2. El juego en el periodismo
3.3. El espectáculo como eje central
People increasingly leave digital‘traces’ relevant to health in multiple different information systems as they interact with health and other services and go about their everyday lives. The potential for such data to inform, and potentially transform, our understanding of health and disease has been recognised by world leaders. For example, George Osborne, the UK Chancellor of the Exchequer from 2010 to 2016, said in a Science Speech
to the Royal Society in November 2013 that the UK has‘some of the world’s best and most complete data-sets
in healthcare’ (© Crown copyright; contains public sector information licensed under the Open Government
Licence v3.0).1Following the 2015 State of the Union address, the White House signalled an initiative that
would‘catalyse a new era of data-based and more precise medical treatment’ (reproduced under a Creative
Commons Attribution 3.0 License).2
Data tapestry
The electronic record relevant to health is diverse. This‘data tapestry’ of information (Figure 1) spans records
in primary care and in hospitals, structured data (codes, values) and unstructured data (text). It includes information on genomics, through to social media, the sensed self and wider social determinants of disease. Deriving information and knowledge from one or more sources of these rapidly developing sources of health-relevant data might be of great benefit to inform improvements in care and outcomes of disease. Translational gap
One model that links the process of science and improvements in health care and that has attracted considerable attention in the UK is the translational model, as laid out, for example, in the Cooksey
Review.4This identified two gaps in the translation of biomedical science to health care, construed as an
essentially linear process. The first translational gap is the generation of new therapies to improve health care based on biomedical science. Epidemiological research using electronic health records (EHRs) can help to close the first translational gap by enabling researchers to quantify the need for new therapies, identify risk factors and (with genomic data) suggest potential new drug targets. Randomised clinical trials with follow-up through EHRs can be cheaper and simpler than conventional investigator-led trials and may be particularly helpful for increasing the evidence base for non-drug interventions.
However, it is in the second translational gap, that is, the introduction of new therapies into clinical practice, on which EHR-based research programmes may have the greatest impact.
For example, new prognostic risk scores may help interventions to be targeted most appropriately, decision support tools embedded in clinical systems may help to improve clinical decision-making and health economic analyses can ensure that treatments are managed in a cost-effective way.
Learning health systems
A second, more recent, model linking data, information and knowledge to improvements in health care
comes from the concept of a learning health system5in which science and informatics, involving the
real-time access to knowledge and digital capture of the care and outcomes, are a central component. Electronic health records: what does the UK contribute?
Although there may be a strong scientific and health system case for making much better use of the
diversity of the‘data tapestry’, in reality, existing research efforts have been largely focused on structured
data collected as part of an interaction with one or more components of the health system.
The UK is one of the few countries in the world in which a picture of the patient journey can be traced through EHRs spanning primary care, hospitals and, ultimately, death registries. The UK has the potential
to address research questions that would currently not be possible in Denmark and Sweden (Table 1), countries that do have outstanding national registries but that have not brought national primary care records to research in the same way as is possible in the UK.
A combination of features in the UK underpin the potential of EHRs: Free at point of use
+
unique health identifier +
≈ every citizen registered with a general practitioner (GP) +
≈ every general practice uses EHRs +
wide range of linkage possibilities, including to the national heart attack registry [Myocardial Ischaemia National Audit Project (MINAP)].
Linkage of EHR and registry data, which are commonly held separately across multiple sources, with bespoke phenotypic and genetic information can generate a unique platform to explore why diseases occur and
Types of data Medication Demographics Encounters Diagnoses Procedures Diagnostics (ordered) Diagnostics (results) Genetics Social history Family history Symptoms Lifestyle Socioeconomic Social network Environment
Structured data Unstructured data
Probabilistic linkage to obtain new types of data
Probabilistic linkage to validate existing data or fill in missing data
Examples of biomedical data Ability to link data to an individual Data quantity
More Less
Pharmacy data
Claims data Registry or clinical trial data
Health-care center (electronic
health record) data Easier to link to individuals
Harder to link to individuals
Only aggregate data exists Data outside of health-care system
Electronic pill dispensers OTC
medication Medication filled Dose
NDC
Employee sick days Death records
23andMe.com Police records Ancestry.com
Indirect from OTC purchases Fitness club memberships, grocery store purchases
Census records, Zillow, LinkedIn
Facebook friends, Twitter hashtags Climate, weather, public health databases,
HealthMap.org, GIS maps, EPA, phone GPS
RxNorm HL7 Visit type and time SNOMED Chief complaint Differential diagnosis ICD-9 ICD-9 Pathology, histology Lab values, vital signs SNPs, arrays Tobacco/alcohol use Radiology CPT LOINC ECG Route Allergies Out-of-pocket expenses Medication prescribed Medication
instructions Medication taken
Diaries Herbal remedies Alternative therapies News feeds Reports Tracings, images Digital clinical notes Paper clinical notes Credit card purchases Home treatments, monitors, tests Personal health records Patients likeme.com 1 1 2 2 Blogs Tweets Facebook postings 1 2 1 2 Physical examinations
FIGURE 1 The tapestry of potentially high-value information sources that may be linked to an individual for use in health care. CPT, current procedural terminology; ECG, electrocardiogram; EPA, US Environmental Protection Agency; GIS, geographic information system; GPS, global positioning system; ICD-9, International Classification of Diseases, Ninth Edition; LOINC, Logical Observation Identifiers Names and Codes; NDC, National Drug Code; OTC, over the counter; SNP, single nucleotide polymorphism. Reproduced with permission from Weber et al. Finding the missing link for big biomedical data. JAMA 2014;311:2479–80.3Copyright © 2014 American Medical Association. All rights reserved.
