DART: The Integration and Analysis of Data using Artificial Intelligence to Improve Patient Outcomes with Thoracic Diseases.
University of Oxford · Academic
In term In term in the September 2026 edition: the latest version runs to 19 March 2027.
- Reference
- DARS-NIC-668928-L9N5F
- Current version
- v2.11
- Term of current version
- 20 March 2026 to 19 March 2027
- Start date
- 26 October 2023
- Data controller
- Sole Data Controller
- Commercial purposes
- Yes
- Sublicensing
- No
- Files released to date
- 28
Why the data was released
Objective for processing
The University of Oxford requires access to NHS England Data for the purpose of the following research project:
DART (The Integration and Analysis of Data using Artificial Intelligence to Improve Patient Outcomes with Thoracic Diseases)
The following is a summary of the aims of the research project provided by University of Oxford:
DART is a study developing integrated diagnostics to enable the earlier diagnosis of lung cancer. This study, which began in 2020, aims to improve patient survival, yielding time and cost efficiencies to the NHS. The overall objective of the study is to link data from the Targeted Lung Health Checks (TLHC) programme, collected by the Lung Health Centres and sent to Oxford University Hospitals NHS Foundation Trust (OUH), with NHS England Data to better predict those who will benefit from lung cancer screening and to develop artificial intelligence systems that could improve understanding (through machine learning) of co-morbidities and confirm or improve standard clinical practices. Costs and survival will be summarised for groups of patients for various health states. It is anticipated that limited data will be received on secondary care resource use via the TLHC Programme. Based on initial inspections, this data will be fragmented, will cover a more limited follow-up period, and will not be available in detail for all participants in the TLHC programme. The Data requested from NHS England under this Agreement is expected to considerably enhance the study by providing less biased data and ensuring that missing secondary care resource use data is minimised.
Another objective of DART is to improve and enhance the existing evidence base on outcome prediction for patients with lung cancer. The requested Data will be used to identify patients with lung cancer who relapse or die; analyses using these Data will assess the efficacy of current and proposed clinical approaches.
OUH has been granted and funded for a Trusted Research Environment (TRE). The data from DART will be stored in this TRE, which will be managed in accordance with the relevant approvals from the Confidentiality Advisory Group (CAG), health Research Authority (HRA), and the Research and Development Department of the OUH. Access to the TRE for researchers will be governed by the above regulations and will require detailed applications and approval by the TRE Data Access Committee (TRE DAC). The NHS England data requested for DART will be stored in the DART section of the TRE.
DART has a total of nine work packages
• WP1 Leadership and Project Management
• WP2 Planning for, and management of Datasets, and Integration
• WP3 AI Model Validation on Lung Cancer Screening (LCS) data
• WP4 Digital Pathology AI and Radiomics Model Development and Validation
• WP5 Chronic obstructive pulmonary disease (COPD)/ Coronary artery disease (CAD) Model Development and Validation
• WP6 Primary Care/Population Health and Health Economics
• WP7 Integration of Blood Biomarkers
• WP8 Outcome prediction and new treatment paradigms
• WP9 Quality assurance: radiologist training and monitoring
The requested Data will be used in multiple DART work packages (WPs 4, 5, 6, 8) to further its overall aims:
> WP4: This WP relates to development of an artificial intelligence algorithm to diagnose lung cancer from digital pathology images. The HES data will enable ground truthing (a Gold Standard that can be used to compare and evaluate model results) of cancer diagnoses and for the algorithm to learn which of the cancers are fatal and incurable.
> WP5: This WP relates to lung cancer comorbidity, and to maximising the clinical information gathered during a single computerised tomography (CT) exam study of the chest, to allow for earlier treatment of patients who have underlying, unrecognised diseases or conditions such as Chronic Obstructive Pulmonary Disease (COPD) and Coronary Artery Disease (CAD). The aim of WP5 is to automate the identification of comorbid states (COPD and CAD) related to or caused by lung cancer, by developing and implementing machine learning models. The aim is to utilise all standard of care data captured during the lung cancer screening programme, if available through the normal course of clinical practice, to identify incidental findings that may be picked up in the TLHC patient cohort through the screening programme.
> WP6: The aim of half of WP6 is to assess whether the addition of an AI component to the Lung Cancer Screening Programme (LCSP) in England is cost-effective, compared to current practice. This is to be achieved by constructing a decision-analytic model to evaluate resource use, costs and health outcomes for patients in the LCSP over time, with and without the AI component in place. A specific type of model will be built – a state transition model – in which patients move between different lung cancer specific health states periodically over a time horizon of several years. By spending time in these health states, patients accrue costs and health outcomes. The requested HES-TLHC linked Data will be used to determine an appropriate cost for each health state, and the requested linked mortality Data will be used to inform estimates of survival outcomes in each health state. To achieve the former, the HES-TLHC linked Data will be combined with NHS Reference Costs to calculate the secondary care costs accrued by patients in specific health states.
> WP8: The aim of WP8 is to improve and enhance the existing evidence base on outcome prediction for patients with lung cancer. The aim is to identify predictors of unexpected (usually poor) outcome. This will require a wide spectrum of data that might help predict, for example, greater mortality, lower recurrence-free survival and overall survival than expected having accounted for all of the known factors that influence these (such as age, performance status and stage). This will be achieved by using the requested Data linked to TLHC data to identify patients with lung cancer who relapse or die. Analyses will be undertaken using this data to confirm the efficacy of current clinical approaches and suggest standardised approaches that may reduce unwarranted variation. The results of these analyses will also be used to challenge existing practice that does not meet the outcomes achieved in better services and suggest new paradigms of care by identifying novel markers of unexpected outcomes. WP6 and WP8 are necessarily linked: the model that will be constructed in WP6 will be adaptable such that the cost-effectiveness of standardised and new clinical approaches can be tested.
The other work packages will not use the NHS England data provided under this Agreement.
The following NHS England Data will be accessed:
> Hospital Episode Statistics; necessary for the analysis of the algorithms developed in WP5 to determine if they can detect and quantify clinically meaningful COPD and CAD from the Lung Cancer Screening CT scans, and the requested HES-TLHC linked Data will be used to determine if the AI algorithms developed relate to clinically important episodes. For WP6, HES Data is required for the cost-effectiveness analyses and the requested HES-TLHC linked Data will be used to determine an appropriate cost for each health state. For WP8, the HES data is required for the analysis of the algorithms developed in this work package to determine if they can determine whether and why, patients from the TLHC Lung Cancer Screening CT scans die from lung cancer and are not cured despite having early-stage disease. Specifically, the following HES Data is required:
- Admitted Patient Care - necessary to determine how often patients from the TLHC are admitted to hospital and whether it is due to COPD, CAD or lung cancer.
