NHS Digital Data Release Register - reformatted

Optum Health Solutions Uk Limited projects

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

Bespoke Extract Request for producing benchmarks within products — DARS-NIC-277499-D3D0X

Type of data: information not disclosed for TRE projects

Opt outs honoured: No - data flow is not identifiable, Anonymised - ICO Code Compliant, No (Does not include the flow of confidential data)

Legal basis: Health and Social Care Act 2012, Health and Social Care Act 2012 - s261 - 'Other dissemination of information', Health and Social Care Act 2012 – s261(1) and s261(2)(b)(ii), Health and Social Care Act 2012 – s261(1) and s261(2)(b)(ii), Health and Social Care Act 2012 – s261(2)(b)(ii)

Purposes: Yes (Supplier)

Sensitive: Non Sensitive, and Non-Sensitive

When:DSA runs 2019-12-01 — 2020-11-30 2017.12 — 2020.10.

Access method: One-Off, Ongoing

Data-controller type: OPTUM HEALTH SOLUTIONS UK LIMITED

Sublicensing allowed: No

Datasets:

  1. Bespoke Extract : SUS PbR A&E
  2. Bespoke Extract : SUS PbR OP
  3. Bespoke Extract : SUS PbR APC Episodes
  4. Bespoke Extract : SUS PbR APC Spells
  5. Secondary Uses Service Payment By Results Accident & Emergency
  6. Secondary Uses Service Payment By Results Episodes
  7. Secondary Uses Service Payment By Results Outpatients
  8. Secondary Uses Service Payment By Results Spells
  9. Secondary Uses Service Payment By Results Accident & Emergency

Objectives:

Optum Health Solutions (UK) Ltd (Optum UK) requires HES/SUS data to calculate benchmarks within two products: the population health analytics tool Health Population Manager (HPM) and the Commissioning & Contracting suite (CCA). Work is undertaken because of requests received from NHS clients - specifically Clinical Commissioning Groups (CCGs), Commissioning Support Units (CSUs), NHS Trusts, NHS England and GP Practices. Clients need to understand how they are performing in different areas (e.g. patient pathways) compared to other similar organisations or national benchmarks. Having the HES/SUS data enables the team to provide this analysis to clients either through HPM or the CCA tool. An example includes CCGs who need to identify Quality, Innovation, Productivity and Prevention (QIPP) initiatives to improve the quality of care they deliver while making efficiency savings that can be reinvested into the NHS. By benchmarking against peer CCGs, it can identify the areas where CCGs can make improvements.

Optum Health Solutions (UK) Ltd currently provides commissioning support services to over 20 CCGs across Lincolnshire, Thames Valley & Wessex and Derbyshire. All of these CCGs require support in benchmarking analysis. Some of the NHS clients would have access to aggregated views of the SUS data in the products so they can carry out benchmarking analysis themselves – access is only given to clients who need such access. All aggregations are compliant with the HES Analysis Guide, with small numbers suppressed in line with the guide. In terms of geographies, the data is aggregated by postcode district. The Optum UK team also provides reports to clients after analysts have analysed the data and added interpretation and insight.

These benchmarks are proportions from aggregate numbers produced at various levels from diagnosis and procedure combinations to Healthcare Resource Groups and Episode Treatment Groups. An example might be the proportion of all non-elective admissions that have a length of stay of zero days. The data is used to estimate the impact of changes in policy from the Department of Health such as changes to the Payment by Results (PbR) guidance or specialised Information Rules. Similarly, if a provider was to ask for a local tariff modification claiming that they were unfairly penalised under PbR then the benchmarks would be used to see if the data backed this up. Sometimes when discussions are under-way with either an existing or a potential new client, the benchmarks might be used to give an indication of whether a proposed service change would indeed save money or improve services.

To do this Optum Health Solutions (UK) requires national HES/SUS data so its team can benchmark clients’ hospital activity numbers against other similar organisations to identify whether there is potential to improve services or to amend tariffs.

This is a long–running project and HES APC,OP,A&E and CC pseudo/anonymised data has been supplied under the legal basis Health and Social Care Act 2012 –S261(1).

