Through DataNet, the patient medical records held by all GPs in Lambeth have been systematically developed, anonymised and combined into a single database for the borough. This enables detailed analysis of the borough population to identify groups who can be followed up for targeted health interventions based on prevalence and risk.
Data security and patient confidentiality are protected through a specially-designed data sharing agreement. Lambeth DataNet began with a clear rationale to improve care and health outcomes for people in Lambeth, particularly those at risk of health inequality.
●The scope of the project, which involved many different organisations spread across London, complex data, software and hardware negotiations, and extensive legal and governance scrutiny
●Under the new Caldicott 2 regulations, GPs are assigned the role of ‘Data Controllers’ for the patient healthcare data which they hold. Release of even anonymised data is strictly controlled. Anonymised data can only be released after the local population have been informed and a legal document called a ‘Data Sharing Agreement’ has been negotiated and signed by each general practice. This whole process took 18 months of intense negotiations with all parties. There are huge fines for data breaches which meant that costly procedures had to be put into place to avoid any risk to data security.
●GPs were concerned that the data could be used for performance management, or for commercial purposes; we received a strong steer from local GPs to ensure that under no circumstances could this data be used commercially, nor for performance management.
●We had to coordinate engagement with practices across Lambeth concerning a number of technical issues and requirements
●The local community and patients had to be consulted. This required a process known as Fair Processing Notices to be issued. We worked with the NHS Lambeth CCG communications team to devise these leaflets and communication channels.
●Setting up a team of data analysts representing the CCG, Public Health and King’s College London (the academic partner)
●Collaboration with clinical colleagues – to discuss detailed findings from DataNet and other sources to prioritise and develop the broad intervention of the programme
●Agreeing how best to present data which was meaningful for the GPs who, ultimately, had provided the data in the first place. What data did they want? And in what format?
●Agreeing that the three principles of using the data were: reducing inequalities, increasing quality, feeding back data in a way which benefited the GPs (who provided the data) and the local community (so that they have a stake in the project
●Meetings with GPs and the LMC to negotiate several iterations of a new DSA.
●The new DSA went through 14 iterations and 18 months of negotiation before being approved and signed by all GP practices in Lambeth. It protected GPs from any financial risk related to information governance and the LMC helped us include a clause ensuring the data would not be used for either performance management or commercial purposes. Once approved, the LMC was able to ‘recommend’ that all GP practices in the CCG should sign the DSA.
●The project has enabled GPs in Lambeth to develop highly targeted initiatives aimed at many different conditions and treatments
●DataNet assesses multi-morbidity to support Locality Care Networks to deliver integrated patient care
●DataNet produces qualitative reports which address health inequalities
●It has enabled the delivery of healthcare planning at scale, including the development of a Lambeth Clinical Effectiveness Group.
●In 2014, significant new investment for the Lambeth Early Action Partnership project, a Big Lottery Fund award of £36 million targeting asthma, epilepsy and cerebral palsy in children, was based on evidence provided by Lambeth DataNet
●The aim to improve value in the delivery of primary care services began with a drive to achieve high levels of patient profile recording: ethnicity, language and country of origin. This data could then be used to identify groups at particular risk – and to target interventions for individual patients.
●Lambeth DataNet provides access to patient data to improve understanding of conditions and reduce variations in care, and hence drive up value in primary care. It also provides unique evidence-based reporting to support commissioning decisions and service transformation.
●DataNet has been used as the data source for National Institute for Health Research funded projects investing in local primary care research, totalling over £2.1 million
●DataNet was used to run a ‘Quality Outcomes Framework (QOF)-Plus’ scheme in which GPs were given additional funding for achievement of QOF-stretch targets, over and above standard QOF targets, in hard-to-reach groups.
As a borough with a large multi-ethnic population and pockets of extreme social deprivation, Lambeth GPs know they must target inequality and develop tailored solutions for the many different communities they serve. Due to the pioneering use of patient data, DataNet required equally ground-breaking information governance regulation. A Data Sharing Agreement was developed and agreed by the London-wide Local Medical Committee and the project has been registered with the Information Commissioner’s Office. Patient safety was also a focus alongside efficiency. DataNet has been used to produce risk summaries of Lambeth patients, including those: prescribed potentially dangerous drug combinations (e.g. amlodipine and simvastatin); prescribed treatments now known to be ineffective at preventing stroke (e.g. aspirin in patients with atrial fibrillation); and those at risk of cardiac rhythm problems (e.g. patients on high dose methadone).