- Accident & Emergency - necessary to determine how often patients from the TLHC are seen in Emergency Departments, and whether it is due to COPD, CAD or lung cancer and to estimate costs associated with lung cancer treatment.
- Critical Care - necessary to determine how often patients from the TLHC are admitted to ICUs and whether it is due to COPD, CAD or lung cancer and to estimate costs associated with lung cancer treatment.
- Outpatients - necessary to determine how often patients from the TLHC are seen in OPs and whether it is due to COPD, CAD or lung cancer and to estimate costs associated with lung cancer treatment.
> Emergency Care Data Set (ECDS) – necessary to determine how often patients from the TLHC are seen in Emergency Departments, and whether it is due to COPD, CAD or lung cancer and to estimate costs associated with lung cancer treatment.
> Civil Registration Mortality - necessary to determine whether patients from the TLHC centres have died from COPD, CAD or lung cancer and because mortality Data will be used to inform estimates of survival outcomes associated with health states.
> Cancer Registration - necessary to determine who in the cohort has / has had cancer and who has died from it.
The level of the data will be:
> Identifiable – NHS number is required to link the NHS England Data to the TLHC data.
The data will be minimised as follows:
> Limited to a study cohort identified by Oxford University Hospitals NHS Foundation Trust (OUH) – approximately 300,000 participants of the TLHC programme
> Limited to data for episodes after the start date of the TLHC programme for each site (TLHC specific; not defined for each cohort member). Cancer and deaths data will not be filtered in this way.
> Data access is minimised according to the research purpose within the TRE. This means that researchers within each distinct work package will have access only to the Data that they need.
The University of Oxford is the sponsor and controller as the organisation responsible for ensuring that the Data will only be processed for the purpose described above.
The lawful basis for processing personal data under the UK GDPR is:
> Article 6(1)(e) - processing is necessary for the performance of a task carried out in the public interest or in the exercise of official authority vested in the controller.
The lawful basis for processing special category data under the UK GDPR is:
> 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.
Early detection and treatment of asymptomatic patients is crucial. Detecting lung cancer early when it is small and seen on a CT scan as a small nodule is currently recognised as the best way to do this. The ambition the DART study is to see if some of the aspects of lung cancer screening can be improved, which is expected to provide public healthcare benefits.
The funding is provided by Innovate UK, with in kind funding (non-monetary) from GE Healthcare (GEHC) and Roche Diagnostics Ltd. The funding is specifically for the component of the project described. Funding is in place until the end of March 2024.
OUH, GEHC, Optellum and Nottingham University Hospital NHS Foundation Trust (NUH) are processors acting under the instructions of the University of Oxford.
Microsoft Limited provides IT hosting services to OUH and will store the Data as contracted by OUH.
A Patient and Public Involvement and Engagement (PPIE) group assisted with the grant application and another PPIE group assisted with the programme design after the grant was received and has an on-going role in advising the project through quarterly meetings and ad hoc email/video calls as needed.
GEHC are a commercial organisation and a processor under this Data Sharing Agreement for the purpose described. They will be using the data obtained under this agreement, in combination with the data DART is obtaining from the Targeted Lung Health Checks, to develop and improve artificial intelligence algorithms to better predict respiratory conditions such as cardiac diseases and COPD.
GEHC make a considerable (over 1 million pounds) kind contribution to the project and, as a commercial company, would naturally expect a commercial gain and the publicity associated with participating in a prestigious project which has attracted Government and national media attention. The expected health and social care benefits of their contribution into the project outweigh any commercial gain they would receive from the project.
Alongside the University of Oxford academic team, Roche Diagnostics Ltd make an in-kind contribution of their expertise into WP4 and have loaned a Digital Pathology Scanner to DART (in Oxford) and provided a staff member to operate this. As a funder and commercial organisation, Roche Diagnostics reasonably expect a commercial return and to receive acknowledgement and publicity for their participation in the project. In turn, their contribution to the earlier detection of lung cancer far outweighs any likely commercial return.
Processing activities
OUH will transfer data to NHS England. The data will consist of identifying details (specifically NHS Number, Gender, Date of Birth, and a DART Study ID) for the cohort to be linked with NHS England Data.
NHS England will provide the relevant records from the HES APC, HES CC, HES OP, HES A&E, ECDS, Civil Registrations of Death, and Caner Registration Data datasets to OUH. The Data will:
> Contain directly identifying data items including NHS number which are required to link the NHS England Data to the lung health check data.
The Data will not be transferred to any other location.
The Data will be stored on servers at OUH. OUH also stores data on the TRE which runs on the Azure Cloud provided by Microsoft Limited.
The Data will be accessed onsite at the premises of OUH.
The record-level Data will also be accessed on the DART TRE by authorised personnel via remote access. The Data will remain on the servers at OUH and Microsoft Limited at all times.
The Controller(s) must confirm and provide evidence upon audit by NHS England that access via any remote device complies with the data security obligations within this DSA and the Data Sharing Framework Contract.
For remote access:
- Remote access will only be from secure locations situated within the territory of use (as further restricted elsewhere within the DSA if so done) stated within this DSA;
- Personnel are both prohibited and technically prevented from downloading or copying NHSE data to local devices;
- Access controls granting users the minimum level of access required are in place;
- Remote access is only via secure connections (e.g., VPNs or secure protocols) to protect data;
- Multifactor authentication (MFA) is required for remote access;
- Device security, including up-to-date software and operating systems, antivirus software, and enabled firewalls are utilised for the remote access;
- All remote access is undertaken within the scope of the organisation’s DSPT (or other security arrangements as per this agreement) and complies with the organisation’s remote access policy.
The above applies in addition to any condition set out elsewhere within the DSA (e.g. who may carry out processing, and for what purpose).
Only aggregated data with small numbers suppressed can be exported from the TRE, via a secure Airlock Policy.
Remote processing will be from secure locations within England/Wales. The data will not leave England/Wales at any time.
All personnel accessing the data have been appropriately trained in data protection and confidentiality.
The Data will be linked at person record level with data from the TLHC programme. The Data will not be linked with any other data.
The identifying details will be stored in a separate database to the linked dataset used for analysis. All analyses will use the pseudonymised dataset. There will be no requirement and no attempt to reidentify individuals when using the pseudonymised dataset.
Analysts from the OUH, NUH and University of Oxford will analyse the data for the purposes described above.
Optellum and GEHC users will access the data for the purpose of developing AI tools to achieve the purpose described above.
Expected output
The expected outputs of the processing will be:
> A summary of updated findings to relevant stakeholders (Innovate UK, CRUK, Roche, Optellum, GEHC): Expected March – May 2027 and 2028
> Submissions to peer reviewed high impact journals: Specific target journals will be identified as the work progresses to ensure the study findings are communicated to appropriate audiences. Expected January – June 2027 and 2028
> Presentations will also be given at national and international conferences relevant to lung cancer and screening and artificial intelligence.
