NHS Digital Data Release Register - reformatted

University Of Hertfordshire projects

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


Towards a value based healthcare system: assessing cost effectiveness of care pathways and mining for patterns and trends in the use of acute services — DARS-NIC-06381-N2B8J

Type of data: information not disclosed for TRE projects

Opt outs honoured: Anonymised - ICO Code Compliant (Does not include the flow of confidential data)

Legal basis: Health and Social Care Act 2012 – s261(1) and s261(2)(b)(ii), Health and Social Care Act 2012 – s261(2)(b)(ii)

Purposes: No (Academic)

Sensitive: Non-Sensitive

When:DSA runs 2019-01-01 — 2020-07-18

Access method: One-Off

Data-controller type: UNIVERSITY OF HERTFORDSHIRE

Sublicensing allowed: No

Datasets:

  1. Hospital Episode Statistics Accident and Emergency
  2. Hospital Episode Statistics Admitted Patient Care
  3. Hospital Episode Statistics Critical Care
  4. Hospital Episode Statistics Outpatients
  5. Hospital Episode Statistics Accident and Emergency (HES A and E)
  6. Hospital Episode Statistics Admitted Patient Care (HES APC)
  7. Hospital Episode Statistics Critical Care (HES Critical Care)
  8. Hospital Episode Statistics Outpatients (HES OP)

Objectives:

A day hardly ever passes in the UK without the NHS hitting the tabloids, with headlines such as “NHS is dying”, “NHS in crisis”, “A&E patients hit by winter crisis”, “NHS cuts 15000 beds in 6 years”, and “worst nurse shortage ever”. With ever increasing demand (mainly due to an ageing population), lack of resources and beds (due to severe financial cuts), coupled with the likely impact of Brexit on staffing within the NHS, which means such headlines will continue now and in the future.

To tackle these issues, since 2010 a team of academics at Hertfordshire Business School (HBS) at University of Hertfordshire has been developing healthcare models, in the form of simulation, statistical modelling, health economics and big data analytics. The goal is to assist key decision makers (such as directors of NHS Trusts, service managers and consultants) achieve the most efficient and effective delivery of high-quality care to patients. In collaboration with industry experts (both the NHS and pharmaceuticals) the team has developed many models for a wide range of diseases, e.g., COPD, ophthalmology (cataract and retinal services), Parkinson’s disease, acute services and neonatal care. For instance, Johnson and Johnson and Allergan helped with the conceptualisation of the ophthalmology pathway for cataract and intravitreal services, respectively.

The models are used in practice to measure the impact of changes, such as re-designing services, care pathway and practices, e.g.
• what happens if 10% of Ophthalmology patients are shifted into primary care;
• what happens if 3 additional consultants and 5 senior nurses for Parkinson’s disease are recruited;
• what if demand for 75+ age group increases by 10% across the board,
• where length of stay for the same age group decreases by 25%.
The models can be adapted to any NHS services, that is, the hospital management can input local estimates or data of service demands and capacities, to create a baseline model. The user can then compare the baseline with potential changes in the patient pathway in the safety of a virtual environment. By making such changes key decision makers can easily understand the impact on activity, cost, staffing levels, skill-mix, utilisation of resources and, more importantly, it allows them to find the interventions that have the highest benefit to patients and provide best value for money.
The University of Hertfordshire wishes to retain pseudonymized HES data for the purpose of continual development and implementation of the following projects (with relevant details provided for each project).

