NHS Digital Data Release Register - reformatted
NHS North and East London Commissioning Support Unit
Project 1 — DARS-NIC-371243-H1P5T
Opt outs honoured: No - data flow is not identifiable (Does not include the flow of confidential data)
Sensitive: Non Sensitive
When: 2016/04 (or before) — 2019/01.
Legal basis: Health and Social Care Act 2012, Health and Social Care Act 2012 – s261(1) and s261(2)(b)(ii)
Categories: Anonymised - ICO code compliant
- Hospital Episode Statistics Accident and Emergency
- Hospital Episode Statistics Admitted Patient Care
- Hospital Episode Statistics Critical Care
- Hospital Episode Statistics Outpatients
- Diagnostic Imaging Dataset
- Bridge file: Hospital Episode Statistics to Diagnostic Imaging Dataset
Some of the yielded benefits of HES data to date are: 1) Assisting NHS England in London, at identifying the rates of GP referrals to hospitals across all the CCGs in London. There is a lot of variation, both between CCGs and between GP Practices within a CCG area and NELCSU have compared these in a way that accounts for these areas very different populations (in statistical terms this is known as “standardisation”. The NHS is experiencing significant pressure and unprecedented levels of demand. The average annual growth in GP referrals between 2009/10 and 2014/15 was 3.9%. Growth in 2015/16 compared to 2014/15 was 5.4%. For the same period, other referrals, which include consultant to consultant referrals grew by 6.7%. There is clearly a significant need for the NHS to manage the demand that flows into hospitals by ensuring that only the most appropriate cases are referred for face to face consultation. There is also evidence to suggest that a referral to hospital is not always necessary. NHS England have published a demand management “Good Practice Guide” covering areas such as “peer review of referrals”, “shared decision making” and “advice and guidance”. The results of this analysis help identify which geographies to target, inform the conversation around appropriate areas to change and help monitor the impact of any implemented demand management schemes. 2) Assisting a CCG in the South of England implement improvements in the area of diabetes and respiratory disease. The improvements will involve the health system – GPs, hospitals, community services – working more effectively together (or in the jargon working in a more “integrated” way). HES data has been used to identify variation in “outcomes” to identify potential areas to target. For example HES data was used to identify the numbers of patients admitted with complications of diabetes, as these are an indicator that a patient’s condition has deteriorated, something that could possibly be counteracted with better management of a patient’s condition within primary care. Are these numbers high relative to other areas? How much do they vary by GP practice? What is the real reason behind this variation? Integrated care schemes internationally have evidenced significant benefits in improving patient outcomes, experience of care and reducing costs to the health system. This project is the first stage of a pilot, and will be extended out to a wider geography and to other clinical areas. 3) Identifying potential influences on high A&E attendance rates. Within London, most A&E departments are under huge pressure from rising A&E demand. However, the rate of increase does vary significantly by geography and by patient group. By using national HES data NHS North and East London Commissioning Support Unit were able to undertake statistical modelling of most of the known drivers of A&E attendance and try to understand the relative importance of each. For example, patient ethnicity appears to be one influence. Those ethnic groups with high attendance rates can be targeted through communication campaigns or through their GPs to encourage use of alternative services where possible. The better that the reasons for growing A&E demand can be understood, the more effective commissioners can be in tackling the root cause of the issue.