The outputs will not contain patient level NHS England Data and will only contain aggregated information with small numbers suppressed as appropriate in line with the relevant disclosure rules for the dataset(s) from which the information was derived.
The outputs will be communicated to relevant recipients through the following dissemination channels:
> Journals
> The results of the planned analyses will be communicated directly to the National Screening Committee (NSC) upon the completion of DART in March 2028 by collaborators who also have a position on this committee
> Results will be shared with the staff at the TLHC centres and other operational staff involved in lung cancer screening
> DART Patient and Public Involvement and Engagement (PPIE) contributors, including the Roy Castle Lung Cancer Foundation, will be consulted to communicate findings to the general public and patients in an accessible way
> National and International conferences
Outputs are expected throughout 2026, 2027 and 2028.
Expected measurable benefits
The requested Data will be used to undertake analyses that are expected to reveal:
• The cost-effectiveness of adding an AI component to the Lung Cancer Screening Programme in England, compared to current practice.
• The efficacy of (a) current clinical approaches to outcome prediction; (b) plausible standardised approaches to outcome prediction, and (c) potential new paradigms of care by identifying novel markers of outcomes.
It is expected that these findings will have the following benefits for patients in the English NHS, and the health sector overall:
• Lung cancer could be more accurately diagnosed, with enhanced prognostic information.
• The time to diagnosis for lung cancer could be reduced.
• The frequency of use of harmful invasive procedures in the diagnostic pathway could be reduced.
• Selection of patients for lung cancer screening could be improved, reducing the overall cost allocation for this.
• The assessment of risks from comorbidities could be improved.
The results of the planned analyses will be communicated directly to the National Screening Committee (NSC) by collaborators who also have a position on this committee, thus ensuring that the findings directly influence and optimise the implementation of the Lung Cancer Screening Programme.
Results will be communicated to the NSC upon the completion of DART in March 2027, and are therefore expected to begin to translate into direct patient benefit quickly. The British Thoracic Society (BTS) guidelines are currently under review and we intend for our findings to contribute to that national update. In parallel, the results will feed into the European Society of Thoracic Imaging’s forthcoming lung cancer screening guidelines, enabling DART to contribute meaningfully to high‑impact international standards that shape clinical practice across Europe.
With around 48,000 people are diagnosed with lung cancer every year in the UK, the potential improvements in early detection, patient outcomes and cost effectiveness could be significant.
Benefits reported so far
Early detection and treatment of asymptomatic patients is crucial. Detecting lung cancer early when it is small
and seen on a CT scan as a small nodule is currently recognised as the best way to do this. The ambition the
DART study is to see if some of the aspects of lung cancer screening can be improved, which is expected to
provide public healthcare benefits.
The requested Data will be used to undertake analyses that are expected to reveal:
• The cost-effectiveness of adding an AI component to the Lung Cancer Screening Programme in
England, compared to current practice.
• The efficacy of (a) current clinical approaches to outcome prediction; (b) plausible standardised
approaches to outcome prediction, and (c) potential new paradigms of care by identifying novel markers
of outcomes.
It is expected that these findings will have the following benefits for patients in the English NHS, and the
health sector overall:
• Lung cancer could be more accurately diagnosed, with enhanced prognostic information.
• The time to diagnosis for lung cancer could be reduced.
• The frequency of use of harmful invasive procedures in the diagnostic pathway could be reduced.
• Selection of patients for lung cancer screening could be improved, reducing the overall cost allocation
for this.
• The assessment of risks from comorbidities could be improved.
The results of the planned analyses will be communicated directly to the National Screening Committee
(NSC) by collaborators who also have a position on this committee, thus ensuring that the findings directly
influence and optimise the implementation of the Lung Cancer Screening Programme.
Results will be communicated to the NSC upon the completion of DART in March 2027, and are therefore expected to begin to translate into direct patient benefit quickly. The British Thoracic Society (BTS) guidelines are currently under review and we intend for DART's findings to contribute to that national update. In parallel, the results will feed into the European Society of Thoracic Imaging’s forthcoming lung cancer screening guidelines, enabling DART to contribute meaningfully to high‑impact international standards that shape clinical practice across Europe. With around 48,000 people are diagnosed with lung cancer every year in the UK, the potential improvements in early detection, patient outcomes and cost effectiveness could be significant.
Datasets on the current version
Legal basis for provision: Health and Social Care Act 2012 - s261(5)(d); National Health Service Act 2006 - s251 - 'Control of patient information'.
| Dataset | Type of data | Sensitivity | Frequency | Confidential data |
|---|---|---|---|---|
| Cancer Registration Data | Identifiable | Non-Sensitive | One-Off | Section 251 NHS Act 2006 |
| Civil Registrations of Death | Identifiable | Sensitive | One-Off | Section 251 NHS Act 2006 |
| Emergency Care Data Set (ECDS) | Identifiable | Non-Sensitive | One-Off | Section 251 NHS Act 2006 |
| Hospital Episode Statistics Accident and Emergency (HES A and E) | Identifiable | Non-Sensitive | One-Off | Section 251 NHS Act 2006 |
| Hospital Episode Statistics Admitted Patient Care (HES APC) | Identifiable | Non-Sensitive | One-Off | Section 251 NHS Act 2006 |
| Hospital Episode Statistics Critical Care (HES Critical Care) | Identifiable | Non-Sensitive | One-Off | Section 251 NHS Act 2006 |
| Hospital Episode Statistics Outpatients (HES OP) | Identifiable | Non-Sensitive | One-Off | Section 251 NHS Act 2006 |
Files released
Files released counts only files released externally by DARS. Access granted in NHS England's own systems, such as its Secure Data Environment, is not included.
Patient opt-outs were applied to all 28 files released under this agreement, across every version. About opt-outs
No files recorded as released under the current version. 28 were released under earlier versions, shown in the version history.
Version history
The register lists each renewal of this agreement as a separate row. This site has 3 versions.
DARS-NIC-668928-L9N5F-v2.11 20 March 2026 to 19 March 2027
- Title
- DART: The Integration and Analysis of Data using Artificial Intelligence to Improve Patient Outcomes with Thoracic Diseases.