Ongoing Projects using HES
1. Title: Patient Pathway Modelling Using Discrete Event Simulation to Improve the Management of Chronic Obstructive Pulmonary Disease (COPD)
Objective: A decision support toolkit (DST) for improving the management and efficiency of COPD care is needed to respond to the needs of COPD patients now and in the future. Using the DST, the goal is to assess key changes in the COPD pathway, such as the impact of post-exacerbation pulmonary rehabilitation (PEPR) policy, increased use of community services and the introduction of smoking cessation. The study team will observe if these changes lead to improvements in quality adjusted life years (QALY), reduction in emergency readmissions, reduction in occupied bed days and associated costs.
Collaborators and their role: Royal Free London NHS Foundation Trust, as well as the local community COPD services, run by Central and North West London NHS Foundation Trust. Partner organisations play a crucial role in every aspect of the DST development, including conceptualisation of the COPD pathway; verification and validation of the simulation model and DST; health economic modelling; testing and running the model many times, and more importantly using the model in practice for change. They do not play a role in data controllership or data processing activities of data supplied by NHS Digital
Date started: September 2015
Target completion date and why: December 2020. The study team will need to test and implement the model at other NHS Trusts (and services), 1) to establish the usefulness of the model, 2) to be able to generalise the model for all NHS services in England (i.e. where necessary the model may need tweaking), and 3) evidence the impact on patients, NHS and beyond. These types of models are in its infancy and additional time is needed for it to be accepted by NHS key decision makers.
Funder: No additional funder is involved in this project.
Data processing: The University of Hertfordshire is the sole data controller and data processor of the data supplied by NHS Digital under this agreement. Royal Free London and Central and North West London NHS Foundation Trust are not involved in any data controller or processing activities of data supplied by NHS Digital
Is it part of a PhD project? Yes, the PHD project is entitled Patient Pathway Modelling Using Discrete Event Simulation to Improve the Management of Chronic Obstructive Pulmonary Disease (COPD) which is the same as the study title. The PHD student is a substantive employee of the University staff as they are employed to teach.

2. Title: A simulation-based decision support toolkit for informing the management of patients with Diabetes
Objective: Explore the impact of a range of changes to diabetes services as namely, what is the impact of increasing the number of diabetes specialist nurse (DSN) and ward based DSNs? What is the impact of increasing the number of patient referrals to community service? Can effective care be continued to be provided to patients in 12 months’ time with the existing resources?
Collaborators and their role: Northampton General Hospital. They do not play a role in data controllership or data processing activities of data supplied by NHS Digital.
Date started: September 2015
Target completion date and why: December 2020. The study team will need to test and implement the model at other NHS Trusts (and services), 1) to establish the usefulness of the model, 2) to be able to generalise the model for all NHS services in England (i.e. where necessary the model may need tweaking), and 3) evidence the impact on patients, NHS and beyond. These types of models are in its infancy and additional time is needed for it to be accepted by NHS key decision makers
Funder: No additional funder is involved in this project.
Data processing: The University of Hertfordshire is the sole data controller and data processor of the data supplied by NHS Digital under this agreement. Northampton General Hospital is not involved in any data controller or processing activities of data supplied by NHS Digital.
Is it part of a PhD project? Yes, the PHD project is entitled ‘A simulation-based decision support toolkit for informing the management of patients with Diabetes. which is the same as the study title. The PHD student is a substantive employee of the University staff as they are employed to teach.

3. Title: A discrete event simulation (DES) model to evaluate the treatment pathways of patients with cataract in the United Kingdom
Objective: A DES model was developed to re-design cataract pathway. The goal is to assess the possible changes to the configuration of cataract services to improve efficiency and enable them to cope with increasing demand. The study team will evaluate the impact of increasing the number of surgeries carried out per week.
Collaborators and their role: Leicester Royal Infirmary. The study team will observe if these changes lead to improvements in quality adjusted life years (QALY), reduction in emergency readmissions, reduction in occupied bed days and associated costs.
Date started: May 2017
Target completion date and why: December 2020. The study team will need to test and implement the model at other NHS Trusts (and services), 1) to establish the usefulness of the model, 2) to be able to generalise the model for all NHS services in England (i.e. where necessary the model may need tweaking), and 3) evidence the impact on patients, NHS and beyond. These types of models are in its infancy and additional time is needed for it to be accepted by NHS key decision makers.
Funder: No additional funder is involved in this project.
Data processing: The University of Hertfordshire is the sole data controller and data processor of the data supplied by NHS Digital under this agreement. Leicester Royal Infirmary do not participate in data controllership or data processing activities of data supplied by NHS Digital.
Is it part of a PhD project? No.

