Florence Nightingale: Passionate Statistician
Lynn McDonald 1, “Florence Nightingale: Passionate Statistician.” Journal of Holistic Nursing 16,2 (June 1998): 267-77
Nightingale’s passionate commitment to statistics was based on her faith in a God of order, who created a world that ran by law. God’s laws could be known through research, as a result of which suitable interventions to better the world could be applied. Statistics were a vital component in her holistic approach to health care as a system. They served both to indicate serious problems and to assist in policy making, and then again to monitor the effects of the new policies. She pioneered the use of evaluative statistics and saw reforms achieved as a result of her advocacy.
This paper explores three key aspects selected from Nightingale’s more than forty years of applied statistical work: her adaptation of Quetelet’s methodological foundations, the use of statistics in monitoring public health care systems and her pioneering study of maternal mortality in childbirth.
Florence Nightingale: Passionate Statistician
Nightingale’s passionate commitment to statistics was intimately related to her spirituality. God was a God of order, who created the world and let it run by law. We should not pray to be “delivered from plague, pestilence and famine,” but learn how to intervene in nature to prevent such disasters. Nightingale complained at length at how impractical the medieval mystics were. Spiritual retreats should rather equip one for practical action in the world, into which one returned refreshed and strengthened. Statistics helped one by indicating precisely what to do to second God’s creative work, to be a fellow-worker with God.
Statistics were a vital component of Nightingale’s systemic approach to health care. She advocated health promotion, disease prevention as a better use of resources than treatment on a case-by-case basis of the ill after they succumbed. Her mentor on all things statistical was the Belgian statistician L.A.J. Quetelet (an astronomer, head of Belgium’s central statistical agency and a widely-respected expert on the collection of official statistics and probability theory). From Quetelet she learned that “Administration saves more hospital patients than the best medical science.” That is, of course, administration that was based on the best possible knowledge, namely good statistical data. Sloppy administration was a worse sin than sexual immorality, because its consequences harmed so many more people. Nightingale’s holistic approach to health care may indeed be said to be based on statistics, for it was statistical analysis that taught her the importance of the environment, social and bio-physical, both on susceptibility to disease and treatment outcomes.
Nightingale was a passionate statistician (the expression comes from a chapter title in E.T. Cook’s Life of Florence Nightingale) not least of all because she could see the individuals in quantitative data (Cook, 1913). In a letter to Sir John McNeill, a close collaborator from Crimea on, she described unnecessarily high mortality rates in the army (17-20 per thousand compared with 11 in civilian life) as being “criminal,” comparable to taking 1100 men out upon Salisbury Plain and shooting them (Kopf 1916-17). Similarly, in arguing for a new hospital to be built at Winchester, she compared the death toll from erysipelas (eight out of twenty-four affected, in a hospital of 100-120 beds) as akin to the “hospital massacre” that occurred, for a short period at Scutari, “so that Winchester aspires to rival the most colossal calamity of history in its small way” (Nightingale 1861b). Thus she used quantitative data to support her arguments for reform in hospital construction.
Nightingale had a prodigious respect for the possibility of unintended harmful results of actions, indeed including those of impeccably humanitarian motivation. She frequently gave as an example the establishment of foundling hospitals for abandoned infants, which resulted in increasing the abandonment of illegitimate children (again Quetelet provided the original examples, but she found subsequent instances of her own). Recourse to statistics, in other words, was essential simply to ensure that one did not cause more harm than good in one’s “good deeds.” As the Victorian era was one of enormous reform activity and social experimentation her cautions were most germane.
Nightingale’s commitment to statistical method spanned her entire working life, from her first royal commission on return from the Crimean War in 1856 to her last work on India in the 1890s, when she was unable to send out questionnaires herself to collect data, but she helped in the analysis of available material. The gathering of the best data possible became an integral part of the Nightingale method, used in all her serious projects, what she called her “business.” Where good data were already collected she used them, where not she designed and sent out (or had a government office send out) her own questionnaires. She worked closely with William Farr, Britain’s leading social statistician, to improve the routine collection of social data, precisely because she believed in their importance for social planning purposes, notably in health care, housing and education.
