Big Data: How codebreakers helped transform healthcare

30 January 2018

 “We can only see a short distance ahead, but we can see plenty there that needs to be done.” Alan Turing [1]

When you arrive at Bletchley Park, you might want to make your way to Hut 8 where Alan Turing and his colleagues deciphered Naval U-boat messages. Building on the idea of the Universal Turing Machine (one machine, for handling all programmable tasks) [2] Turing and colleagues developed the Enigma machine.

Towards the end of World War II, the first programmable computer in the world, Colossus, was developed at Bletchley by Tommy Flower’s and colleagues to decode high level German military intelligence throughout occupied Europe.  (A working Colossus can be seen at Bletchley Park).

In 2015 Royal Mail honoured Tommy Flowers with a first-class Colossus stamp

The birth of modern computing can be traced back to many of these early developments at Bletchley Park. Beyond the war, the pace of development continued until in 1953, IBM, introduced the first commercial, scientific computer with memory, operating systems, storage and the ability to print. Within a decade computers had found their way into healthcare, mainly at this time to computerise billing and administrative tasks. It wasn’t until the late 80s, though, that computers were integrated into clinical care as a means to improve health. [3]

By the early 90s, the majority of UK general practices were computerised with repeat and acute prescribing, the collation of annual data and audits being the most used applications.  While GP’s were quicker to adopt computers; hospitals were not: 76% of consultants had access to NHSnet for email and browsing but only a few used computer-based patient records to facilitate care. [4]

Automation of health care orders, results and care plans led to the accumulation of large databases of clinical findings. The availability of these large-scale databases along with the development of relational database software that connects one set of data to another and the wide-scale use of the Internet means we can better understand and transform healthcare through the use of ‘Big Data’

Today’s systems hold your complete medical history. Along with the continual accrual of administrative data, organisations like the NHS now hold millions of patient’s records. And when you use these records in an anonymised way for research purposes then you have what is known as Big Data.

One example in the UK is the Clinical Practice Research Datalink (CPRD), which holds records for over 5 million currently registered and active patients in primary care. Research in the Department of Primary Care Health Sciences recently provided NHS England with the first comprehensive data on GP workload.  By using CPRD records from over 100 million patient consultations, the results, published in The Lancet, [5] showed that workload increased by 16% over a ten year period, indicating GPs are working the equivalent of one extra day week.

Research in the EBM Datalab, for instance, is using Big Data to build innovative, live tools to help make healthcare data more impactful in the real world. Every month, the NHS publishes anonymised data about prescription medicines in the UK. The data is overwhelming (600 million rows), which makes it hard to determine anything useful. With OpenPrescribing, the team are making it easy to monitor prescribing trends, spot unusual patterns, and see who is prescribing what to improve care

Digital health applications that use individuals data has the potential to reduce inefficiencies in healthcare delivery, improve access, reduce costs, increase quality, and make medicine more personalised and precise.

Digital tools are currently in use by patients to manage their blood pressure better through online portals, and People with type 1 diabetes are improving their care through text message reminders. A substantial number of smartphone Apps are in development that aims to use your data to improve healthcare. In Asthma, A 2017 review found 38 different apps (13 available for both iOS and Android), that had some potential to develop asthma self-management.

Bletchley Park Week

Kellogg College and Oxford have a unique relationship with Bletchley Park. Nearly 10,000 people worked in the broader Bletchley Park organisation. Recruitment to Bletchley occurred from Oxford and Cambridge universities. Searching the Bletchley Park Roll of honour, of those that worked at Bletchley, I found 179 named individuals with a connection to Oxford. One of these, for example, was my wife’s great aunt, Daphne Mary Moss, who attended St Hugh’s College Oxford and worked in Hut 10.

Work at Bletchley has informed how we can use data to improve health services. Turing’s universal machine concept (you can read a copy of the 1936 paper here), along with an understanding of the speed and reliability of the electronic technology and the inherent inefficiency in designing different machines for different processes were basic computing principles developed at Bletchley, which still underpin much of our current thinking and work.

We can learn an immense amount from reflecting on past experiences to consider what we might do, and not do, differently in the future.  I look forward to seeing you at the Bletchley Park Week where we will continue the conversation on the past and present relationship between Bletchley Park and its role in transforming healthcare.

Bletchley Park Week 2018

For a full list of events happening during Bletchley Park Week 2018 (4th – 10th March), please visit this page.

Carl Heneghan is Professor of Evidence-Based Medicine and Director of Programmes in EBHC at the Centre for Evidence-Based Medicine, Dept of Primary Care Health Sciences, University of Oxford and a Fellow of Kellogg College. 

References
[1] A.M. Turing. Computing machinery and intelligence Mind, 59 (1950), pp. 433-460

[2] Turing, A.M. (1936), “On Computable Numbers, with an Application to the Entscheidungsproblem”, Proceedings of the London Mathematical Society, 2 (published 1937), 42 (1), pp. 230–65, doi:10.1112/plms/s2-42.1.230 (and Turing, A.M. (1938), “On Computable Numbers, with an Application to the Entscheidungsproblem: A correction”, Proceedings of the London Mathematical Society, 2 (published 1937), 43 (6), pp. 544–6, doi:10.1112/plms/s2-43.6.544). (Online versio.)

[3] Alan A Montgomery, Tom Fahey.  A systematic review of the use of computers in the management of hypertension. Epidemiol Community Health 1998; 52:520–525 http://jech.bmj.com/content/jech/52/8/520.full.pdf

[4] Why general practitioners use computers, and hospital doctors do not—Part 1: incentives BMJ 2002; 325 doi: https://doi.org/10.1136/bmj.325.7372.1086 http://www.bmj.com/content/325/7372/1086.full.print

[5] Hobbs FDR, Bankhead C, Mukhtar T, Stevens S, Perera-Salazar R, Holt T, Salisbury C; Clinical workload in UK primary care: a retrospective analysis of 100 million consultations in England, 2007-14.Lancet. 2016 Jun 4;387(10035):2323-2330. doi: 10.1016/S0140-6736(16)00620-6.

Comments