NHS Digital Data Release Register - reformatted

Mckinsey & Company

Project 1 — DARS-NIC-368233-L2N0W

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) — 2021/04.

Repeats: Ongoing

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

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


NHS organisations, including NHS Trusts and Foundation Trusts, CCGs, CSUs, and NHS England, commission McKinsey and Company to work on projects which are procured by the NHS organisation within and outside of specific procurement framework agreements. The scope of this work is developed by the client organisation and covers a broad range as specified by the client including strategy, performance transformation, and organisational development. Examples are listed in the “specific outputs” section. McKinsey and Company use HES and PbR data in order to provide fact-based answers to McKinsey’s NHS clients questions regarding identification, assessment and quantification of opportunities to improve the quality and efficiency of the NHS services that they deliver, or are responsible for overseeing and regulating. McKinsey and Company have applied for permission to use both HES and PbR datasets because, though derived from the same source, they each support different aspects of advanced analytics: (a) PbR allows the most accurate assessment of income and expenditure under the activity-based funding model, Payment by Results, used in the NHS, while, (b) HES allows for the most granular analysis of operational performance, capacity, utilisation and demand. McKinsey and Company have applied for a license renewal for HES and PbR data from 2012/13 to the present (ongoing monthly managed service subscription) in order to be able to look at trends in performance, expenditure, utilisation and demand. The specific purposes and types of analysis that McKinsey perform are the following: (1) Benchmarking and analysis of operational performance (2) Benchmarking and analysis of variation in utilisation rates and tariff spending (3) Analysis of historic trends in rates of activity and spending (4) Analysis of the impact of different service configuration options HES and PbR data will only be used in the context of services by McKinsey in England and will not be used for non-NHS (or social care) organisations or for organisations outside of England. McKinsey and Company are requesting to maintain access to three years of historical data in order to monitor trends in performance, expenditure, utilisation and demand. Access to three years of data allows for the analysis of trends to identify cyclical patterns in utilisation as well as directional trends in performance, while also allowing for the identification of anomalies in activity. Furthermore this permits the measurement of the effectiveness of performance and cost improvement initiatives such as in tracking activity and expenditure following implementation of a cost improvement plan or QIPP initiative.

Yielded Benefits:

McKinsey has used HES data to support its work with NHS clients for the past 10 years. The benefits of some of this past work is set out in a range of case studies below. 1. Addressing a London health economy deficit of £40m (2017) McKinsey & Company completed a 5 week project in July 2017, working with a London health economy to address a system deficit of £40m. The project team used HES data to drive benchmarking of the CCG and trust’s historic performance against comparators and review internal activity trends such as A&E and ambulatory care sensitive condition attendance rates by GP practice across the borough. The team used these benchmarks to align the different health economy partners (including CCG and trust) around a shared understanding about the drivers of deficit, and identify opportunities to drive improvements to care quality and access. The benchmarking exercise using HES data has helped the local health economy identify geographic areas in need of short-term resource investments to improve access to urgent care. In particular, detailed historic benchmarking indicated that the health system had significantly higher than extended spending on acute care, and non-elective admissions in particular. Both the CCG and the trust agreed that an investment in community services would dramatically improve healthcare access and quality, as well as their deficit. The CCG and trust are now in ongoing discussions to move towards joint ways of working towards improving care and reducing the system deficit. 2. Financial recovery and improvement planning in the Midlands (2017). McKinsey & Company completed a 4 week study in June 2017 to provide financial recovery, improvement and sustainability support to a health economy in the Midlands. The health system had delivered their lowest level of QIPP savings since 2013, and were seeking to develop recurrent and transformational QIPP plans for the current fiscal year. The project team used HES data to benchmark the CCG’s historic performance against comparator CCGs, as well as to compare internal variability in secondary care activity by GP practice. These benchmarks were used to assess the size of the improvement opportunity in the region, evaluate the ambition of current QIPP schemes, and support the development of detailed delivery plans to implement the schemes. The outputs of the financial review, including the use of HES-derived outside-in productivity benchmarks, supported the delivery plans of 15 QIPP initiatives with an expected savings of c. £30m at the end of the fiscal year. 3. Access improvement for elective care at a teaching trust (2017) McKinsey completed an 8-week study in July 2017 to support access improvement to elective care for an NHS teaching trust serving a catchment population of 650,000. The project team used HES data to develop a single version of the truth on planned care performance to align stakeholders to a common understanding on changes in demand and their drivers over time. Analyses included an historical review of elective care activity volumes across inpatient and outpatient settings, compared volume increases to patterns in referral to treatment waiting times, and peer benchmarking on performance measures with other trusts. The analysis revealed that increases in volumes were driven by referrals from out of area CCGs, and by faster than average rise in consultant to consultant referrals. This analysis also supported subsequent prioritisation of improvement initiatives and the quantification of their impact. The anticipated impact of the planned interventions, once fully implemented, include: - Removed outpatient waiting list backlogs across five priority specialities within 12 months - Reversed deterioration in inpatient backlogs across five priority specialities within 12 months - Streamlined and more convenient services, such as faster email and telephone advice, one-stop shops to reduce patient visits, and greater use of patient decision aids and decision making in their treatment 4. Productivity of an ambulance trust (2014) McKinsey has been engaged in ongoing work since 2014 with an ambulance trust to review productivity opportunities across their operations. The client was facing deteriorating operational performance and was looking to implement a new operational approach. The project team performed detailed analysis and modelling of patient demand, service capacity and service efficiency. HES data were used to model historic trends in conveyances to A&E. Insights derived from HES analysis formed the basis for discussions with stakeholders and experts to diagnosis drivers of deteriorating performance. HES data was also used to compare historic performance prior to the adoption of a new operational pilot, with trust-supplied data following implementation. Impact of the new operating model was validated using actual observed pilot data. This included a 10% improvement in the proportion of ambulances arriving on scene within 8 minutes. 5. Financial improvement for an acute trust in the North of England (2016) McKinsey completed a 12 week project in July 2016 leading a large acute trust in the North of England through a large-scale financial improvement programme. The client faced an underlying financial challenge of £90m, having historically achieved ~£45m in annual financial improvements. McKinsey worked in consortium with MoorHouse and Four Eyes to review financial improvement opportunities across the whole of the hospital system. The project ran across two phases, with the first 2 weeks dedicated to a rapid baseline assessment to identify top-down opportunities. During this period the McKinsey team used HES data to benchmark productivity KPIs such as case-mix adjusted ALOS and historic changes to activity within key specialties against comparator trust peers to size the overall productivity potential. In the second 10-week delivery phase, the consortium supported hospital divisions to develop and strengthen plans for delivery around the nursing workforce, medical workforce, theatres, outpatients, length of stay and admin and clerical workforce. The team’s work, supported by peer benchmarks using HES data, helped to strengthen 300 existing financial turnaround initiatives and identify an additional 100 plans, for a total achieved in-year savings of £79m (or £100m in annualised savings). 6. Clinical service redesign for an NHS trust (2016) McKinsey completed an 18 week project in clean sheet redesign across six functional and clinical service lines in November 2016. The team used HES data to conduct a diagnostic of orthopaedic productivity metrics including length of stay, operations per consultant, DNA rates and activity rates. The trust’s performance was benchmarked internally across the hospital sites, and nationally against comparable trusts. Metrics were designed to align with national best practice. Analyses of case-mix adjusted length of stay were conducted to confirm that the trust’s higher length of stay post-surgery was related to productivity rather than complexity of cases. McKinsey worked with a triumvirate of consultant, nurse and manager from the service line to develop aspiration targets derived from the benchmarking. HES outputs were presented to a broad range of staff in large design workshops in terms of aggregated PowerPoint tables. The result of the work has been an end to end pathway redesign built around the aspirational productivity metrics and agreed upon by the hospital’s clinicians and non-clinical leads, and a modelled impact of the redesign on the people and infrastructure requirements. . The end to end pathway redesign is expected to improve patient care by improving quality of care in line with national best practice guidelines, reducing variation in clinical quality, and reducing referral to treatment times. Referral to treatment times are expected to fall from the current median of >36 weeks to 5 weeks.

Expected Benefits:

Benefits achieved to date are: 1. McKinsey and Company completed a project with a large acute Trust to support them in developing a reconfiguration strategy, affecting a number of adjacent NHS acute Trusts. The purpose of the work was to improve clinical care by organising services in such a way that national clinical standards, in particular for 7 day working and 24/7 consultant cover in critical care, could be met; and to address a forecast deficit of £150m by 2019/20 through identifying efficiency opportunities and new sources of income. The outputs of the work included operating models by specialty (and setting), showing workforce requirements, activity, costs and income, for a range of different configuration options. Models were developed using Hospital Episode Statistics and Payment by Results data as the source for baseline activity and income assumptions and involved the analysis of the impact of different service configuration options as described above. McKinsey and Company prepared a set of written reports setting out the implications of each option and supporting documents for use in stakeholder engagement and alignment activities. The work lasted approximately 4 months finishing in May 2015. Following the work, the hospital has transitioned from being an amalgamation of four discrete sites, into a cohesive hospital with a clear identity, vision and strategy shared across management, front line staff and service lines. Furthermore, the hospital has moved to the front of the pack in being clear and assertive about its vision, and is now busy implementing it, working in a very complex external environment. (2) McKinsey and Company worked with a large acute Trust for approximately four months to support them to improve overall performance in their A&E department and, in particular, to increase the proportion of patients treated within 4 hours in line with national quality targets (of 95% achievement), and to improve clinical care for all admitted and non-admitted emergency patients. Outputs included a detailed modelling of patient flows through the Emergency Department; analysis of patterns of demand for A&E services nationally, sub-nationally and locally; identification of the most useful performance metrics to support management of the emergency pathway; analysis of the impact of A&E demand and performance on the rest of the hospital (in terms of non-elective admissions; bed occupancy rates; length of stay; delayed discharges) for all medical and surgical specialties relevant to the emergency pathway. Benchmarks of operational performance (as described above) compared the trust against size and activity profile for similar NHS trusts across England. Benchmarking analysis was used to develop targets and conclusions about demand factors (demand factors measured using the types of analysis described in analysis of variation in utilisation and analysis of historic trends as described above) affecting emergency care performance. The work concluded in February 2015. Since then the trust has seen an improvement of 11.5% percentage points in the achievement of their waiting times standard (with the proportion of patients waiting less than 4 hours in A&E averaging at 95.9% in June 2015). (3) McKinsey and Company worked with a group of NHS organisations across a large local health economy to benchmark operational performance in A&E and non-elective inpatient care looking at demand, capacity and productivity. The outputs included analysis of trends in demand overall and for specific segments including out-of-area patients, elderly patients and other sub-groups; detailed analysis by admission type and method, case mix, treatment received and length of stay. This involved analysis of operational performance, variation in utilisation and historic trends as described above. The output resulted in a written report including an agreed set of performance indicators. The project took 12 weeks and completed in February 2015. Since then the participating trusts have seen an average improvement of 5.5% from their Q3 waiting time standards achievement to the end of June 2015, compared to a national improvement of 2.6 percentage points. In addition to impact on the headline waiting times in A&E units, a number of other direct benefits were realised from this work, including: • Reduced length of stay for long-waiting patients including those experiencing delayed transfers of care (DTOC) • Improved real time data across the whole urgent care system enabling more effective escalation and therefore preventing risks associated with “surge” impacts on A&E departments or assessment units • Improved understanding of system wide capacity and demand, allowing Systems Resilience Groups to plan interventions to re-balance capacity Expected future benefits for individual projects vary, but in almost all cases involve identification and quantification of opportunities to improve the quality of patient care and population health, and to deliver more effective, efficient care. Target dates (for expected improvements) also vary but in almost all cases are within 3 years and often include within year opportunities for service improvements and/or savings. Examples of the benefits expected for the projects described above are: (1) The outputs of the analysis of different configuration options which will support the reconfiguration programme are expected to improve patient care by supporting the delivery of 7 day services and 24/7 consultant-led care in critical care. The absence of 7 day services and 24/7 consultant-led services have contributed to observed excess mortality at nights and weekends in NHS hospitals. The reconfiguration programme is also expected to deliver £26-36m of cost savings through more efficient use of estates and workforce; and to support the delivery of a further £111m in identified cost improvement programme (CIP) savings opportunities over the next two years. (2) The outputs of the operational performance benchmarking, variation in utilisation benchmarking and analysis of historic trends (described above) undertaken for this project looking at the performance of the A&E department in a large acute Trust, are expected to increase the proportion of patients in A&E treated within 4 hours from just over 80% (rolling weekly average prior to this work) to over 95% within a four month period. This will improve patient care and experience. Associated objectives include improving health outcomes by quicker placement of patients into appropriate wards and on swifter discharge placement in alternative care settings. (3) The outputs of the operational performance benchmarking, variation in utilisation benchmarking and analysis of historic trends (described above) undertaken for this project looking at how to improve the delivery of non-elective health services across a large local health economy, are expected to support NHS Trusts to improve care for patients arriving at A&E by increasing the proportion of patients treated within 4 hours and to allow providers and commissioners to develop the most appropriate range of services for people in need of urgent care. Improvements are expected within a time-frame of 6-12 months.