progress, to investigate quality of care and to identify opportunities to improve health outcomes. This new data revolution has attracted substantial investment and promotion with initiatives to improve access to and use of linked EHR data, which will serve to decrease the burden of obtaining data in a research-ready format and encourage research collaboration.
In what senses are linked electronic health records‘big data’?
One widely used definition of big data concerns the 4 Vs: volume, variety, veracity and velocity. EHR data are
potentially‘big data’ in every sense; not only do they cover a large number of people (entire populations)
with a large amount of information [e.g. 2 million patients, 5 billion rows of data in our ClinicAl disease research using LInked Bespoke studies and Electronic health Records (CALIBER) data platform, which forms the foundation of this programme] per patient, but the data are highly varied and complex, with different ways of coding information (e.g. by means of International Classification of Diseases and Read codes). The data validity (‘veracity’) is simultaneously a concern for clinical care and for research. Clearly, having data in real time (‘velocity’) can be crucial for clinical decision support. Harnessing EHR data for research requires a deep understanding of the health-care system from which the data originated, as well as the statistical and computing skills to analyse large data sets.
In 2011 the UK government published a Strategy for UK Life Sciences,16which placed EHR research as a
central part of the strategy to accelerate health research in the UK. The UK is unique in being the only country with national cardiovascular disease (CVD) registries and primary care data (including important information on cardiovascular risk factors and their management in the community) available at a scale for research. The recent report on personalised health care by the National Information Board and Genomics
England17urged the NHS to transform its use of information in order to provide better care as well as to
enable better research. This will require a substantial increase in the maturity of EHRs in the UK as well as a framework to make these data available for research.
TABLE 1 Availability of primary care data for research in different countries
Country
National or
regional Primary and ambulatory care data available for research linkages
UK National CPRD,6
accessed through Academic Health Sciences Networks Sweden National Primary care is organised regionally; national initiatives in SwedeHeart7
and the National Registry of Secondary Prevention
Denmark National Register of Medicinal Product Statistics8,9
Canada Regional Ontario, Institute for Clinical Evaluative Sciences10
Regional Ontario Health Insurance Plan Physician claims database USA National Medicare (for people aged≥ 65 years) (see www.medicare.gov/)
National Million Veteran Program (see www.research.va.gov/mvp/veterans.cfm) Regional Mayo Clinic11
Regional Rochester Epidemiology Project, Olmsted County (see http://rochesterproject.org/) Regional Kaiser Permanente California Research Program on Genes, Environment, and Health12
Regional Intermountain Healthcare13
Republic of Korea
National National health insurance claims database from the Health Insurance Review & Assessment Service14
CPRD, Clinical Practice Research Datalink.
Reproduced from Denaxas et al.15Published by Oxford University Press on behalf of the International Epidemiological
Association. © The Author 2012. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by-nc/3.0/), which permits non-commercial reuse, distribution, and reproduction in any medium, provided the original work is properly cited.
Current state of electronic health records in UK hospitals relevant to quality and outcomes research
Although UK general practices have, via the Clinical Practice Research Datalink (CPRD)6and other
initiatives, contributed EHR data that have led to> 1000 peer-reviewed publications, the use of data
within hospitals beyond the simple Hospital Episode Statistics (HES) returns to generate useful knowledge has led to very few publications. One of the reasons for this is that UK hospitals, like those in Europe, have been at a low level of maturity, according, for example, to the Healthcare Information Management Systems Society classification (Table 2). This is changing rapidly in the UK in light of the emerging strategy from the National Information Board and NHS Transformation programmes such as the Genomics
England-led 100,000 Genomes Project.19
However, part of the government’s plan for making health-care information available for research, the
care.data programme,20aroused national anxieties about patient confidentiality and had to be suspended
pending further public consultation. This illustrates the sensitivities that must be respected in EHR research.
Technical solutions such as data pseudonymisation and secure computer workspaces (‘safe havens’) can
enable sensitive patient data to be safely used for research, but it is vital to involve patients and the public in balancing privacy concerns against the benefits of research.