This amendment is to now request SUS PbR data on a monthly basis, on receipt of the historical SUS PBR data for 2014/15, 2015/16 and 2016/17, Optum will destroy the HES data so only four years of data is retained at any one time. The benefit to Optums customers of having monthly SUS PbR data is that the data will remain up-to-date for them so that decisions are based on the latest available data and the data will include tariffs calculated by NHS Digital that will match local SUS data, thus providing consistency and like-for-like comparisons. In the past, with the HES data, by the time the annual extracts were received and then loaded and processed them through Opium’s system, the data is already over a year out of date, which means customers are making decisions based on old data. Benchmarking against the latest year of data ensures customers can see where they are outliers compared to similar organisations and have the time to address issues before they can escalate and grow. The tariffs in the PbR data will enable Optum and customers to benchmark against costs/spend with the assurance that the tariffs have been accurately calculated by NHS Digital. As an example, in the past Optum have identified for clients where their hospital providers are outliers for a number of key performance metrics such as readmissions within 30 days, first to follow up ratios and average length of stay for different specialties. As the data was already a year out of date, the clients were not receiving the latest view of the provider performance and were not able to effectively challenge the providers because of this. The providers argued that their performance had improved in the subsequent year and Optum had no recent comparison points to challenge this. There is more and more demand for data as close to real-time as possible so that changes can be made within health systems when it really matters. The monthly PbR data will allow Optum to provide information that is far more actionable than yearly HES data. Monthly data allows Optum to align analysis to the latest data potentially available to clients comparing latest provider performance against like for like periods, and providing the most accurate timely analysis and comparisons between providers/geographies.

The data is used only for NHS clients or organisations who work with the NHS.

Yielded Benefits:

Optum UK have used the data to develop RightCare reports and other benchmarking reports for the CCGs Optum UK support in Lincolnshire and Milton Keynes. This has identified to the CCGs where they are outliers and allowed them to investigate further to develop QIPP schemes. Optum UK have also used the data in Thames Valley to identify where provider casemix is inconsistent to support contracting discussions. In Buckinghamshire, the CCG was alerted to an issue with excess bed days by NHS Improvement. Optum UK were able to use the national data to verify that the local SUS data was correct and the issue was not as serious as highlighted by NHS Improvement.

Expected Benefits:

Optum Health Solutions (UK) Ltd currently provides commissioning support services to over 20 CCGs across Lincolnshire, Thames Valley & Wessex and Derbyshire. The measurable benefits of services are preventing unnecessary acute admissions, ensuring earlier intervention in primary care and better control of patients’ conditions. In addition, commissioners save money that is then reinvested to provide better services for their target populations.

Please see the below case studies from Thames Valley & Wessex on the CCA tool.

For Thames Valley & Wessex CCG clients, Optum Health Solutions (UK) Ltd undertook a benchmarking analysis exercise using the HES data to identify pathways where improvements could be made to improve outcomes, spend and patient care. One of the pathways identified was Coronary Heart Bypass Graft (CABG) procedures at the Royal Brompton Hospital (RBH). The data indicated that the length of stay for patients in critical care following the procedure was higher at RBH than other providers.

Optum Health Solutions (UK) Ltd mapped out the patient pathway at RBH and then compared with evidence based best practice to measure the gap in provider performance. This revealed that the average number of bed days in the Intensive Therapy Unit (ITU) and the High Dependency Unit (HDU) at RBH were 1.8 days for each, compared with the best practice of 1 day according to Milliman guidelines. Further analysis also identified that not all patients were having a pre-operative assessment, with best practice indicating that this would reduce the length of stay in hospital. Optum Health Solutions (UK) Ltd also found that not all patients were having a routine follow-up appointment at RBH and some patients were having a follow-up at home with a homecare nursing service.

On behalf of the CCGs, Optum Health Solutions (UK) Ltd negotiated with RBH and agreed a change in the CABG pathway to:
• Reduce critical care bed days from 1.8 to 1;
• Decommission the homecare nursing service;
• Commission outpatient first and follow-up attendances with surgeon/physician for pre-operative assessments.

This resulted in a better experience for patients in line with best practice and a reduced cost for the CCGs. Overall, £223,072 was identified as potential health economy savings and £150,000 implemented within the contract with RBH.