- Commercial
- Yes
- Sublicensing
- No
- Datasets
- 7
- Files released
- 0
Datasets: Cancer Registration Data; Civil Registrations of Death; Emergency Care Data Set (ECDS); Hospital Episode Statistics Accident and Emergency (HES A and E); Hospital Episode Statistics Admitted Patient Care (HES APC); Hospital Episode Statistics Critical Care (HES Critical Care); Hospital Episode Statistics Outpatients (HES OP)
What changed from DARS-NIC-668928-L9N5F-v1.3
Text removed is struck through; text added is underlined. Unchanged paragraphs are summarised rather than repeated.
| Field | Was | Became |
|---|---|---|
| Title | DART: The Integration and Analysis of Data using Artificial Intelligence to Improve Patient Outcomes with Thoracic Diseases. | |
| Start date | 2026-03-20 | |
| End date | 2027-03-19 |
Expected output
[1 paragraph unchanged]
> A report of findings to funding bodies (Innovate (part of UK Research and Innovation); Cancer Research UK (CRUK)): expected March 2024
> A summary of updated findings to relevant stakeholders (Innovate UK, CRUK, Roche, Optellum, GEHC): Expected March – May 2027 and 2028
> A summary of findings to relevant stakeholders (Innovate UK, CRUK, Roche, Optellum, GEHC): Expected March – May 2024
> Submissions to peer reviewed high impact journals: Specific target journals will be identified as the work progresses to ensure the study findings are communicated to appropriate audiences. Expected January – June 2027 and 2028
> Submissions to peer reviewed open-access journals: Specific target journals will be identified as the work progresses to ensure the study findings are communicated to appropriate audiences. Expected January – June 2024
> Presentations will also be given at national and international conferences relevant to lung cancer and screening and artificial intelligence.
> Presentations will also be given at national and international conferences relevant to lung cancer and screening
The outputs will not contain patient level NHS England Data and will only contain aggregated information with small numbers suppressed as appropriate in line with the relevant disclosure rules for the dataset(s) from which the information was derived.
The outputs will not contain NHS England Data and will only contain aggregated information with small numbers suppressed as appropriate in line with the relevant disclosure rules for the dataset(s) from which the information was derived.
[2 paragraphs unchanged]
> The results of the planned analyses will be communicated directly to the National Screening Committee (NSC) upon the completion of DART in March
2024
2028
by collaborators who also have a position on this committee
[2 paragraphs unchanged]
> Results will be communicated to the National Screening Committee upon the completion of DART in March 2024
[1 paragraph unchanged]
Outputs are expected throughout
2024.
2026, 2027 and 2028.
Expected measurable benefits
[9 paragraphs unchanged]
The results of the planned analyses will be communicated directly to the
[19 words unchanged]
directly influence and optimise the implementation of the Lung Cancer Screening Programme.
Results will be communicated to the NSC upon the completion of DART in March 2024, and are thus expected to begin to translate into direct patient benefit quickly. Around 48,000 people are diagnosed with lung cancer every year in the UK, so the potential health benefits and cost impacts could be significant.
Results will be communicated to the NSC upon the completion of DART in March 2027, and are therefore expected to begin to translate into direct patient benefit quickly. The British Thoracic Society (BTS) guidelines are currently under review and we intend for our findings to contribute to that national update. In parallel, the results will feed into the European Society of Thoracic Imaging’s forthcoming lung cancer screening guidelines, enabling DART to contribute meaningfully to high‑impact international standards that shape clinical practice across Europe.
With around 48,000 people are diagnosed with lung cancer every year in the UK, the potential improvements in early detection, patient outcomes and cost effectiveness could be significant.
Benefits reported
[20 paragraphs unchanged]
influence and optimise the implementation of the Lung Cancer Screening Programme.
Results will be
communicated to the NSC upon the completion of DART in March 2024, and are thus expected to begin
Results will be communicated to the NSC upon the completion of DART in March 2027, and are therefore expected to begin to translate into direct patient benefit quickly. The British Thoracic Society (BTS) guidelines are currently under review and we intend for DART's findings to contribute to that national update. In parallel, the results will feed into the European Society of Thoracic Imaging’s forthcoming lung cancer screening guidelines, enabling DART to contribute meaningfully to high‑impact international standards that shape clinical practice across Europe. With around 48,000 people are diagnosed with lung cancer every year in the UK, the potential improvements in early detection, patient outcomes and cost effectiveness could be significant.
to translate into direct patient benefit quickly. Around 48,000 people are diagnosed with lung cancer every
year in the UK, so the potential health benefits and cost impacts could be significant.
Changed only in punctuation, spacing or capitalisation: Processing activities.
Unchanged: Objective for processing.
DARS-NIC-668928-L9N5F-v1.3 20 March 2024 to 19 March 2026
- Title
- DART: The Integration and Analysis of Data using Artificial Intelligence to Improve Patient Outcomes with Thoracic Diseases
- Commercial
- Yes
- Sublicensing
- No
- Datasets
- 7
- Files released
- 26
Datasets: Cancer Registration Data; Civil Registrations of Death; Emergency Care Data Set (ECDS); Hospital Episode Statistics Accident and Emergency (HES A and E); Hospital Episode Statistics Admitted Patient Care (HES APC); Hospital Episode Statistics Critical Care (HES Critical Care); Hospital Episode Statistics Outpatients (HES OP)
What changed from DARS-NIC-668928-L9N5F-v0.6
Text removed is struck through; text added is underlined. Unchanged paragraphs are summarised rather than repeated.
| Field | Was | Became |
|---|---|---|
| Start date | 2024-03-20 | |
| End date | 2026-03-19 |
Objective for processing
[35 paragraphs unchanged]
> Limited to data for episodes after the start date
(individual
of the TLHC programme for each site (TLHC
specific;
this will be the date the
not defined for each
cohort
member attended their screening).
member).
Cancer and deaths data will not be filtered in this way.
[14 paragraphs unchanged]
Benefits reported
Yielded Benefits is not a requirement for new applications.
Early detection and treatment of asymptomatic patients is crucial. Detecting lung cancer early when it is small
and seen on a CT scan as a small nodule is currently recognised as the best way to do this. The ambition the
DART study is to see if some of the aspects of lung cancer screening can be improved, which is expected to
provide public healthcare benefits.
The requested Data will be used to undertake analyses that are expected to reveal:
• The cost-effectiveness of adding an AI component to the Lung Cancer Screening Programme in
England, compared to current practice.
• The efficacy of (a) current clinical approaches to outcome prediction; (b) plausible standardised
approaches to outcome prediction, and (c) potential new paradigms of care by identifying novel markers
of outcomes.
It is expected that these findings will have the following benefits for patients in the English NHS, and the
health sector overall:
• Lung cancer could be more accurately diagnosed, with enhanced prognostic information.
• The time to diagnosis for lung cancer could be reduced.
• The frequency of use of harmful invasive procedures in the diagnostic pathway could be reduced.
• Selection of patients for lung cancer screening could be improved, reducing the overall cost allocation
for this.
• The assessment of risks from comorbidities could be improved.