4. Name: Modelling the neonatal system: analysis of patient pathways and length of stay
Objective: An examination of the association between neonate’s pathway to discharge and length of stay within a neonatal setting. The goal is to establish whether there is a positive association between babies discharged home alive and how long they stay in hospital.
Collaborators and their role: University College London Hospital
Date started: September 2019
Target completion date and why: December 2020. The study team will need to test and implement the model at other NHS Trusts (and services), 1) to establish the usefulness of the model, 2) to be able to generalise the model for all NHS services in England (i.e. where necessary the model may need tweaking), and 3) evidence the impact on patients, NHS and beyond. These types of models are in its infancy and additional time is needed for it to be accepted by NHS key decision makers.
Funder: No additional funder is involved in this project.
Data processing: The University of Hertfordshire is the sole data controller and data processor of the data supplied by NHS Digital under this agreement. University College London Hospital are not involved in any data controller or processing activities of data supplied by NHS Digital
Is it part of a PhD project? No.

5. Name: A discrete event simulation model to evaluate the use of community services in the treatment of patients with Parkinson’s disease (PD) in the United Kingdom
Objective: A model is developed to represent the PD care structure including
patients’ pathways, treatment modes, and the mix of resources required to treat PD patients. Scenarios involving increased use of community services will be simulated to examine the impact on hospital activity, finances, consultant/nurse utilisation, and patient outcomes.
Date started: June 2018
Target completion date and why: December 2020. The study team will need to test and implement the model at other NHS Trusts (and services), 1) to establish the usefulness of the model, 2) to be able to generalise the model for all NHS services in England (i.e. where necessary the model may need tweaking), and 3) evidence the impact on patients, NHS and beyond. These types of models are in its infancy and additional time is needed for it to be accepted by NHS key decision makers.
Funder: No additional funder is involved in this project
Data processing: The University of Hertfordshire is the sole data controller and data processor of the data supplied by NHS Digital under this agreement.
Is it part of a PhD project? No.

6. Name: An innovative entire hospital level decision support system for efficient and effective delivery of healthcare services
Objective: To develop a whole hospital level decision support system (DSS) to assess and respond to the needs of local populations. The model integrates every component of a hospital (including A&E, 28 outpatient and inpatient specialties) to aid with efficient and effective use of scarce resources (e.g. staff and beds). For example, an evaluation of the impact of increasing demand for elective and non-elective inpatient admissions; the study team will calculate the required number of beds, consultation clinics slots and theatre capacity to ensure effective, efficient and timely delivery of care
Collaborators and their role: The Princess Alexandra Hospital NHS Foundation Trust. They do not play a role in data controllership or data processing activities of data supplied by NHS Digital.
Date started: September 2014
Target completion date and why: December 2020. The study team will need to test and implement the model at other NHS Trusts (and services), 1) to establish the usefulness of the model, 2) to be able to generalise the model for all NHS services in England (i.e. where necessary the model may need tweaking), and 3) evidence the impact on patients, NHS and beyond. These types of models are in its infancy and additional time is needed for it to be accepted by NHS key decision makers.
Funder: No additional funder is involved in this project
Data processing: The University of Hertfordshire is the sole data controller and data processor of the data supplied by NHS Digital under this agreement. The Princess Alexandra Hospital NHS Foundation Trust are not involved in any data controller or processing activities of data supplied by NHS Digital
Is it part of a PhD project? No.

As a research group, each member of the team is responsible for the development of the above models, made up of senior researchers and PhD students. Without HES the above models will not be possible to be developed, as data plays a crucial role in populating the model, not just for development purposes but for testing and implementing. Currently there is a phase of testing/implementing the models at other NHS Trusts, it is essential to extend the HES dataset. For example, the COPD model described in (1) will be tested at University Hospital Birmingham and Southampton General hospital; the cataract pathway model in (3) will be implemented at York Teaching Hospital NHS Foundation Trust; the entire hospital simulation model in (6) will be implemented at Norfolk and Norwich NHS foundation Trust to examine the impact of a 3 store building for cancer patients.

