IT management consultants often seem to use the analogy between data and water. One thing that their glossy websites don’t usually say, though, is that, as well as sailing to your destination on water, you can also drown in it. Several years ago someone calculated that the amount of information transmitted globally, now measured in zettabytes, is the equivalent of each person on earth receiving 174 newspapers every day. We have now, of course, entered the mis-information age, of alternative facts and fake news. Mis-information and mis-representation are so problematic that the chair of the UK Statistics Authority has twice in the last few months publicly rebuked senior politicians for inaccurate statements about key matters of fact.
We should be able to do better than that in the NHS. The universal registration system of patients with GPs, the cycles of National Audits, the accurate collection of prescribing and clinical activity data in general practice and in hospitals, the use of large sentinel clinical networks such as the Royal College of General Practitioners’ Research Surveillance Centre, and the careful curation of the cancer registries together provide an unparalleled resource for examining and improving patient care, measuring health service performance, and undertaking a wide range of research. Ironically, the first researchers to exploit the potential of the General Practitioner Research Database (GPRD), established in 1987, were working in Boston and Madrid. The GPRD, as part of the Clinical Practice Research Datalink (CPRD), now contains over 14 million years of patient data and is an exceptional resource. Much of what we know about the early diagnosis of cancer and the measurement of cardiovascular risk is derived from GPRD/CPRD studies. In this issue of the journal the various sources of information held in NHS databases are described by Lucy McDonnell and colleagues, with a number of references to the valuable research for which they have formed the basis.
Accurate data collection underpins some important papers in this month’s issue of the BJGP. The careful work conducted by Sally Hull and her colleagues, using linked primary and secondary care data, has demonstrated that the population burden of comorbidity is the strongest predictor of attendance at A&E departments, explaining much of the association with social deprivation, and countering the argument that access to general practice is at fault. As a corrective, however, the interesting study by Peter Tammes and colleagues on excess winter mortality among the elderly cautions against over-confidence in the ability of routinely collected data to identify older patients at most risk from cold weather. The study by Luke Mordecai and colleagues, demonstrating the seemingly inexorable rise in opioid prescribing for chronic pain, drew its information from other publicly available data sources, including the Health and Social Care Information Centre and the Office for National Statistics.
In Life & Times Mila Petrova and Stephen Barclay take a critical look at the government’s response1 to the National Data Guardian’s report2 on data security, consent, and opt-outs, and are concerned that it may make matters even worse, although anything worse than care.data is difficult to imagine. They conclude that:
‘If we do not want a repeat of the care.data experience, we need, among other things, to keep data sharing for individual care strongly in focus, be extra thoughtful about words and context, and avoid turning simplification into misinformation.’
I have just finished reading The Ministry of Utmost Happiness by Arundhati Roy. As well as being exhilarated by her absolutely brilliant writing, and transported by her astonishing imagination, I became more and more aware of how ignorant I have been over the last decades of the seismic political and social events that have taken place, and are still happening, across the Indian sub-continent. The information age may be here, but it certainly isn’t evenly distributed.
- © British Journal of General Practice 2018