In November 2019, the US Federal Government announced that it would conduct an enquiry into Project Nightingale, a collaboration between the ubiquitous Google and an American hospital group. So what is Google doing in hospitals collecting big data and why has it appropriated the name of the world’s most famous nurse?
Florence Nightingale (1820–1910) led the first official team of British military nurses to Turkey during the Crimean War, fought between Britain and Russia (1853–56). She is often known today by her nickname – ‘the Lady with the Lamp’– for the night rounds she made tending to the wounded and sick.
But one of the less celebrated distinctions of Nightingale is that in 1859 she was the first woman to be elected a Fellow of the (later Royal) Statistical Society. She had used statistical analysis to identify relationships between death rates and unhealthy living conditions such as poor sanitation, ventilation and food. She was helped by the pioneering work of William Farr (1807–83), who in 1838 was appointed as ‘Compiler of Abstracts’ (Chief Statistician) to the Registrar General of Births, Marriages and Deaths – the first such post in the world. Farr was a qualified doctor who was largely self-taught in the compilation and analysis of statistics, but he used his position to collect data on causes of death, using categories agreed with the heads of the royal colleges. The 27 categories included such headings as dysentery, typhus, diarrhoea and “suddenly”. In the 1840s, Farr began to issue annual reports that analysed causes of death in relation to age, gender, parish and occupation. By the standards of its time, this was truly what now might be termed ‘big data’.
Saving lives with statistics
Florence Nightingale was, like Farr, a self-taught statistician whose father regarded the study of mathematics as unfeminine and the profession of nursing as degrading, but he was powerless in the face of his daughter’s determination. Her work in Crimea had convinced her that foul living conditions, poor food, inadequate ventilation and polluted water were responsible for seven times more deaths among the soldiers than were injuries sustained in conflict.
Upon returning from that war she began to apply statistical techniques to the health of the military in times of peace and observed that mortality among previously healthy young soldiers living in barracks was double that of the civilian population. She observed that it was criminal to tolerate the living conditions in barracks and that the War Office might as well “take 1,100 men out on Salisbury Plain and shoot them”. She impressed Queen Victoria with her reports on living conditions for soldiers, both during the Crimean War and later in India. The monarch commented “I wish we had her at the War Office”, and summoned Lord Panmure, Minister of War, to Balmoral to listen to the nurse, with Victoria herself hovering over their deliberations. To remove any doubts, the queen wrote: “Lord Panmure will be much gratified and struck with Miss Nightingale.” In turn, Nightingale was so despairing of the innumeracy of Victoria’s university-educated ministers with their degrees in classics that she devised a “Coxcomb diagram” (an early pie chart) to make her points. As a result of her work, death rates in barracks fell by 75 per cent.
Learning from early mistakes
Nurse Nightingale and Dr Farr didn’t get everything right – and the formidable pair almost set back the cause of sanitary reform during the mid-19th century. They were both convinced that the four cholera epidemics that ravaged Great Britain between 1831 and 1866, carrying off almost 40,000 citizens in London alone, were caused by a miasma of foul air rather than polluted water (the true cause). For that reason, they campaigned against the practice of laying sewers beneath houses, believing that any smell that escaped would carry germs into the home.
Fortunately, Nightingale and Farr were powerless against the determination of Sir Joseph Bazalgette (1819–91), who constructed a system of sewers that, to this day, intercepts London’s sewage and conducts it to treatment works in Beckton and Abbey Wood, downstream of the capital. William Farr was eventually convinced, by his own data, that cholera was indeed caused by polluted water, but Nightingale went to her grave in 1910 believing that air was the true culprit.
Google’s project evidently aims to do the same thing as William Farr and Florence Nightingale albeit on a much more ambitious scale, establishing relationships between sociological factors (age, living conditions, diet, exercise) and outcomes such as health and longevity. But in the intervening centuries we have started to gain a much more sophisticated understanding of factors like heredity, and in particular genetics, with the emerging possibility of personalised medicine. These developments have given rise to a number of well-founded fears. The first concerns confidentiality. Do patients in hospitals want information about their genes and lifestyles to be divulged in such a way that individuals can be recognised? Almost certainly not. And is there a possibility that, even if the information is anonymous, certain group characteristics could be used by insurance companies to increase premiums, for example for people who are overweight, diabetic or have certain genetic characteristics? Would that be socially desirable?
- Elizabeth Blackwell: the pioneering ‘first female doctor’, who entered medicine to prove a point
Big data and the NHS
Closer to home, the issues are becoming associated with the United Kingdom’s beloved National Health Service (NHS). The UK has three components that make it well placed to lead the way in the application of patient data to health care and advances in science. The first component is the NHS itself, with cradle to grave health records of almost 60 million citizens of every age, race and condition. Most of these are now held on computers in hospitals, health centres and GP surgeries – a veritable treasury of health and sickness information beyond measure. The second component is the UK’s vibrant software and biotech sectors, which are particularly evident in the ‘Golden Triangle’ of London, Cambridge and Oxford. And the third component is a highly successful pharmaceutical industry, encompassing companies like GSK and AstraZeneca (and a host of start-ups, many of them in the same Golden Triangle). Although many advanced nations have pharmaceutical industries and software/bioscience enterprises – the USA springs to mind –none, in my view, can also add the records and experience of the NHS.
Lessons to be learned from Nightingale?
So why doesn’t the UK recapture the work of the Crimean nurse for itself and launch its own Project Nightingale under the leadership of the NHS? There are several hurdles to be negotiated.
First, not all hospitals and GPs have the same computer systems. These systems don’t all talk to one another very coherently and attempts to unify the system have been met with only partial success. One approach could start with the patient records of a few large hospitals that treat a wide variety of conditions.
Second, the problem arises of the confidentiality of patient data. It is surely not beyond the wit of man to combine data in such a way that personal identities are removed while making key facts like age, gender, and health histories available for research purposes.
And finally, a lead organisation would be needed to pull together the data and make it available in such a way that it helps those who are doing research into illnesses and those, the patients, who can benefit from it. A start has been made. A health technology enterprise called Sensyne Health, founded in 2018, is using artificial intelligence in collaboration with the NHS to detect hidden patterns in anonymised patient data to throw light on causes of disease, just as Florence Nightingale did with her pie charts. And in 2019, NHSXhas been set up under a former Director General of Digital Policy in the Department for Digital, Culture, Media and Sport. One of the first tasks of the group is to make NHS systems compatible.
So perhaps we will be able to reclaim Florence Nightingale for the NHS, even if we do replace her pie charts with DNA sequences, spreadsheets and personalised medicine. As we have seen, Nightingale was particularly interested in sanitation – and personalisation has lately taken a step in the direction of the lavatory in the form of the FitLoo. This results from an unlikely combination of the Massachusetts Institute of Technology, the European Space Agency and some sanitation engineers. It will screen human waste for signs of incipient disease, such as cancer or diabetes, and send a message to the user’s mobile phone or GP. In this way, personalised medicine joins big data – with or without Google – to improve human health. Florence Nightingale (and William Farr) would certainly approve.
Stephen Halliday is the author of An Underground Guide to Sewers: Or Down, Through and Out in Paris, London, New York &c (Thames and Hudson, 2019)