The results of the planned analyses will be communicated directly to the National Screening Committee
(NSC) by collaborators who also have a position on this committee, thus ensuring that the findings directly
influence and optimise the implementation of the Lung Cancer Screening Programme. Results will be
communicated to the NSC upon the completion of DART in March 2024, and are thus expected to begin
to translate into direct patient benefit quickly. Around 48,000 people are diagnosed with lung cancer every
year in the UK, so the potential health benefits and cost impacts could be significant.
Unchanged: Processing activities, Expected output, Expected measurable benefits.
Objective for processing
The University of Oxford requires access to NHS England Data for the purpose of the following research project:
DART (The Integration and Analysis of Data using Artificial Intelligence to Improve Patient Outcomes with Thoracic Diseases)
The following is a summary of the aims of the research project provided by University of Oxford:
DART is a study developing integrated diagnostics to enable the earlier diagnosis of lung cancer. This study, which began in 2020, aims to improve patient survival, yielding time and cost efficiencies to the NHS. The overall objective of the study is to link data from the Targeted Lung Health Checks (TLHC) programme, collected by the Lung Health Centres and sent to Oxford University Hospitals NHS Foundation Trust (OUH), with NHS England Data to better predict those who will benefit from lung cancer screening and to develop artificial intelligence systems that could improve understanding (through machine learning) of co-morbidities and confirm or improve standard clinical practices. Costs and survival will be summarised for groups of patients for various health states. It is anticipated that limited data will be received on secondary care resource use via the TLHC Programme. Based on initial inspections, this data will be fragmented, will cover a more limited follow-up period, and will not be available in detail for all participants in the TLHC programme. The Data requested from NHS England under this Agreement is expected to considerably enhance the study by providing less biased data and ensuring that missing secondary care resource use data is minimised.
Another objective of DART is to improve and enhance the existing evidence base on outcome prediction for patients with lung cancer. The requested Data will be used to identify patients with lung cancer who relapse or die; analyses using these Data will assess the efficacy of current and proposed clinical approaches.
OUH has been granted and funded for a Trusted Research Environment (TRE). The data from DART will be stored in this TRE, which will be managed in accordance with the relevant approvals from the Confidentiality Advisory Group (CAG), health Research Authority (HRA), and the Research and Development Department of the OUH. Access to the TRE for researchers will be governed by the above regulations and will require detailed applications and approval by the TRE Data Access Committee (TRE DAC). The NHS England data requested for DART will be stored in the DART section of the TRE.
DART has a total of nine work packages
• WP1 Leadership and Project Management
• WP2 Planning for, and management of Datasets, and Integration
• WP3 AI Model Validation on Lung Cancer Screening (LCS) data
• WP4 Digital Pathology AI and Radiomics Model Development and Validation
• WP5 Chronic obstructive pulmonary disease (COPD)/ Coronary artery disease (CAD) Model Development and Validation
• WP6 Primary Care/Population Health and Health Economics
• WP7 Integration of Blood Biomarkers
• WP8 Outcome prediction and new treatment paradigms
• WP9 Quality assurance: radiologist training and monitoring
The requested Data will be used in multiple DART work packages (WPs 4, 5, 6, 8) to further its overall aims:
> WP4: This WP relates to development of an artificial intelligence algorithm to diagnose lung cancer from digital pathology images. The HES data will enable ground truthing (a Gold Standard that can be used to compare and evaluate model results) of cancer diagnoses and for the algorithm to learn which of the cancers are fatal and incurable.
> WP5: This WP relates to lung cancer comorbidity, and to maximising the clinical information gathered during a single computerised tomography (CT) exam study of the chest, to allow for earlier treatment of patients who have underlying, unrecognised diseases or conditions such as Chronic Obstructive Pulmonary Disease (COPD) and Coronary Artery Disease (CAD). The aim of WP5 is to automate the identification of comorbid states (COPD and CAD) related to or caused by lung cancer, by developing and implementing machine learning models. The aim is to utilise all standard of care data captured during the lung cancer screening programme, if available through the normal course of clinical practice, to identify incidental findings that may be picked up in the TLHC patient cohort through the screening programme.
> WP6: The aim of half of WP6 is to assess whether the addition of an AI component to the Lung Cancer Screening Programme (LCSP) in England is cost-effective, compared to current practice. This is to be achieved by constructing a decision-analytic model to evaluate resource use, costs and health outcomes for patients in the LCSP over time, with and without the AI component in place. A specific type of model will be built – a state transition model – in which patients move between different lung cancer specific health states periodically over a time horizon of several years. By spending time in these health states, patients accrue costs and health outcomes. The requested HES-TLHC linked Data will be used to determine an appropriate cost for each health state, and the requested linked mortality Data will be used to inform estimates of survival outcomes in each health state. To achieve the former, the HES-TLHC linked Data will be combined with NHS Reference Costs to calculate the secondary care costs accrued by patients in specific health states.
> WP8: The aim of WP8 is to improve and enhance the existing evidence base on outcome prediction for patients with lung cancer. The aim is to identify predictors of unexpected (usually poor) outcome. This will require a wide spectrum of data that might help predict, for example, greater mortality, lower recurrence-free survival and overall survival than expected having accounted for all of the known factors that influence these (such as age, performance status and stage). This will be achieved by using the requested Data linked to TLHC data to identify patients with lung cancer who relapse or die. Analyses will be undertaken using this data to confirm the efficacy of current clinical approaches and suggest standardised approaches that may reduce unwarranted variation. The results of these analyses will also be used to challenge existing practice that does not meet the outcomes achieved in better services and suggest new paradigms of care by identifying novel markers of unexpected outcomes. WP6 and WP8 are necessarily linked: the model that will be constructed in WP6 will be adaptable such that the cost-effectiveness of standardised and new clinical approaches can be tested.
The other work packages will not use the NHS England data provided under this Agreement.
The following NHS England Data will be accessed:
> Hospital Episode Statistics; necessary for the analysis of the algorithms developed in WP5 to determine if they can detect and quantify clinically meaningful COPD and CAD from the Lung Cancer Screening CT scans, and the requested HES-TLHC linked Data will be used to determine if the AI algorithms developed relate to clinically important episodes. For WP6, HES Data is required for the cost-effectiveness analyses and the requested HES-TLHC linked Data will be used to determine an appropriate cost for each health state. For WP8, the HES data is required for the analysis of the algorithms developed in this work package to determine if they can determine whether and why, patients from the TLHC Lung Cancer Screening CT scans die from lung cancer and are not cured despite having early-stage disease. Specifically, the following HES Data is required:
- Admitted Patient Care - necessary to determine how often patients from the TLHC are admitted to hospital and whether it is due to COPD, CAD or lung cancer.