The outputs of the benchmarking analysis as outlined will be used to help the BLMK STP bridge its financial funding gap. Optum Health Solutions (UK) Ltd supported the STP in calculating its total cost of care and projected financial funding gap. Using the HES data, Optum UK benchmarked BLMK against other regions in England to identify opportunities for savings and better care. The benchmarking analysis was used to identify QIPP and CIP savings initiatives.

Optum Health Solutions (UK) Ltd is looking to use the HES/SUS data in a similar way for other clients, including the Lincolnshire STP, the Modality vanguard and the Symphony vanguard. This would include the use of RightCare methodology to identify savings initiatives for Optums clients to inform QIPP and CIP schemes.

This amendment is to now request SUS PbR data on a monthly data instead of the yearly HES data. The benefit to customers of having monthly SUS PbR data is that the data will remain up-to-date for them so that decisions are based on the latest available data and the data will include tariffs calculated by NHS Digital that will match local SUS data, thus providing consistency and like-for-like comparisons.

Outputs:

The data is used to sense check data Optum Health Solutions (UK) Ltd get from CCGs or spot areas that might be worth further investigation by Optum UK’s client(s). For instance, Optum UK might spot that an area has a particularly high level of neurology compared to the rest of the country and suggest to their client to look at local feeds to see if there really is or not and if so, find out why. These outputs are used for existing clients only, the data is not used to identify new clients.

When reports are provided to clients, data is only included at an aggregated level, for example, proportion of hernia daycase admissions as a proportion of all admissions. Small numbers are suppressed in line with the HES analysis Guide. No record level data ever leaves the server and is only used for the purposes listed above.

Optum Health Solutions (UK) Ltd has used the HES data in CCA for sense checking data and benchmarking received from a number of clients in early 2016 (verifying that the local data looks accurate in terms of total admissions/attendances) and for benchmarking purposes. For example, Optum Health Solutions (UK) Ltd supported the Bedfordshire, Luton and Milton Keynes (BLMK) Sustainability and Transformation Plan (STP) group in understanding the total cost of care across the region today and in the future.

Optum Health Solutions (UK) Ltd collated local activity and cost data from across the region and developed an actuarial model to forecast activity and cost 10 years into the future (the unmitigated trend). Using the HES data, Optum UK then benchmarked BLMK against other regions in England to identify opportunities for savings and better care. The results of this analysis were then used to create mitigated activity and cost trends. This information was used by the BLMK in their STP submission to NHS England in October 2016.

Optum Health Solutions (UK) Ltd is looking to use the HES/SUS data in a similar way for other clients, including the Lincolnshire STP, the Modality vanguard and the Symphony vanguard. This would include the use of the NHS RightCare methodology to identify savings initiatives for the clients to inform Quality, Innovation, Productivity and Prevention (QIPP) and Cost Improvement Plan (CIP) schemes.

The outputs derived from HES/SUS data allow comparisons between the client’s organisation or population and others nationally for a range of performance metrics such as lengths of stay, emergency admissions, A&E attendance, etc. These National comparators are used by NHS organisations to improve the quality of care delivered by comparing their performance as set out by a specific range of care quality and performance measures, detailed activity and cost reports. The comparators are also used in service redesign and Health Needs Assessment (identifying underlying disease prevalence within the local population compared with the national picture). A customer typically looks at areas of activity that they are outliers for and use these as a way of prioritising service redesign activity and to target areas of deeper analysis and service improvement using more detailed data sources.

Processing:

Optum Health Solutions (UK) Ltd will store the data on a SQL server located in a N3 connected data centre in Heathrow, England [premises rented from SunGard Availability Services]. SunGard will have no access to the data or servers, but provide a hosting service via N3 for Optum Health Solutions (UK) Ltd’s servers - the infrastructure is owned by Optum UK but the premises and network connections to the N3 are rented from SunGard. Optum staff in the UK will remotely access the server via dedicated Virtual Private Network (VPN) over N3, additionally protected by a VPN token and login. Access is on a strictly limited basis. The server is only accessible by a limited number of Optum Health Solution (UK) Ltd staff, with the Optum UK I.T. infrastructure team having access to the HES server but not the databases thereon.