The HES data will only be used by North and East London Commissioning Support Unit (NELCSU) and its approved users (via secure access to the HES Cube) to support CCGs to meet their statutory duties under the Health & Social Care Act 2012 and NHS health economy wide transformation projects that require detailed hospital level data. The data is proposed to be used on two levels: Objective 1 HES extracts: pseudonymised record level HES data is interrogated only by approved in-house NELCSU analysts to provide benchmarking and comparative information to NELCSU CCG clients and NHS health economy wide transformation projects that require detailed hospital level data. The full set of HES data allows complex and detailed modelling and benchmarking of activity, essential to successful commissioning of services and contract monitoring, including analysing relationships and influences between A&E, Inpatient and outpatient care. This will especially support benchmarking work for CCG clients taking part in health economy wide transformation projects that require detailed and comprehensive hospital level data (for example from local SUS data feeds the applicant only receives data for their CCG’s registered population, which does not allow whole trust activity to be considered) and allow CCGs to benchmark and identify best practice in similar health economies anywhere in England. Outputs released will be aggregated with small number suppression and provided, by data sharing agreement, only for use in the NHS Health and Social Care sector to meet statutory requirements under the Health and Social Care Act 2012 and is not for commercial use. Objective 2 HES Cube: to populate a a stand-alone secure data tool containing summary stand-alone data to allow a broader range of users (than objective 1) without access to the full HES datasets to get quick and easy access to information on common and simple counts of hospital activity, for the same purposes as output 1, to support the successful commissioning. This tool can generate small numbers from its restricted number of derived data fields (please see ‘Processing activities’ section below for further information on what the tool contains). Any output released from the cube must be aggregated and have small numbers suppressed (see Summary section above for details of data sharing agreement in place). The HES Cube will be used by the following: - in-house NELCSU analysts, - by approved analytics staff in NELCSU customer CCGs and by approved public health analysts within Local Authorities(LA) within CCG customer areas that have signed a local data sharing agreement (see Appendices A and B – supporting documentation SD3 and SD4) which include agreement to comply with the HES protocol and suppress the release of small numbers. This permits access to the tool and to aggregated summary data containing small numbers. Access to Local Authority Public Health teams is given to support CCGs to meet their obligations to Public Health teams through the provision of aggregated data in support of joint duties to under the Health & Social Care Act 2012. Public Health teams lost existing access to HES and SUS data in 2013 when they moved from PCTs to Local Authorities, and therefore their ability to support CCGs has been difficult since. These purposes are; Aggregate outputs with small numbers suppressed to support health needs assessment, including supporting the delivery of Joint Strategic Needs Assessments (JSNAs) [joint with CCGs] Aggregate outputs with small numbers suppressed to support the development of CCG commissioning intentions, commissioning strategies and QIPP plans Aggregate outputs with small numbers suppressed to support Public Health activity relating to protecting the health of the population in XXXX LA’. Recipients are not permitted to onwardly share the data unless the HSCIC gives explicit permission for them to do so and should permission be granted this would be for anonymized aggregated data with small numbers suppressed only. Access to the HES data will not be given to any other organisations, other than those stated above.
CCGs and Local Authorities (Public Health teams) have joint statutory duties under the Health and Social Care Act 2012 to plan and commission services and jointly assess the needs of their patients and populations, to ensure that health improvements and better outcomes are measurable, identifiable and attainable. Access to HES data helps these organisations achieve this by providing the greatest scope to evaluate outcomes of care and improvement in their health services against peer groups and national achievement – providing a more extensive and complete base of knowledge for decision making than data on their own patients alone (SUS data). Measurable benefits can occur, for example, through gradual improvement in outcome over a number of years, to more immediate commissioning new services where a gap is identified, or de-commissioning failing services by identifying lower outcomes than is acceptable, compared to the norm.
Objective One and Two Outputs are on an ongoing basis (i.e., no target date) as the HES data in general and the HES Cube are used to support general commissioning and public health needs, and is not aimed at a specific report or deadline for use. All outputs informed by information retrieved from the HES data tables and the HES Cube are governed by adherence to the HES guidance on suppression of small numbers. Users of the data are bound by signed data sharing agreements (see Appendices A and B - supporting documentation SD3 and SD4) with conditions that they abide by HES guidance which means that all outputs released must be aggregate with small numbers suppressed. HES allows NELCSU to provide intelligence for programmes whose scope demands activity benchmarking of the CSU's clients (CCGs) against similar health economies or populations in England. SUS data does not allow this scope. National data also supports NHS health economy wide transformation projects that require detailed and comprehensive hospital level data Commissioners can compare with any service known to have better outcomes or new pathways, or support large scale transformation projects that may impact several commissioners across, for example, the North London area. Outputs expected are aggregated data to support reports or decisions across examples such as the following: • Elements of Joint Strategic Needs Assessments (JSNA) - to support CCGs/Local Authorities to consider the needs of their local populations and in how they respond with effective commissioning of services to properly meet those needs, by enabling, for example views of the use of secondary services by different patient groups by condition, ethnicity, etc. • Quality, Innovation, Productivity and Prevention (QIPP) development - identifying and benchmarking areas across England with better practice than locally, to help evaluate high costs and poor outcomes in hospital care. • Providing data on hospital admissions in-year to support monitoring of national ambitions, such as avoiding unnecessary admissions across CCGs, by practice, condition, hospital trust. CCGs are required to monitor and make progress on national outcome measures and ambitions by NHS England, and use of national benchmarking is promoted heavily by initiatives such as Right Care ‘ Commissioning for Value’ (on behalf of NHSE). Without access to national data such as HES, CCGs cannot be ultimately certain that they are making progress or making decisions on the best basis possible. Aggregated outputs will only be used for NHS Health and Social Care use; or by organisations who are commissioned by NHS Health and Social Care for a specific purpose as outline above.