It is telling that one of the last projects of her life, 1891, was the (unsuccessful) attempt to establish a chair or readership in “social physics and their practical application” at Oxford University, as a means of training future civil servants how to use statistics in social planning (Oxford was the chief training grounds for the public service) (Pearson, 1924).
The following seven areas give an indication of the depth and breadth of Nightingale’s statistical achievements.
- The collection and presentation of data for the Royal Commission on the Sanitary State of the Army in the East. Nightingale was instrumental in the establishment of the royal commission, briefing witnesses etc. She gave a succinct analysis of the data, pioneering the use of “coxcombs” (area charts) for the graphical presentation of mortality data. Government statistics would never be the same as a result, for her example of providing easy-to-visualize bar charts and pie charts was taken up in subsequent routine publications of Census and other data (Cohen, 1984; Nightingale 1858a, 1858b and 1859).
- The comparative analysis of mortality rates of the army at home with the civilian population. This resulted in substantial reform in barracks construction, diet and the provision of medical and nursing services (Nightingale, 1858b).
- The collection and analysis of data for the Royal Commission on the Sanitary State of the Army in India. Nightingale designed the questionnaires that were sent to all stations and analyzed the data returned (Nightingale, 1863a). Her work here led ultimately to massive reforms in the administration of India, but only after decades of agitation.
- Reform of hospital statistics and the establishment of uniform criteria for reporting disease. Nightingale, with the collaboration of Quetelet and Farr, obtained the support of the International Statistical Congress, held in London 1860, which led to the first attempts to standardize statistics on disease and its treatment (Keith, 1988; Nightingale, 1860; Stolley and Lasky, 1995).
- Fighting the Contagious Diseases Acts, for the compulsory inspection/treatment of prostitutes. Nightingale recognized the high toll venereal diseases took in the armed forces, but she used statistics in a briefing note to show that they were not reduced where compulsory inspection and treatment of suspected prostitutes, were in use. She was not successful in preventing the legislation, first passed in 1864 and which was not repealed until 1886 after a long struggle led by Josephine Butler. Nightingale’s statistical arguments were used throughout the campaign, as well as her and others’ arguments based on religious principles, civil liberties etc.
- The collection of data on mortality and morbidity of native children in colonial schools. Nightingale devised the questionnaire that was sent out by the Colonial Office. She analyzed the collected data and made recommendations, which resulted in improved administration (Nightingale, 1863b).
- The analysis of data on surgical outcomes. Nightingale used a paper to the 1863 International Statistical Congress in Berlin as the means to provide an early example of evaluative statistics.
All the above-mentioned projects will be related in my Collected Works of Florence Nightingale currently in preparation.
Selected for fuller discussion here are her work on Quetelet, which elucidates her philosophical approach to statistics, the use of statistics in monitoring the public health care system and the use of statistics in her pioneering study of mortality of women in childbirth (with related recommendations for midwifery practice).
Nightingale and Quetelet
Nightingale carefully worked her way through Quetelet’s major publications on statistics. Quetelet was well known in Britain, where he had assisted in the founding of the statistics section of the British Association for the Advancement of Science, 1834. He gave Nightingale a copy of his two-volume Physique sociale (Quetelet, 1869), which she extensively annotated and translated, and then drew on for an essay “In Memoriam” she wrote on his death in 1874 (Diamond and Stone, 1981). She knew a number of his earlier works but it seems that she did not know the English translation of the shorter (1835), first edition, published as A Treatise on Man and the development of his Faculties (Quetelet, 1842).
