Health and data strata 01 10 12 final

Editor's Notes

• #4: <Need to add pictures of projects to this>
• #5: What is innovation?And how does Big Data relate to itDefinitions – new ideas put into practiceA more radical view – new actors – and that’s the theme we’d like to use for this presentation.Where might the new actors come from in healthcare? And how is data making it possible for new actors to get involved?
• #6: Data can be used in many different ways – we wanted to try and break down the monolith of ‘big data’ projects to look at how they create innovation.At the left-hand end, data is used to expose information to new audiences, and can deliver huge improvements and efficiency savings. But it doesn’t change the model or admit new actors. They take existing processes and approaches, and use data to find ways to make them better.At the right-hand end are new approaches and models that wouldn’t be possible without the data to support them. These are more transformative, and create opportunities for new actors from outside the existing system to get involved.
• #7: Health is always changing:Particular nature of innovation in healthGrowing demand Changing disease burden “Patients are waiting” Who are the radically innovative consumers going to be?long-term, incurable, chronic conditionsCancer survivors and their familiesData scientists and geeksThese groups are ready to be experimental, ready for a new approach, willing to invest time, happy to balance privacy and progress
• #8: Longstanding demand for innovation in health – huge user groupIn previous centuries, health was about short, sharp, often violent interventions.But it has always generated innovation – a huge array of surgical tools displayed here
• #9: In the following centuries, it moved to an institutional model, creating the structures and organisations to care for episodes of ill-health.Development of nursing practices, asylums, disinfection.
• #10: Modern healthcare is still based largely on institutions. The big components of the system haven’t changed all that much since Victorian times. We still recognise hospitals, mental health institutes, family doctors.
• #11: Now we need to look at healthcare differently.It becomes a continuous process, that happens both with and outside the formal healthcare institutions.
• #12: These four generate data about the patient Data vs. knowledgeSpeed and reliability of clinical research not as good as it could be Clinical knowledge difficult to interpret and apply Applied knowledge (comparatively) neglected This is a standard, institution-based model.
• #13: The new way of describing data in health acknowledges that institutions and their records are only one part of a broader system.
• #14: When we think about health and long-term conditions, it’s not only the clinical and institutional measures that are important (highlighted in blue)A whole range of more social, lifestyle, behavioural factors have a huge influence, and can’t always be captured by institutional data.
• #15: Changes the gatekeepersMore and better data: Patient data (records) Patient data (lifestyle, experience, informal) Admissions and hospital dataSensors and monitors Open science Social networks
• #16: Some of these trends are common to other spheres of big data, and others are more specific to health or life sciences.Capacity for individuals to gather more data Ability to manage distributed problem solving Improvements in organisation of knowledge Tools to interpret data (modelling etc) Ability to integrate data Semantic data – from Royal Society (2012) ‘Science as an open enterprise’ – map from the Linking Open Data Community Project
• #17: We wanted to break down the range of innovations in health, and to try and tie together some of the separate ideas and discussions that link healthcare and data.We think that innovation is coming from both outside and inside the established health system; and is being used to both create and capture data and knowledge in difference ways; and to take data and apply it to care in new ways.
• #18: There are a series of separate constituencies that are interested in these questions:Top left: medical research, patient data for research consultation, joining up medical databasesBottom left: quantified selfers, peer health networks, citizen science enthusiastsTop right: stratified and genetic medicine, clinician decision support, evidence-based medicineBottom right: health bots, online health services
• #19: We’d like to walk through an example of each of these areas of innovation.
• #20: National diabetes register for Scotland covers 995 of 1000 GP practices in Scotland – estimated 99.5% completenessAllows you to do population–level analysis on gender, mortality, deprivation, etc.Diabetes increases the risk of dying by an average of: - 35% for men - 62% for women270,000 people, updated daily by GPs and practice nurses all over ScotlandLinking data to hospital admissions data, mortality data, economic deprivation indicesFuture – intending to link to genetic data and biomarkers, to education data.Dealt with using a safe haven to link data, without seeing identifying dataNot just for research and analysis – GPs see the benefit as well, which helps take up. Helps to monitor progress of diabetes patients and meeting targetsAlso making the data available back to patients – ‘My Diabetes, My Way’ The Scottish Diabetes Network can best be described as a nation-wide clinical electronic system for the management of diabetes, which complements the existing primary care networks and diabetes treatment systems. It accumulates information on a daily basis from all GP’s surgeries and hospitals. Relevant medicinal information, such as biochemical details relating to diabetes and cardiovascular health, are recorded and included in the database. Data for 205,000 people living with Diabetes is recorded on the system, and used to refine treatment and participation in further research and trials. The network falls into this category due to its focus on generating information from within the NHS Scotland, using the existing institutional structures and current technologies in a novel way to achieve a quantitative and qualitative improvement in data on Diabetes in Scotland. It operates entirely within the Scottish NHS, and its focus is on the incremental improvement of existing techniques using the data gathered.   The goals of the network are to expand the use of patient data obtained from the Scottish Diabetes Research Network in clinical trials, and to apply the data to challenges other than diabetes. Another goal is to further enhance the co-operative working of researchers in diabetes. Enhancing treatment for diabetes sufferers in a range of qualitative senses is also high on the agenda. Utilising the data in a more efficient way, in line with the more widespread efforts at co-ordination amongst researchers, seems key to the future of the issue.