The project has enabled GPs in Lambeth to develop highly targeted initiatives aimed at many different conditions and treatments, including chronic obstructive pulmonary disease, heart failure, blood pressure, diabetes, stroke, hypertension, asthma, epilepsy, cerebral palsy, and atrial fibrillation. Qualitative success indicators include improved access to primary care data for the Lambeth Public Health team, and collaborations with research, such as King’s College London has supported development of the anonymised database and offered support with statistical analysis (particularly with displaying age and sex-adjusted data, funnel-plots etc.) and with geographical mapping.
DataNet has produced annual Health Equity Reports for each GP practice in Lambeth. These summarise the key health inequalities based on ethnicity, deprivation, age and gender in terms of achieving public health targets in atrial fibrillation, cardiovascular disease risk and diabetes. The Lambeth Early Action Project, which provides a wide range of programmes to support parents and change the way services work with families, is using DataNet to monitor progress. DataNet found high rates of smoking were demonstrated in patients with serious mental illness, which led to the commissioning of a dedicated smoking cessation service.
DataNet supported a project for improving diabetes care, which led to significant improvements in the quality of care and health outcomes of local people living with diabetes, including: earlier detection, more people being supported with pre diabetes, increasing people agreeing goals and care plans, and healthier blood pressures, cholesterol, and blood sugars. DataNet was used on a study of prevention of cardiovascular disease, which found that women with CVD should be offered statin prescription and may require higher statin dosages for improved control. Lambeth has seen huge improvements in anticoagulation rates in atrial fibrillation, which prevents large numbers of strokes. We are also currently working with clinical colleagues on blood pressure control and detection.
The systems devised for DataNet are scalable and replicable, and can be adopted by any CCG. This has been made available to several other interested CCGs in England. Similarly we have made freely available all the technical details of the IT infrastructure built to house the data; other CCGs can either share the resource or be offered a template to construct a similar system. Two London CCGs are already replicating an approach using the same methodology. To further increase scale we have negotiated the first CCG contract with EMIS Web, the GP clinical software supplier, for anonymised data extraction.
The data extraction module has been piloted and after much negotiation, is now suitable for other CCGs to adopt. DataNet has surpassed its original goal in several respects, for example, in August 2016, we supported the healthcare system in Sierra Leone. They had no data for rates of long-term conditions and noted high rates of hypertension and diabetes in the Sierra Leonean population. DataNet has data for 990 adults from Sierra Leone and living in Lambeth. Long-term conditions prevalence data from this group supported the Sierra Leone Ministry of Health in its healthcare planning.
There is clear evidence DataNet has improved value in primary care. Qualitative evidence of improved value includes support for public health initiatives, research collaborations, attracting funding into the borough, and even spreading knowledge internationally. Professor Paul McCrone, Professor of Mental Health Economics at King’s College, used DataNet to look at the health economics of discharging mental health patients from the Community Mental Health Team (CMHT) and out into the community. This project used Lambeth DataNet data linked to data from South London and Maudsley Trust. Those discharged to primary care had costs which were 53 per cent lower than those who continued with CMHT care. Overall, care costs were 71 per cent lower in primary care than in secondary care.
All parties have been involved in the redesign. Strong community engagement from the outset has meant that DataNet is one of the few anonymised databases of primary care health data which has explicit community support – this means it has avoided the fate of national schemes which failed to secure support and had to be abandoned (e.g. the care.data scheme). HealthWatch Lambeth worked with the project team and led an 18-month programme of community consultation about the use of anonymised health data. As a result, before access is granted to the DataNet data for researchers, a clear benefit to local patients must be demonstrated. Similarly, for each initiative which is developed using the data, opportunities for co-design are sought. For example, we know from DataNet that the West African community has the worst blood pressure control in Lambeth.
HealthWatch is working to engage the West African community to develop culturally appropriate initiatives to improve blood pressure control, targeting both lifestyle issues and access to more proactive primary care. Several DataNet projects have supported patients to self-care. For example, one locality, Stockwell, has a high proportion of Portuguese residents with some small areas having over 30 per cent Portuguese. DataNet data demonstrated high cardiovascular risk with this national group, and supported development of a drive to reduce the high rates of smoking and obesity in the local Portuguese community. The Portuguese Community Project, won the NHS Lambeth Clinical Commissioning Group (CCG) 2016 Lammy Award for community engagement and promoting engagement with health initiatives.