Note that PhD students are substantive employees at the University of Hertfordshire due to teaching commitments and are subject to the same disciplinary procedures that all members of staff are.
The University of Hertfordshire has determined that there are no moral or ethical issues from the dissemination of data for this purpose. Once received from NHS Digital, the data being processed by the University of Hertfordshire will be pseudonymised.

GDPR

In accordance with Article 6 (1)(e), data processing is necessary for the performance of the tasks specified above, which are carried out in the public interest or in the exercise of official authority vested in the controller. University of Hertfordshire is a public authority completing research for the public interest.

In accordance with Article 9 (2)(j) processing is necessary for archiving purposes in the public interest, scientific or historical research purposes or statistical purposes in accordance with Article 89(1) based on Union or Member State law which shall be proportionate to the aim pursued, respect the essence of the right to data protection and provide for suitable and specific measures to safeguard the fundamental rights and the interests of the data subject. University of Hertfordshire is a public authority completing research for the public interest.

Yielded Benefits:

In collaboration with the Ophthalmology services at Leicester Royal Infirmary the study team demonstrated the impact of re-designing their cataract pathway to key decision makers, made up of Ophthalmology consultants, service manager and senior nurses and showcased the possible changes to the configuration of cataract services to improve efficiency and enable them to cope with increasing demand. At the time of the study, according to the service manager, the number of patients waiting for cataract surgery (i.e. the backlog) was in the region of thousands. The findings showed how to increase the number of surgeries per list (e.g. from 6 surgeries to 8) thus increasing in the number of cataract surgeries carried out per year (approximately 36%) at no extra cost. Starting from a baseline of 5,542 surgeries per year, increasing to 7,521. The expected improvements at Leicester are in line with the model results. The findings from the model led to the approval of changes by the Chief Executive of Leicester Royal Infirmary Hospital and the service has increased the number of surgeries taking place each month, thus reducing the waiting list. This reduction in waiting times is vital to patients who, whilst cataract is a condition which is totally reversible, are at significantly increased risk of falls and potential fractures. Reduced vision is also known to inhibit patient’s capacity to socialise which is vital in maintaining a healthy state. This was achieved in January 2019. The Director of Market Access of Johnson & Johnson Vision has also expressed the usefulness of the model back in December 2018 (after having read the published article), particularly around evidencing the impact of change, such as increasing/decreasing human resources and the impact of increasing the number of surgeries per week. As demand for cataract procedures is already high and expected to increase sharply over the near future, Johnson and Johnson would like to support all NHS cataract services in the UK by developing similar simulation models, to evidence the process of re-designing their services and finding the most efficient and effective delivery of care to patients. The Parkinson’s disease (PD) project explored innovative ways of delivering care to patients with the specific focus on integrating Community Services (CSs) into the treatment pathways for PD patients. A simulation model was developed to test the possible impact of shifting more patients to CSs on both the operational and cost performance of the NHS PD care pathway. The model is being tested on several hospitals in the UK to determine the operational and financial gains, which the hospitals could achieve by adopting the results and recommendations of the research. Parkinson’s UK (a leading PD charity) is currently preparing a policy brief to be uploaded to their website, with the objective of promoting research, encouraging hospitals and primary care to increase use of CS, in line with The Department of Health and Social Care initiatives towards shifting services from hospital to CS.

Expected Benefits:

Without building the simulation models and populating them with the HES data it is impossible to predict what specific measurable benefits these Trusts will experience. However, the key purpose of the development is to enable numerous different service improvement options to be explored and understood.
The patient pathway simulations will provide invaluable tools for forecasting future activity for these Trusts. Often such forecasts take the Trusts months to prepare and are not statistically validated, resulting in inaccurate predictions. The simulations will replace this with a few minutes’ work.
The simulation models (or decision support systems) for these Trusts will:
• enable the full impact of changes to these Trust’s services to be explored in the safety of a ‘virtual’ environment. At present due to the complexity in services, changes are introduced without a full understanding of the impact they will have on other areas of the service.
• the outputs from the simulation (the results of the impact on resources, numbers of patient treated etc.) will be able to be used to support strong business cases for change to board-level executives and holders of budgets, thus dramatically shortening the time from identifying the service improvements to be made, to implementing the improvements. Currently this can often take around a year. Using the simulation, these Trusts will be able to shorten this to a matter of a few weeks.
• the simulation will be fully statistically validated so that the results can be relied on to be accurate. At present service improvements are often based on subjective opinions of what improvements should be made.