- Accident & Emergency - necessary to determine how often patients from the TLHC are seen in Emergency Departments, and whether it is due to COPD, CAD or lung cancer and to estimate costs associated with lung cancer treatment.
- Critical Care - necessary to determine how often patients from the TLHC are admitted to ICUs and whether it is due to COPD, CAD or lung cancer and to estimate costs associated with lung cancer treatment.
- Outpatients - necessary to determine how often patients from the TLHC are seen in OPs and whether it is due to COPD, CAD or lung cancer and to estimate costs associated with lung cancer treatment.
> Emergency Care Data Set (ECDS) – necessary to determine how often patients from the TLHC are seen in Emergency Departments, and whether it is due to COPD, CAD or lung cancer and to estimate costs associated with lung cancer treatment.
> Civil Registration Mortality - necessary to determine whether patients from the TLHC centres have died from COPD, CAD or lung cancer and because mortality Data will be used to inform estimates of survival outcomes associated with health states.
> Cancer Registration - necessary to determine who in the cohort has / has had cancer and who has died from it.
The level of the data will be:
> Identifiable – NHS number is required to link the NHS England Data to the TLHC data.
The data will be minimised as follows:
> Limited to a study cohort identified by Oxford University Hospitals NHS Foundation Trust (OUH) – approximately 300,000 participants of the TLHC programme
> Limited to data for episodes after the start date of the TLHC programme for each site (TLHC specific; not defined for each cohort member). Cancer and deaths data will not be filtered in this way.
> Data access is minimised according to the research purpose within the TRE. This means that researchers within each distinct work package will have access only to the Data that they need.
The University of Oxford is the sponsor and controller as the organisation responsible for ensuring that the Data will only be processed for the purpose described above.
The lawful basis for processing personal data under the UK GDPR is:
> Article 6(1)(e) - processing is necessary for the performance of a task carried out in the public interest or in the exercise of official authority vested in the controller.
The lawful basis for processing special category data under the UK GDPR is:
> 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.
Early detection and treatment of asymptomatic patients is crucial. Detecting lung cancer early when it is small and seen on a CT scan as a small nodule is currently recognised as the best way to do this. The ambition the DART study is to see if some of the aspects of lung cancer screening can be improved, which is expected to provide public healthcare benefits.
The funding is provided by Innovate UK, with in kind funding (non-monetary) from GE Healthcare (GEHC) and Roche Diagnostics Ltd. The funding is specifically for the component of the project described. Funding is in place until the end of March 2024.
OUH, GEHC, Optellum and Nottingham University Hospital NHS Foundation Trust (NUH) are processors acting under the instructions of the University of Oxford.
Microsoft Limited provides IT hosting services to OUH and will store the Data as contracted by OUH.
A Patient and Public Involvement and Engagement (PPIE) group assisted with the grant application and another PPIE group assisted with the programme design after the grant was received and has an on-going role in advising the project through quarterly meetings and ad hoc email/video calls as needed.
GEHC are a commercial organisation and a processor under this Data Sharing Agreement for the purpose described. They will be using the data obtained under this agreement, in combination with the data DART is obtaining from the Targeted Lung Health Checks, to develop and improve artificial intelligence algorithms to better predict respiratory conditions such as cardiac diseases and COPD.
GEHC make a considerable (over 1 million pounds) kind contribution to the project and, as a commercial company, would naturally expect a commercial gain and the publicity associated with participating in a prestigious project which has attracted Government and national media attention. The expected health and social care benefits of their contribution into the project outweigh any commercial gain they would receive from the project.
Alongside the University of Oxford academic team, Roche Diagnostics Ltd make an in-kind contribution of their expertise into WP4 and have loaned a Digital Pathology Scanner to DART (in Oxford) and provided a staff member to operate this. As a funder and commercial organisation, Roche Diagnostics reasonably expect a commercial return and to receive acknowledgement and publicity for their participation in the project. In turn, their contribution to the earlier detection of lung cancer far outweighs any likely commercial return.
Expected output
The expected outputs of the processing will be:
> A report of findings to funding bodies (Innovate (part of UK Research and Innovation); Cancer Research UK (CRUK)): expected March 2024
> A summary of findings to relevant stakeholders (Innovate UK, CRUK, Roche, Optellum, GEHC): Expected March – May 2024
> Submissions to peer reviewed open-access journals: Specific target journals will be identified as the work progresses to ensure the study findings are communicated to appropriate audiences. Expected January – June 2024
> Presentations will also be given at national and international conferences relevant to lung cancer and screening
The outputs will not contain NHS England Data and will only contain aggregated information with small numbers suppressed as appropriate in line with the relevant disclosure rules for the dataset(s) from which the information was derived.
The outputs will be communicated to relevant recipients through the following dissemination channels:
> Journals
> The results of the planned analyses will be communicated directly to the National Screening Committee (NSC) upon the completion of DART in March 2024 by collaborators who also have a position on this committee
> Results will be shared with the staff at the TLHC centres and other operational staff involved in lung cancer screening
> DART Patient and Public Involvement and Engagement (PPIE) contributors, including the Roy Castle Lung Cancer Foundation, will be consulted to communicate findings to the general public and patients in an accessible way
> Results will be communicated to the National Screening Committee upon the completion of DART in March 2024
> National and International conferences
Outputs are expected throughout 2024.
Benefits reported
Early detection and treatment of asymptomatic patients is crucial. Detecting lung cancer early when it is small
and seen on a CT scan as a small nodule is currently recognised as the best way to do this. The ambition the
DART study is to see if some of the aspects of lung cancer screening can be improved, which is expected to
provide public healthcare benefits.
The requested Data will be used to undertake analyses that are expected to reveal:
• The cost-effectiveness of adding an AI component to the Lung Cancer Screening Programme in
England, compared to current practice.
• The efficacy of (a) current clinical approaches to outcome prediction; (b) plausible standardised
approaches to outcome prediction, and (c) potential new paradigms of care by identifying novel markers
of outcomes.
It is expected that these findings will have the following benefits for patients in the English NHS, and the
health sector overall:
• Lung cancer could be more accurately diagnosed, with enhanced prognostic information.
• The time to diagnosis for lung cancer could be reduced.
• The frequency of use of harmful invasive procedures in the diagnostic pathway could be reduced.
• Selection of patients for lung cancer screening could be improved, reducing the overall cost allocation
for this.
• The assessment of risks from comorbidities could be improved.
The results of the planned analyses will be communicated directly to the National Screening Committee
(NSC) by collaborators who also have a position on this committee, thus ensuring that the findings directly
influence and optimise the implementation of the Lung Cancer Screening Programme. Results will be
communicated to the NSC upon the completion of DART in March 2024, and are thus expected to begin
to translate into direct patient benefit quickly. Around 48,000 people are diagnosed with lung cancer every
year in the UK, so the potential health benefits and cost impacts could be significant.