It was Quetelet who solved for her the great dilemma of how to reconcile an orderly universe, run by law, and an active program of intervention in it. Social research became the investigation of God’s laws, conceptually the same for the social world as for the bio-physical. In opening her “In Memoriam” essay she referred to the “application of his discoveries to explaining the plan of God in teaching us by these results the laws by which our moral progress is to be attained, or rather explaining the road we must take if we are to discover the laws of God’s government of His moral world.” Quetelet was no less than the “discoverer” of the science “upon which alone social and political philosophy” could be based, which should not be limited to the administrative or legislative domain but “be the interpreter of all theodikè, all divine government and its laws, embracing the smallest and most accidental to the greatest and most universal actions and phenomena of our moral and physical life.” It was “the germ of a vast reform to be made in the world’s morality, not by confessing and bewailing our desperate wickedness but by practically growing the new moral world out of the discovery of what the laws are.” This would permit us, no less, than to exchange original sin for original goodness, by “discovering the laws of God’s vast scheme of universal order.”
Moral laws of God can be found by induction as by physical laws. Indeed God’s moral, social and physical laws act and reaction one another. By Quetelet’s method moral [social] laws can be stated in exact numerical results.
A law does not “govern” or “subordinate,” does not compel people to commit crime or suicide. On the contrary, it put means into our hands to prevent them, if we did but observe and use these means. It simply reduces to calculation observed facts; this is all that a law means…The causes influencing the social system are to be recognized and modified. From the past we may predict the future.
Monitoring the Health Care System
Nightingale’s health care system had a strong social/environmental component. The key was to build health by giving infants/children a good start, which required good nutrition, safe water and adequate housing (all rare for the great majority of the poor at that time in Britain). Onto that base measures would be added for treating ill people, both at home with nurse/health visitors, and in hospitals, with trained nurses of course. There would be a variety of institutions and asylums for convalescent care (in the country) and for the care of the chronically ill, disabled, mentally ill and handicapped. All of these institutions would collect and centrally report their statistics (mortality, admissions, duration of stay etc.) Comparative analysis of success and failure would permit wiser administrative decisions to be made in the future. Nightingale was a “tax and spend” Democrat in contemporary political terms, but she always wanted good value in spending the taxpayer’s money. With uniform reporting procedures comparisons could be made not only regionally but internationally, permitting the recognition of particularly good (or bad) methods of treatment (Nightingale, 1861a).
As well as the institutional data to be collected Nightingale envisaged the routine collection and analysis of benchmark health data. As early as the preparation for the 1861 Census she proposed the addition of questions on health status and housing for precisely this purpose. She was not successful at that time; health surveys now play the role Nightingale projected for the Census.
A good health care system extended from the international level to the local, in the case of Britain with county sanitary committees to investigate local conditions and monitor progress when reforms were instituted. Nightingale herself analyzed local data in Buckinghamshire, the area where her sister and brother-in-law had their country home and which she visited frequently (Nightingale, 1894a). Notes on Nursing itself includes attention to occupational health, not least of all for nurses and doctors. Nightingale as well pointed out that “the places where poor dressmakers, tailors, letter-press printers and other similar trades have to work for their living are generally in a worse sanitary condition than any other portion of our worst towns” (Nightingale, 1860b).
Midwifery for Nightingale would be a key part of regular practice of trained nurses. Provisions for training nurse midwives was indeed the second project undertaken by the Nightingale Fund after the Crimean War (the first of course being the institution of secular training for nurses at St. Thomas’s Hospital). Yet Nightingale soon closed the lying-in institution, associated with King’s College Hospital, when statistics showed an unacceptably high rate of puerperal fever—26 deaths out of 780 women, or 33.3 per 1000 compared with a national rate of 5.1 (Nightingale, 1871). Her book, Introductory Notes on Lying-in Institutions, 1871, is a remarkable document for the lucidity of its arguments for its basic principles (childbirth being a natural phenomenon which should have minimal medical involvement) and the practicality of the reform measures arrived at. Because basic data on mortality in childbirth were lacking Nightingale had to begin with the collection of statistics, by sending out questionnaires to a variety of institutions, such as hospitals and workhouse infirmaries.