Outputs:

The primary output of this phase is the development of the toolkits. Secondary outputs will be the validation and utilisation of the toolkits to support local NHS Trusts.

From the perspective of aggregating data from HES to assess effectiveness and cost effectiveness of care pathways the outputs from the simulation models will include the following:

• Staff utilisation (in percentages) – Could staff numbers be reduced so that they can be used in other area? Do staff need to be relocated from quieter areas to the area being monitored? What percentage of staff resources are used for a particular specialty and could this be better managed?
• Bed/theatre/clinic slot utilisation (in percentages) – Are these being used effectively?
• Diagnostic/Treatment procedures (in numbers) – Were all procedures necessary? Could other treatments have been considered for the patient to reduce the time from referral to treatment (RTT 18-week pathway)
• Revenue and costs - To help finalise budget requirements for departments and aid in cost-saving across the NHS
• Patient Outcomes – study readmission rates and why they occur, look at mortality rates within a Trust to test effectiveness of pathway management using the tool.
• Cost benefit analysis of care pathways based on quality-adjusted life-years and incremental cost effectiveness ratio.

Using the above metrics, the research team will establish the care pathways that are found to be most effective and cost-effective. Furthermore, the team will capture the difference in patient pathways between providers and establish pathways that have the highest/lowest utilisation of resources and the pathways that have the highest/lowest cost of care. The pathways will be shared across NHS Trusts and providers so that they can be utilised to improve efficiency across the NHS.

These models will enable rapid exploration of future service capacity and demand across the Trusts. The outputs from the simulation will be graphs, charts and tabulations in aggregated small number suppressed. They will not contain any record level HES data. All findings will be presented to Trusts and conferences in the form of PowerPoint presentations.

Published articles for the projects described above
1) Demir, E., Southern, D., Rashid, S., and Lebcir, R. (2018). A discrete event simulation model to evaluate the treatment pathways of patients with cataract in the United Kingdom. BMC Health Service Research. 18(1):933. doi: 10.1186/s12913-018-3741-2.
2) Demir, E., Southern, D., Verner, A., and Amoaku, W. (2018). A simulation tool for better management of retinal services. BMC Health Service Research. 18(1):759. doi: 10.1186/s12913-018-3560-5.
3) Lebcir, M. & Demir, E., Ahmad, R., Vasilakis, C., and Southern, D (2017). A Discrete Event Simulation model to evaluate the use of community services in the treatment of patients with Parkinson’s disease in the United Kingdom. BMC Health Services Research. 17(1):50. doi: 10.1186/s12913-017-1994-9.
4) Demir, E., Vasilakis, C., Lebcir, M. & Southern, D (2015). A simulation-based decision support tool for informing the management of patients with Parkinson’s disease. 2015. International Journal of Production Research. 53, 24, p. 7238-7251.
5) Ordu, M., Demir, E., and Tofallis, C. (2019). A decision support system for demand and capacity modelling of an accident and emergency department. Health Systems. DOI: 10.1080/20476965.2018.1561161.
6) Ordu, M., Demir, E., and Tofallis, C. (2019). A comprehensive modelling framework to forecast the demand for all hospital services. The International Journal of Health Planning and Management. doi: 10.1002/hpm.2771.