• #21: A real example of the need to bring together lots of data in one place to make treatment decisions.Cancer is not one thing – it’s many different diseases, with different causes, that have in common unchecked and invasive cell growth.Many of the new generation of cancer treatments now being launched and in development only work in cancers driven by a single genetic fault or protein.In order to make treatment decisions, you need to test and stratify patients according to what characterises their cancer, and bring that information together with knowledge about the range of drugs and what they target, plus side effects and other considerations, to make a useful treatment decision.Cancer Research UK want to make genetic testing part of standard NHS practice, as well as developing a database of genetic cancer data for research, and are running this pilot scheme with the Technology Strategy Board to make sure that happens. 95% of patients approached have agreed to be enrolledThis is a preview of an approach which will need to be used with more and more diseases in future.In terms of new actors, this creates a challenge, as you rely not just on the companies developing the drugs, but also those developing the clinical tests, which are often different companies, and often develop separate tests for every drug. There’s the potential for a big co-ordination problem there.Story of getting involved:Cancer patients approached and asked to participateIf they agree, a blood sample is taken for genetic analysis – this is a sample of normal DNA for referenceAfter the patient has surgery or a biopsy, a tissue sample is taken to analyse DNA from the cancerous cellsGenetic sequencing of the mutations are sent back to the clinic to be added to the patient’s medical records, as well as stored in a research DB.At present, this doesn’t change the treatment decision, but it will do in the near future, and in the meantime, this is creating a huge research resource, and embedding the practice of genetic testing into the NHS.===
• #22: Story of PLM – ALS diagnosis.Story of PLS research – enough data on the progression of this rare variation (5% of ALS population) to identify a different pattern of decline, distinct from ALS. No-one else had been able to assemble enough patients with this disorder to gather this data.Peer supportGenerating research – now have 27 published studies from thisPatientsLikeMe is a social networking website focused on the sharing of health information and experiences. Its primary function is to connect patients experiencing similar health issues, and to collaboratively find solutions and treatments to those problems based on prior patient experience. It then shares the data it accumulates from patients with its partners – mainly in industry, academia and the pharmaceutical world. The sale of data is its primary profit-driver. It is the first sophisticated model of collaborative health data sharing based primarily on the outside of the system, but including the internal healthcare world too.PatientsLikeMe is driven by the voluntary input of patient data, meaning its primary means of gathering information comes from outside of the established health care system. Its role as an accumulating network is enhanced by the voluntary component of its model – patients are not encouraged or pressured into sharing, but come to the site with the explicit aim of seeking information and advice from outside of the formal health structures. The data is then sold to stakeholders, but PatientsLikeMe has no role in the application of data to product or service development. The evolution of PatientsLikeMe is likely to be a process of incremental growth and improvement, without expanding into new channels. As a relatively new company – and one of the first applications of a new concept – PatientsLikeMe is likely to expand within its current category rather than attempt to spread to new ones. This is particularly true as PatientsLikeMe is so explicitly outside of the formal structures, that expansion that crosses this boundary would radically change the nature of its operation and weaken its unique selling point.
• #23: A system to monitor and, if you wish, share your health and wellness dataOwned by citizens – they have rights to delete data whenever they want, to grant access to third parties and to revoke that access at any time.Platform helps companies to create new services, knowing that the data is secure and citizen-controlledEstablishes a new marketplace for health service innovation in FinlandTaltioni is an online database and service platform that allows Finnish citizens to record and monitor their health, as well as giving them direct access to all past information on their own health and well-being. Users have the option to share their own data with certain external actors, including private companies. It is run and managed by the Finnish Innovation Fund, which is accountable to the Finnish Parliament. Taltioni is placed in this category because it is both an open, patient-driven data accumulating and sharing system, and has strong links to government and the formal health structures. In the future, Taltioni looks set to expand its base of users, and become more deeply integrated with the Finnish health system as more technological enablers become available. It is possible that users will be encouraged to adopt more quantified self style technologies.
• #24: ContinuousPeople-poweredNetworkedInvolves lots of new actors and services