Conceptually she had to work through the question of what was “the real normal death-rate,” or the rock-bottom mortality that could not be avoided even with the best care. This was the work on which she most relied on Quetelet’s methodology. In setting up the training institution in the first place she had been convinced that good training required an institutional setting. When high rates of mortality became evident (although they were still lower than in some Continental hospitals) she had to consider that training midwives at home deliveries might be the only acceptable solution. Throughout, her judgments were based on statistical analysis:
With all their defects, midwifery statistics point to one truth, namely that there is a large amount of preventible mortality in midwifery practice and that, as a general rule, the mortality is far, far greater in lying-in hospitals than among women lying-in at home (Nightingale, 1871).
Nightingale also considered the role of secondary influences like the age of the mother, number of pregnancies and duration of labour, general health and stamina, and social conditions such as the patient’s social class. It might have been expected that women giving birth in workhouse infirmaries, because of their poor health generally, would have had higher mortality rates than women giving birth in a specialized midwifery institution that charged fees. Yet the workhouse mortality rates were lower. Nightingale concluded that the effect of the institution was greater than that of social conditions. Midwifery wards at general hospitals should be closed and home births encouraged. Training institutions for midwives should be small enough so that conditions could emulate those of home births. Training institutions could be justified at all only if they were as safe for birthing mothers as home deliveries. Otherwise this was to “ensure killing a certain number of mothers for the sake of training a certain number of midwives” (Nightingale, 1871).
Nightingale’s data also showed higher mortality rates where the women were in contact with doctors and medical students. She concluded that there should be as little medical involvement as possible, and medical students should be entirely banned (Nightingale, 1871). Small wonder that Introductory Notes on Lying-in Institutions got a nasty (anonymous) review in the British Medical Journal. Laced with sexist sarcasm (including references to the “kind womanly heart” of the “authoress” and her “purely sentimental” arguments), yet the review rages most against her statistical analysis, “ingenuously urged as if it were a logical thunderbolt.”
“On considering these figures” (says she) “the first impression they convey is not that either the Registrar-General or LeFort [a medical statistician] is wrong. But it is a very painful impression of another kind altogether. One feels disposed to ask whether it can be true that, in the hands of educated accoucheurs, the inevitable fate of women undergoing, not a diseased, but an entirely natural condition, at home, is that one out of every 128 must die? If the facts are correct, then one cannot help feeling that they present a very strong prima facie case for inquiring, with the view of devising a remedy for such a state of things.”
Nightingale’s proposals for remedy, including improved standards for hospital construction, are qualified as “hints” which the book “throws out.”
Introductory Notes on Lying-in Institutions are as much imbued with a holistic philosophy as Notes on Nursing, yet they have effectively disappeared from the literature. They show how much Nightingale was a pioneer in midwifery, instituting a training institution for nurse midwives, opposing the medicalization of a natural process and, not least, insisting that all measures be rigorously scrutinized with the use of quantitative data. She herself wanted to do more, statistically, on the issue, notably to collect better data on mortality from home births, but illness prevented her.
Nightingale had the satisfaction of seeing many of her reforms realized, never as fast or as completely as she would have hoped (and much more slowly for reforms in India), but enormous changes were made in sanitary reform and the provision of health care services. The administrative reforms arising from her first royal commission were promptly applied not only within the British Army but in the Northern United States for use in the Civil War. Her statistical achievements were recognized by the leaders of the profession. Her colleague William Farr, himself later a president of the organization, in 1858 proposed her for membership in the London Statistical Society (later the Royal Statistical Society). In 1874 the American Statistical Association elected her an honorary member. Farr not only helped her with her projects over decades, but took advantage of her expertise to solve problems at the central statistical bureau where he worked, the Registrar-General’s Office. Unpublished correspondence shows him thanking her for advice that saved him from making a professional blunder. Oxford University finally honored her in 1997! by instituting an annual lecture on statistics in her name.
Nightingale never wavered on these great principles of a holistic framework, statistically tested, for example addressing issues of rural health as late as 1894 (Nightingale, 1894a and b). An unpublished note still later exclaims: “Oh teach health, teach health, teach health, to rich and poor, to educated and, if there be any uneducated, oh teach it all the more: to men—to women especially—to mothers, to young mothers especially…for the health of their children comes before Greek and grammar.”