Publications in progress for the above projects (2019/2020)

1) Davari, S., Demir, E., and Ordu, M. Healthcare Delivery Reorganisation Problem. Submitted to European Journal of Operational Research.
2) Demir, E. and Gunal, M. A decision support tool with health economic modelling for better management of Diabetes patients. In preparation for British Medical Journal.
3) Ordu, M., Demir, E. and Davari, S. A Hybrid Forecasting-Simulation-Optimisation Model for Multi-period Healthcare Resource Allocation. Submitted to European Journal of Operational Research.
4) Ordu, M., Demir, E., Tofallis, C. and Gunal, M. An innovative entire hospital level decision support system for efficient and effective delivery of healthcare services. Submitted to Decision Sciences.
5) Yakutcan, U., Demir, E., Hurst J. and Taylor, P. The impact of PEPR and increased use of community services for COPD patients: a simulation and health economic modelling approach. In preparation for European Respiratory Journal.
Conferences

The study team will be attending the following conferences in 2019, disseminating the findings to a wide range of national/international academics and practitioners:

1) Operational Research Applied to Health Services (ORAHS). 28 July - 02 August 2019, Karlsruhe, Germany. https://orahs2019.de/
2) Production and Operations Management Society (POMS). 2 – 4 September 2019, Brighton, UK. https://www.poms2019.com

Processing:

The datasets will be held on a password protected server and encrypted drive at the Hertfordshire Business School. The room can only be accessed by a dedicated card reader who are given permission by the estates. Maximum security is in place, e.g., there are always securities patrolling the campus. Access to the room and data is restricted to members of the research team employed by the University of Hertfordshire who are specifically assigned to work on this study. All members have a full-time contractual work agreement, i.e., permanent members of staff.

The data will be used exclusively for the purpose of the specified studies. The data will not be made accessible to any third parties. At the end of the study the data will be safely held on a password protected and encrypted drive at the Hertfordshire Business School and accessed only to answer questions arising from any publication and other publicity if required Subject to a current data sharing agreement with NHS Digital being in place to hold data.

The modelling framework is generic such that it could be applied to any disease area or patient pathway, hence the extracted data is always the same and will therefore be extracting/processing the following data from HES:

• Monthly number of inpatient elective/non-elective admissions.
• Outpatient attendances, follow-ups, did not attends and cancellations. These will be counted for each month (i.e. 36 historical data points) broken down by specialty.
• Weekly A&E arrivals (i.e. counts), number of laboratory tests and investigations
• Average length of stay
• Average waiting time for an Outpatient appointment
• Average number of follow-ups
• Percentage of transfers to inpatient services from A&E
• Average cost of care using HRG tariff codes
• Monthly counts of patient readmissions
• Monthly counts of diagnostic procedures

These will then become key input parameters for the simulation model (discrete event simulation using SIMUL8). There will be no data linkages or matched with other publicly available datasets. There will also be no requirement nor attempt to re-identify individuals, nor attempt to re-identify individuals, there will be counting based on monthly activity there will be no small dataset issues either, i.e., less than 100 records.

Individual records at the episode level (with non-sensitive data) over an extended period is necessary, as it is vital to be able to track individual patients over a period of time to inspect changes in the pathway, e.g., the number of follow-ups in 2006 may not be the same as 2007, 2008 and 2009. The same applies to other metrics of interest, including length of stay, waiting time for appointments, cost of treatment, etc. Episodes are used to establish a patient’s treatment journey by the involvement of a variety of consultants during their stay in hospital.

Furthermore, a patient’s journey typically involves treatment within an entire acute setting, where many fields within A&E, inpatient and outpatients are extracted for analysis. Therefore, it is difficult to reduce the number of datasets and fields that are requested, as missing information within a patient’s pathway can lead to inaccurate models, limiting its use in practice.

The study team work very closely with a large body of collaborators from the NHS (see projects listed above), geographical data is crucial. For instance, the COPD model developed in collaboration with the Royal Free Hospital will be tested at other Trusts, e.g., University Hospital Southampton NHS Foundation Trust, Oxford University Hospitals NHS Foundation Trust, etc. Time to time a new NHS Trust may be part of the study, thus having the entire dataset enables the team to draw their expertise in to the research.

The study team do not develop models for specific demographics (e.g. age groups). It is applicable to all patient categories, thus narrowing down to demographics is not possible.

Data processing will only be carried out by members of the team who are appropriately trained in data protection and confidentiality. For instance, all members of the team have attended the latest GDPR course back in April 2019 at the University of Hertfordshire. This was a compulsory attendance due to the involvement of the team in the use of patient level data (i.e. HES).