It is not possible to provide full details of all specific outputs and timings because McKinsey work on multiple projects for a large number of different national, regional and local organisations across the NHS, including providers, commissioners and regulators. Some examples of outputs expected are set out below. During the course of the projects that McKinsey do with NHS organisations, McKinsey test the data and analysis with McKinsey’s clients, and where necessary update and replace the data with summary data provided by them. This is the case with commissioners and providers, but is not always possible due to limitations in analytical capabilities, resources, and their own access to data. Data is only shared with clients in aggregated, non-patient identifiable formats with small numbers suppressed in line with the HES analysis guide’. McKinsey shares outputs in the following ways with clients: • McKinsey include aggregated, non-patient identifiable data in line with the small numbers guidance into Excel models which McKinsey hand over to the NHS client • McKinsey publish graphs based on the aggregated, non-patient identifiable results of quantitative analysis in line with the small numbers guidance in reports given to McKinsey’s NHS clients • McKinsey present the aggregated, non-patient identifiable results in line with the small numbers guidance at meetings with NHS client stakeholders McKinsey do not directly publish the outputs in any journal articles or other public documents (e.g., white papers) nor do McKinsey directly present any data outputs Current projects with NHS clients requiring access to HES and SUS PbR data include: • Ongoing 18 month project with a London CCG to review their strategic and organisational development due to end in November 2016 • Ongoing 3.5 month strategy review with an acute trust in Northern England, due to end in December 2015 • 2 month urgent care pathway study with a London acute trust scheduled to start in September 2015 and run through to mid November 2015. • 2 month health system reconfiguration study with a CCG collaborative in Midlands and East scheduled to start in September 2015 and run through to mid November 2015. • 10 week hospital productivity study with 8 London trusts McKinsey are currently planning to start in September 2015 and run through to December 2015. • McKinsey are currently responding to a tender for the provision of Health Economy Success Regime support for a large CCG in the South of England which would last 3 months and is estimated to end in December 2015. The data or outputs may not be used (directly or indirectly) for sales or marketing purposes by McKinsey Ltd or by any other non NHS organisation and can only be used for the purposes of the promotion of health.


Data is extracted from the SAS (http://www.sas.com) database in which it is stored in the form of data queries and these are analysed further in Excel. McKinsey and Company currently only use SAS and SAS Enterprise Guide to extract the data. Further analysis on extracted data is currently conducted in Excel spreadsheets. (1) Benchmarking and analysis of operational performance McKinsey have a standardised tool which is created annually using HES data. This tool is a "Hospital Diagnostic" which compares all NHS acute Trusts on a range of operational performance metrics (including case-mix adjusted average length of stay, day case and day of surgery admission rates by setting and specialty; proportion of A&E attendances resulting in admission by length of stay of that admission etc) against a peer group (tailored to each individual Trust). This analytical tool is created in Excel. McKinsey also conduct ad hoc analyses for the same measures to look in more detail at performance, for example at site level, or for specific types of patients (e.g. sub-groups defined by age, gender and diagnosis cluster). Ad hoc analysis is conducted in excel using subsets of data extracted using standardised data queries from the SAS database. (2) Benchmarking and analysis of variation in utilisation rates and tariff spending McKinsey have standardised approaches to measure variation in utilisation rates (by setting, patient type or demographic sub-group, specialty and different activity clusters) and associated tariff expenditure both within (at GP practice level) and between CCG commissioner peer groups (defined using ONS cluster groupings). Utilisation is measured as an activity rate (or associated tariff value) per 1,000 ageneeds weighted population (or most appropriate population measure) and compared to CCG (or GP practice) peer group median, quartiles and deciles. This analysis is conducted in excel using subsets of data extracted using standardised data queries from the SAS database. (3) Analysis of historic trends in rates of activity and spending Operational performance and utilisation rates are measured over time at different frequencies, including yearly, monthly and weekly, in order to understand cyclical patterns and directional performance trends. This analysis is conducted in excel using subsets of data extracted using standardised data queries from SAS. (4) Analysis of the impact of different service configuration options HES data is used to develop best estimates of baseline activity and capacity (defined as bed days for admitted patient care) for commissioners and providers, aggregated at service line level (defined by specialty and point of delivery). This is then forecasted forward using a range of sources of insight, data and triangulation methods (including, but not limited to, local and national historic trends described above), to develop growth assumptions and scenarios. A simulation is created, in excel, to analyse how these baseline levels would change over time if service configuration changed.