Data will only be accessed by Optum staff in England who have authorisation to access the data for the purpose(s) described, all of whom are substantive employees of Optum Health Solutions (UK) Ltd. The data is not accessible to Optum staff outside of England. Authorisation to access the data is granted through a request process by Optum’s Caldicott Guardian and Senior Information Risk Owner (SIRO).

Optum UK analysts will load SUS data into a SQL server, which is dedicated solely for the processing of national SUS data. There are no other data sets stored on the server other than lookups for codes. Within SQL server, stored procedures have been developed to calculate or look up various fields in an automated way (for example Healthcare Resource Groups, specialised service types and costs). After that the relevant counts are taken and proportions calculated, the historic SUS data is used to calculate a probability of admission for various conditions / interventions. In addition, similar proportions or percentages are used to spot coding anomalies or unusual intervention rates in different providers, much as the Atlas of Healthcare Variation does. These aggregated results are transferred to the production servers and fed into various models, predominantly in Excel although other statistical software is also used.

The monthly SUS PbR data will also be loaded into the Health Population Manager (HPM) system and aggregated by the system to GP Practice level for benchmarking purposes. The tool has a graphical user interface that is available to NHS clients via a N3 URL. Access is governed by secure roles based access control (RBAC), with users requiring a username and password to log in to the system. The user is able to select a "norm" (benchmark organisation) to benchmark their organisation against. Norms are available at GP Practice, locality and CCG level. The users have no access to the underlying patient-level data.
No record level data therefore ever leaves the server.

All Optum UK analysts complete annual training on handling sensitive records.

All processing is done on a standalone physical server which is physically sat behind the N3 firewall on its own subnet at the SunGard data centre. The data will not be linked with any record level data. There will be no requirement nor attempt to reidentify individuals from the data. The data will not be made available to any third parties other than CCGs, CSUs, NHS Trusts, NHS England and GP Practices. The data will not be available to any third parties except in the form of aggregated outputs with small numbers suppressed in line with the HES Analysis Guide.

The number of years of data is requested as four years of data gives sufficient data for trending. On receipt of the historical SUS PbR data for 2014/15, 2015/16 and 2016/17, Optum will destroy the HES data prior to 2013/14 so that Optum have four years of data. Once Optum have a full year for 2017/18 Optum will destroy 2013/14 HES data. Annually Optum will apply for a renewal to obtain a further year of monthly extracts and once Optum have received a further year the oldest year of data will be deleted (to maintain a rolling 4 year dataset). Having 4 years of data is the minimum required to effectively forecast activity and spend in the future. The four years allows Optum to effectively factor in seasonal and disease-cycle variation.

Optum Health Solutions (UK) Ltd have a number of Clinical Commissioning Groups (CCG) clients spread across England and benchmark against their RightCare peers, which are typically CCGs in another part of the country. The RightCare peers for a CCG are the 10 most similar CCGs in the country as determined by NHS RightCare. The Optum UK team also benchmark against national numbers and other breakdowns from a national level, such as quartiles and deciles. The national data gives the flexibility to benchmark in all these different ways.

Optum Health Solutions (UK) benchmark high-level utilisation such as total elective admissions and total non-elective admissions. The analysis does not have a specific set of conditions that are used for benchmarking purposes. The conditions used are dependent on the client and the issues that they have in their area. This is often dictated by the initial analysis that NHS RightCare provide, which provides an indication of the patient pathways that could be improved. NHS RightCare is a proven approach that delivers better patient outcomes and frees up funds for further innovation. It gives CCGs and local health economies practical support in gathering data, evidence and tools to help them improve the way care is delivered for their patients and populations. Optum UK supports the CCGs in gathering data, evidence and tools, including benchmarking tools and analysis as outlined here.

None of the clients have access to the raw patient-level HES/SUS data. The only access they have is to aggregated data in web-based tools. The websites can only be accessed from within N3 by approved users. Users require login details to access the tools and their access is roles based (RBAC). Small numbers are suppressed in line with the HES Analysis Guide.

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 ie: employees, agents and contractors of the Data Recipient who may have access to that data).
The Data will only be used for the purposes described in this agreement.