbR data. These tools are free to the public. 4) The requirement for the consultant code is to support a project for the specialist surgical societies to develop an intelligence tool (SWORD) for their members to access measures and metrics about their own performance which will be accessible via the association’s members’ portals (therefore password protected). Only consultant surgeon members of the associations can access the SWORD tool. This is further secured by the request for access being generated by the association and sent to Methods Analytics, with Consultant name, GMC number and nhs.net email address that is used for communication with the individual. Method Analytics creates an account for that consultant and the user has to separately log into the SWORD tool. When the user logs in the system validates a link between their user name and GMC number, so when they click the consultant view they see only their own data with a national mean. At this level only, data for the named consultant is visible with small number suppressed in line with the HES Analysis Guide. However as requested by the surgical associations individual consultants may see their own activity and outcomes without suppression, and national mean data (with small numbers suppressed in line with the HES analysis guide)to enable local discussion amongst surgeons of low volume activity and outcomes. There is no option to view other consultant’s data in this view. If the user does not have a valid GMC number linked to their user account, then when a user clicks on consultant view no information is presented. 5) The HSCIC developed the Summary Hospital Mortality Indicator (SHMI) and provides quarterly publications for each Trust in England. This includes an observed number of deaths within that period that occurred in hospital plus the number of deaths which occurred within 30 days of discharge from hospital. Using the HES-ONS linked dataset, Methods Analytics were able to reproduce the exact methodology and figures in a timely manner which will allow subscribing NHS medical directors, chief executives, clinicians and managers to explore how the SHMI has changed over time and how their own trust is performing against other trusts in the country in terms of mortality rates. This means the data can be used to identify any issues and to improve the quality of care and to reduce patient mortality. ONS data has been used to create SHMI and variants of SHMI and include it as content in items 1-4 above as an indicator. Under this application, the ONS data is no longer held and thus cannot be reprocessed to create new such indicators, but indicators already produced may continue to be used.