DARS-NIC-668928-L9N5F-v0.6 26 October 2023 to 25 October 2024
- Title
- DART: The Integration and Analysis of Data using Artificial Intelligence to Improve Patient Outcomes with Thoracic Diseases
- Commercial
- Yes
- Sublicensing
- No
- Datasets
- 7
- Files released
- 2
Datasets: Cancer Registration Data; Civil Registrations of Death; Emergency Care Data Set (ECDS); Hospital Episode Statistics Accident and Emergency (HES A and E); Hospital Episode Statistics Admitted Patient Care (HES APC); Hospital Episode Statistics Critical Care (HES Critical Care); Hospital Episode Statistics Outpatients (HES OP)
Objective for processing
The University of Oxford requires access to NHS England Data for the purpose of the following research project:
DART (The Integration and Analysis of Data using Artificial Intelligence to Improve Patient Outcomes with Thoracic Diseases)
The following is a summary of the aims of the research project provided by University of Oxford:
DART is a study developing integrated diagnostics to enable the earlier diagnosis of lung cancer. This study, which began in 2020, aims to improve patient survival, yielding time and cost efficiencies to the NHS. The overall objective of the study is to link data from the Targeted Lung Health Checks (TLHC) programme, collected by the Lung Health Centres and sent to Oxford University Hospitals NHS Foundation Trust (OUH), with NHS England Data to better predict those who will benefit from lung cancer screening and to develop artificial intelligence systems that could improve understanding (through machine learning) of co-morbidities and confirm or improve standard clinical practices. Costs and survival will be summarised for groups of patients for various health states. It is anticipated that limited data will be received on secondary care resource use via the TLHC Programme. Based on initial inspections, this data will be fragmented, will cover a more limited follow-up period, and will not be available in detail for all participants in the TLHC programme. The Data requested from NHS England under this Agreement is expected to considerably enhance the study by providing less biased data and ensuring that missing secondary care resource use data is minimised.
Another objective of DART is to improve and enhance the existing evidence base on outcome prediction for patients with lung cancer. The requested Data will be used to identify patients with lung cancer who relapse or die; analyses using these Data will assess the efficacy of current and proposed clinical approaches.
OUH has been granted and funded for a Trusted Research Environment (TRE). The data from DART will be stored in this TRE, which will be managed in accordance with the relevant approvals from the Confidentiality Advisory Group (CAG), health Research Authority (HRA), and the Research and Development Department of the OUH. Access to the TRE for researchers will be governed by the above regulations and will require detailed applications and approval by the TRE Data Access Committee (TRE DAC). The NHS England data requested for DART will be stored in the DART section of the TRE.
DART has a total of nine work packages
• WP1 Leadership and Project Management
• WP2 Planning for, and management of Datasets, and Integration
• WP3 AI Model Validation on Lung Cancer Screening (LCS) data
• WP4 Digital Pathology AI and Radiomics Model Development and Validation
• WP5 Chronic obstructive pulmonary disease (COPD)/ Coronary artery disease (CAD) Model Development and Validation
• WP6 Primary Care/Population Health and Health Economics
• WP7 Integration of Blood Biomarkers
• WP8 Outcome prediction and new treatment paradigms
• WP9 Quality assurance: radiologist training and monitoring
The requested Data will be used in multiple DART work packages (WPs 4, 5, 6, 8) to further its overall aims:
> WP4: This WP relates to development of an artificial intelligence algorithm to diagnose lung cancer from digital pathology images. The HES data will enable ground truthing (a Gold Standard that can be used to compare and evaluate model results) of cancer diagnoses and for the algorithm to learn which of the cancers are fatal and incurable.
> WP5: This WP relates to lung cancer comorbidity, and to maximising the clinical information gathered during a single computerised tomography (CT) exam study of the chest, to allow for earlier treatment of patients who have underlying, unrecognised diseases or conditions such as Chronic Obstructive Pulmonary Disease (COPD) and Coronary Artery Disease (CAD). The aim of WP5 is to automate the identification of comorbid states (COPD and CAD) related to or caused by lung cancer, by developing and implementing machine learning models. The aim is to utilise all standard of care data captured during the lung cancer screening programme, if available through the normal course of clinical practice, to identify incidental findings that may be picked up in the TLHC patient cohort through the screening programme.
> WP6: The aim of half of WP6 is to assess whether the addition of an AI component to the Lung Cancer Screening Programme (LCSP) in England is cost-effective, compared to current practice. This is to be achieved by constructing a decision-analytic model to evaluate resource use, costs and health outcomes for patients in the LCSP over time, with and without the AI component in place. A specific type of model will be built – a state transition model – in which patients move between different lung cancer specific health states periodically over a time horizon of several years. By spending time in these health states, patients accrue costs and health outcomes. The requested HES-TLHC linked Data will be used to determine an appropriate cost for each health state, and the requested linked mortality Data will be used to inform estimates of survival outcomes in each health state. To achieve the former, the HES-TLHC linked Data will be combined with NHS Reference Costs to calculate the secondary care costs accrued by patients in specific health states.
> WP8: The aim of WP8 is to improve and enhance the existing evidence base on outcome prediction for patients with lung cancer. The aim is to identify predictors of unexpected (usually poor) outcome. This will require a wide spectrum of data that might help predict, for example, greater mortality, lower recurrence-free survival and overall survival than expected having accounted for all of the known factors that influence these (such as age, performance status and stage). This will be achieved by using the requested Data linked to TLHC data to identify patients with lung cancer who relapse or die. Analyses will be undertaken using this data to confirm the efficacy of current clinical approaches and suggest standardised approaches that may reduce unwarranted variation. The results of these analyses will also be used to challenge existing practice that does not meet the outcomes achieved in better services and suggest new paradigms of care by identifying novel markers of unexpected outcomes. WP6 and WP8 are necessarily linked: the model that will be constructed in WP6 will be adaptable such that the cost-effectiveness of standardised and new clinical approaches can be tested.
The other work packages will not use the NHS England data provided under this Agreement.
The following NHS England Data will be accessed:
> Hospital Episode Statistics; necessary for the analysis of the algorithms developed in WP5 to determine if they can detect and quantify clinically meaningful COPD and CAD from the Lung Cancer Screening CT scans, and the requested HES-TLHC linked Data will be used to determine if the AI algorithms developed relate to clinically important episodes. For WP6, HES Data is required for the cost-effectiveness analyses and the requested HES-TLHC linked Data will be used to determine an appropriate cost for each health state. For WP8, the HES data is required for the analysis of the algorithms developed in this work package to determine if they can determine whether and why, patients from the TLHC Lung Cancer Screening CT scans die from lung cancer and are not cured despite having early-stage disease. Specifically, the following HES Data is required:
- Admitted Patient Care - necessary to determine how often patients from the TLHC are admitted to hospital and whether it is due to COPD, CAD or lung cancer.