Objective 1 Only an approved list of NELCSU analysts have access to the full set of pseudonymised data tables, via secure server based (Structured Query Language (SQL) querying. CSU analysts interrogate the data to produce aggregated output for monitoring care outcome and activity for a CCG’s population, patient group, condition or service provider, including trends over time in any given activity or care process. For example, trends over time can be modelled to produce forecasts of future activity, taking into account population growth or changes in service configuration. National data is necessary to benchmark against any CCG peer groups (as defined by NHS England), or any other care pathway or group of patients. Benchmarking allows an individual CCG to evaluate its own care processes and outcomes against other similar commissioning populations, with a view to identifying areas for improvement or to identify best practice. National data also supports NHS health economy wide transformation projects that require detailed and comprehensive hospital level data Analysts must only release aggregate data with small numbers suppressed. Objective 2 HES Cube: The HES Cube seeks to provide a quick and flexible way to get information on common questions on hospital activity, such as: • How many admissions, what was the source and type? • How much activity by diagnosis, procedure, HRG? • How many discharges, how long were stays in hospital? • How much activity does a population generate – by resident, commissioner? • How do age bands, gender, ethnicity relate to activity? To do this, NELCSU summarises data in a selection of HES fields to create a stand-alone data table aggregated with small numbers included. The full list of HES fields is shown below in this section. Data periods included are from 2007 to the latest held. A software application (currently Microsoft Excel) is then applied to a summarised data table to enable the user to produce different views of the data according to their needs, via a simple pivot table presentation. No other data is included in the cube or can be linked to it. For any user (both in-house and external clients, i.e., customer CCGs and associated Public Health Teams), the HES Cube as a package is made available in an existing secure online environment, the NELCSU Information Exchange (NELIE). Only named approved users are allowed to access NELIE, and the list and access rights are maintained by NELCSU. NELIE users agree to a data sharing agreement. Users have varying access rights to levels of NELIE information, which includes, for example, general health service activity and contracting/financial reporting, all governed by agreements with the applicants CCG clients. Access specifically to the HES Cube itself is governed by access rights separate from the other information on NELIE. As specified above, external analysts who need access to the HES cube (such as Public Health teams) must sign a data sharing agreement that includes abiding by the HES protocol and they also need approval from their CCG partners. External clients are restricted to two analysts each with access to the HES cube. The “HES Cube” is a self-contained application with no links to other datasets. Also there are no record level identifiers within the cube that would make this possible and the data is aggregated by default but small numbers may be included. What users actually see: The users see a pivoted summary view of counts of hospital activity in which any of the 20 fields can be included as variable to enable flexible and meaningful counts of data. Although a drill down facility exists this will only drill to the limited amount of summary data physically held within the cube. For the tool to be flexible and capable of sufficient combinations to support a useful and meaningful range of work, it must contain enough variables (HES data fields) to meet these needs. As a by-product (but not a requirement) of the combination of variables, it is possible to drill down to counts of small numbers in the tool. The cube summary table contains 20 fields (out of 400+ possible already pseudonymised HES fields) with derived items such as age banding (rather that single ages) and Ward level geography (instead of Lower Super Output Areas or postcode), to help ensure that the occurrence of small numbers is reduced to a minimum. The majority of output from the cube will be unlikely to result in small numbers due to its prime use for benchmarking of a large number of providers/population areas - but any aggregate output with small numbers is required to be suppressed before wider dissemination, as specified in the data sharing agreements. The set of HES data fields processed into a separate data table for the HES Cube are the following: Admission date [ADMIDATE] - Admission Method [ADMIMETH] - Five Year Age Bands derived from [STARTAGE] - Diagnosis [DIAG_01] - Discharge Date [DISDATE] - Discharge Destination [DISDEST] - Discharge Method [DISMETH] - Electoral Ward [RESLADST_CURRWARD] - Episode Order [EPIORDER] - Episode Type [EPITYPE] - Ethnicity [ETHNOS] - Gender [SEX] - Commissioner CCG [GPPRAC] - GP practice [GPPRAC] - HRG [SUSHRG] - IMD Decile [IMD04RK] - Patient Classification [CLASSPAT] - Primary Procedure [OPERTN_01] - Responsible Commissioner [PCTCODE06] - Service Provider [PROCODE] - Source of Admission [ADMISORC] - Specialty [MAINSPEF]