Anonymous review. 1871. British Medical Journal. Nov. 11:559.
Cohen, I. Bernard 1984. “Florence Nightingale.” Scientific American. 246:128-33, 136-37.
Cook, E.T. 1913. Life of Florence Nightingale. 2 vols. London: Macmillan.
Diamond, M. and Stone, M. 1981. “Nightingale on Quetelet.” Journal of the Royal Statistical Society. (A) 144. Pt. 1 66-79, Pt 2 176-213, Pt 3 332-51.
Hawkins, F. 1829. Elements of Medical Statistics. London: Longman.
Keith, J. 1988. “Florence Nightingale: Statistician and Consultant Epidemiologist.” International Nursing Review. 35 (5):147-50.
Kopf, E. 1916-17. “Florence Nightingale as Statistician.” Publications of the American Statistical Association. 15:388-404.
McDonald, L. 1993. Early Origins of the Social Sciences. Montreal: McGill-Queen’s University Press.
——-1994. Women Founders of the Social Sciences. Ottawa: Carleton University Press.
——-ed., 1998. Women Theorists on Society and Politics. Waterloo, ON: Wilfrid Laurier University Press.
——-ed., 2001-2012. Collected Works of Florence Nightingale. Waterloo ON: Wilfrid Laurier University Press.
Nightingale, F. 1858a Notes on Matters affecting the Health, Efficiency and Hospital Administration of the British Army. London: Harrison.
——-1858b. Mortality of the British Army at home, at home and abroad, and during the Russian War, as compared with the Mortality of the Civil Population in England. re-printed from the Report of the Royal Commission on the Sanitary State of the Army. London: Harrison.
——-1859. A Contribution to the Sanitary History of the British Army during the late war with Russia. London: John Parker.
——-1860a. Notes on Nursing, ed. Victor Skretkowicz. London: Baillière Tindall 1996.
——-1860b. “Hospital Statistics.” London: Proceedings of the Second Section of the International Statistical Congress.
——-1861a. “Hospital Statistics and Hospital Plans.” Transactions of the National Association for the Promotion of Social Science. 554-60.
——-1861b. Letter to Sir Harry Verney 12 September, British Library Nightingale Manuscripts, Add Mss 45791.
——-1862. “Note on the Supposed Protection afforded against Venereal Disease by recognizing Prostitution and putting it under Police Regulation.” London (unpublished briefing note).
——-1863a. “Observations by Miss Nightingale on the Evidence contained in Stational Returns,” Report of the Royal Commission on the Sanitary State of the Army in India. London: Stanford 347-70.
——-1863b. Sanitary Statistics of Native Colonial Schools and Hospitals. London: Transactions of the National Association for the Promotion of Social Science 475-88.
——-1863c. “Proposal for Improved Statistics of Surgical Operations.” International Statistical Congress (Berlin), reprinted in the 3d. ed. of Notes on Hospitals.
——-1871. Introductory Notes on Lying-in Institutions. London: Longmans Green. reprinted in Florence Nightingale on Hospital Reform, ed. C. Rosenberg. New York: Garland 1989.
——-1894a. “Health and Local Government.” Aylesbury.
——-1894b. “Rural Hygiene.” Official Report of the Central Conference of Women Workers. Leeds.
Pearson, K., ed. 1924. 3 vols. Life, Letters and Labours of Francis Galton. Cambridge: Cambridge University Press 2:416-24.
Quetelet, L. 1842. A Treatise on Man, trans. R. Knox. Scholars Facsimile reprint 1969.
——-1869. Physique sociale. 2 vols. Brussels: Muquardt.
Spiegelhalter, D. 1997. “Surgical audit: statistical lessons from Nightingale and Codman.” Cambridge: MRC Biostatistics Unit.
Stolley, P. and Lasky, T. 1995. Investigating Disease Patterns: the Science of Epidemiology. New York: Scientific American Library 39-43.
1 Dr Lynn McDonald (university professor emerita) Dept. of Sociology and Anthropology University of Guelph Guelph ON N1G 2W1 Canada