- Accident & Emergency - necessary to determine how often patients from the TLHC are seen in Emergency Departments, and whether it is due to COPD, CAD or lung cancer and to estimate costs associated with lung cancer treatment.
- Critical Care - necessary to determine how often patients from the TLHC are admitted to ICUs and whether it is due to COPD, CAD or lung cancer and to estimate costs associated with lung cancer treatment.
- Outpatients - necessary to determine how often patients from the TLHC are seen in OPs and whether it is due to COPD, CAD or lung cancer and to estimate costs associated with lung cancer treatment.
> Emergency Care Data Set (ECDS) – necessary to determine how often patients from the TLHC are seen in Emergency Departments, and whether it is due to COPD, CAD or lung cancer and to estimate costs associated with lung cancer treatment.
> Civil Registration Mortality - necessary to determine whether patients from the TLHC centres have died from COPD, CAD or lung cancer and because mortality Data will be used to inform estimates of survival outcomes associated with health states.
> Cancer Registration - necessary to determine who in the cohort has / has had cancer and who has died from it.
The level of the data will be:
> Identifiable – NHS number is required to link the NHS England Data to the TLHC data.
The data will be minimised as follows:
> Limited to a study cohort identified by Oxford University Hospitals NHS Foundation Trust (OUH) – approximately 300,000 participants of the TLHC programme
> Limited to data for episodes after the start date (individual specific; this will be the date the cohort member attended their screening). Cancer and deaths data will not be filtered in this way.
> Data access is minimised according to the research purpose within the TRE. This means that researchers within each distinct work package will have access only to the Data that they need.
The University of Oxford is the sponsor and controller as the organisation responsible for ensuring that the Data will only be processed for the purpose described above.
The lawful basis for processing personal data under the UK GDPR is:
> Article 6(1)(e) - processing is necessary for the performance of a task carried out in the public interest or in the exercise of official authority vested in the controller.
The lawful basis for processing special category data under the UK GDPR is:
> 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.
Early detection and treatment of asymptomatic patients is crucial. Detecting lung cancer early when it is small and seen on a CT scan as a small nodule is currently recognised as the best way to do this. The ambition the DART study is to see if some of the aspects of lung cancer screening can be improved, which is expected to provide public healthcare benefits.
The funding is provided by Innovate UK, with in kind funding (non-monetary) from GE Healthcare (GEHC) and Roche Diagnostics Ltd. The funding is specifically for the component of the project described. Funding is in place until the end of March 2024.
OUH, GEHC, Optellum and Nottingham University Hospital NHS Foundation Trust (NUH) are processors acting under the instructions of the University of Oxford.
Microsoft Limited provides IT hosting services to OUH and will store the Data as contracted by OUH.
A Patient and Public Involvement and Engagement (PPIE) group assisted with the grant application and another PPIE group assisted with the programme design after the grant was received and has an on-going role in advising the project through quarterly meetings and ad hoc email/video calls as needed.
GEHC are a commercial organisation and a processor under this Data Sharing Agreement for the purpose described. They will be using the data obtained under this agreement, in combination with the data DART is obtaining from the Targeted Lung Health Checks, to develop and improve artificial intelligence algorithms to better predict respiratory conditions such as cardiac diseases and COPD.
GEHC make a considerable (over 1 million pounds) kind contribution to the project and, as a commercial company, would naturally expect a commercial gain and the publicity associated with participating in a prestigious project which has attracted Government and national media attention. The expected health and social care benefits of their contribution into the project outweigh any commercial gain they would receive from the project.
Alongside the University of Oxford academic team, Roche Diagnostics Ltd make an in-kind contribution of their expertise into WP4 and have loaned a Digital Pathology Scanner to DART (in Oxford) and provided a staff member to operate this. As a funder and commercial organisation, Roche Diagnostics reasonably expect a commercial return and to receive acknowledgement and publicity for their participation in the project. In turn, their contribution to the earlier detection of lung cancer far outweighs any likely commercial return.
Expected output
The expected outputs of the processing will be:
> A report of findings to funding bodies (Innovate (part of UK Research and Innovation); Cancer Research UK (CRUK)): expected March 2024
> A summary of findings to relevant stakeholders (Innovate UK, CRUK, Roche, Optellum, GEHC): Expected March – May 2024
> Submissions to peer reviewed open-access journals: Specific target journals will be identified as the work progresses to ensure the study findings are communicated to appropriate audiences. Expected January – June 2024
> Presentations will also be given at national and international conferences relevant to lung cancer and screening
The outputs will not contain NHS England Data and will only contain aggregated information with small numbers suppressed as appropriate in line with the relevant disclosure rules for the dataset(s) from which the information was derived.
The outputs will be communicated to relevant recipients through the following dissemination channels:
> Journals
> The results of the planned analyses will be communicated directly to the National Screening Committee (NSC) upon the completion of DART in March 2024 by collaborators who also have a position on this committee
> Results will be shared with the staff at the TLHC centres and other operational staff involved in lung cancer screening
> DART Patient and Public Involvement and Engagement (PPIE) contributors, including the Roy Castle Lung Cancer Foundation, will be consulted to communicate findings to the general public and patients in an accessible way
> Results will be communicated to the National Screening Committee upon the completion of DART in March 2024
> National and International conferences
Outputs are expected throughout 2024.
Benefits reported
Yielded Benefits is not a requirement for new applications.
Register history
When this agreement appeared in, or was edited in, each monthly edition of the register. Built by comparing every edition this site holds.
-
December 2023 —
first listed. 1 version: DARS-NIC-668928-L9N5F-v0.6
-
April 2024
1 version added: DARS-NIC-668928-L9N5F-v1.3
-
April 2026
1 version added: DARS-NIC-668928-L9N5F-v2.11
Cite this page
NHS England (2026) Data Uses Register, September 2026 edition, agreement DARS-NIC-668928-L9N5F, “DART: The Integration and Analysis of Data using Artificial Intelligence to Improve Patient Outcomes with Thoracic Diseases.”. Read via NHS Data Access Explorer (unofficial), https://healthdatauses.uk/agreements/dars-nic-668928-l9n5f/ (accessed [date]).
This address stays the same, but the page is rebuilt with each monthly edition, so the citation names the edition it shows. Every edition's data is kept in the facts store.
Source: datausesregister_september2026.xlsx, September 2026 edition of the NHS England Data Uses Register. Search that workbook for DARS-NIC-668928-L9N5F to see the original rows.