Epidemiology and Health-Care Reform

EPIDEMIOLOGY and Health Care Reform The National Health Survey of 1935-1936

| George Weisz, PhD

The National Health Survey undertaken in 1935 and 1936 was Irving Fisher used them to great received wide popular support the largest morbidity survey until that time. It was also the ﬁ rst effect early in the century to and was organized with industrial national survey to focus on chronic disease and disability. The launch a debate about ensuring efficiency.3 About 2 800 000 decision to conduct a survey of this magnitude was part of the national vitality, a debate that people in 19 states were sur- larger strategy to reform health care in the United States. The deeply influenced many of those veyed regarding illnesses suffered focus on morbidity allowed reformers to argue that the health status participating in the events dis- during the preceding year. The of Americans was poor, despite falling mortality rates that suggested cussed here.1 However, during NHS may well have been the the opposite. The focus on chronic disease morbidity proved to be an especially effective way of demonstrating the poor health of the the interwar years, mortality data first morbidity survey carried population and the strong links between poverty and illness. The were fiercely contested and con- out by welfare relief recipients. survey, undertaken by a small group of reform-minded epidemiologists stituted a double-edged sword for Around 6000 unemployed work- led by Edgar Sydenstricker, was made possible by the close interaction those seeking change. American ers were paid by the Works Prog- during the Depression of agencies and actors in the public health health reformers used mortality ress Administration (WPA) to and social welfare sectors, a collaboration which produced new statistics but also created new conduct the survey at an initial ways of thinking about disease burdens. (Am J Public Health. kinds of data and, in the process, cost of 3.5 million dollars. 2011;101:438-447. doi:10.2105/AJPH.2010.196519) took the field of epidemiology in Although the survey was not innovative directions. I describe the first one to verify interviews one particularly salient effort. by contacting doctors, it may have THESE DAYS, REFORMERS The National Health Survey been the first to pay physicians for seeking to show the deficiencies (NHS) undertaken in 1935 and such work. Data from business and of American health care point to 1936 was technically not the first industrial health plans were also embarrassing international statis- national morbidity survey under- collected. This survey was cer- tics indicating that the United taken in the United States. John tainly the first morbidity survey States pays far more for health Shaw Billings used the censuses using a sizable public relations care than other nations and gets of 1880 and 1890 to collect data apparatus. Its planning was a public inferior results. Early in the 20th about sickness incidence, but his event in which cities lobbied century, however, such statistics controversial effort was largely actively to be included. National were harder to come by. The unsuccessful because of the media announced the start of the tradition of using mortality statis- refusal of both doctors and lay- survey, and a local media cam- tics to mobilize investment in men to convey private health paign was launched as soon as public health is an old one. In information to canvassers.2 In the survey hit a city. An analyst the United States, the economist contrast, the well-financed NHS reported on media reactions.

438 | Public Health Then and Now | Peer Reviewed | Weisz American Journal of Public Health | March 2011, Vol 101, No. 3 PUBLIC HEALTH THEN AND NOW

Even before the survey was com- epidemiologist of his generation. and Falk as research associate. pleted, results were leaked by A researcher for the US Public Sydenstricker’s brother-in-law, influential public figures. The Health Service (USPHS) since George St. John Perrott, a mining publicity barrage intensified after 1915 (including a two-year leave chemist who had lost his job dur- publication and set new standards to work for the League of ing the Depression and gone to of media attention. Not least, this Nations), he had participated in work for Sydenstricker, was listed survey was the first national sur- Joseph Goldberger’s famous pella- as a consultant for this study. The vey designed, in the words of H. S. gra investigations and gone on to committee as a whole included Cumming, the about-to-retire sur- pioneering work in the famous other individuals who would play geon general, “to study the extent Hagerstown morbidity studies of a key role in the NHS. Harry and nature of disability in the the 1920s. In 1928, he became Hopkins, responsible for New Deal general population, with special scientific director of the Milbank Welfare policies, was among the reference to chronic disease and Memorial Fund while remaining a five cabinet-level members who physical impairment.”4 consultant to the USPHS. His signed the letter submitting the Why was such an unprece- work exemplified growing collab- committee’s report, and Josephine dented study undertaken at this oration between private founda- Roche, assistant secretary of the time? I suggest that it was part of tions and governmental agencies. treasury in charge of public health, the larger strategy to reform Among his many activities was was a member of the technical health care in the United States membership on the Committee board. Michael Davis, another that had been going on since the on the Costs of Medical Care veteran of the CCMC, was on 1920s. The answer to a second (CCMC), which began its work in the hospital advisory board.9 question—why a morbidity 1927 with funding from eight Sydenstricker and Falk wrote a study?—has to do with the way private foundations. Charles report calling for provision of a mortality statistics could be inter- Winslow of Yale, who was at the health program, including health preted and the need to counter center of a group of progressive insurance. This proposal was not that interpretation. A third ques- members of the commission that included in the committee’s final tion—why the focus on chronic included Sydenstricker, arranged report to the president. disease and disability?—is the for Falk, a former student, to be Despite this setback, health most difficult to answer. Although appointed chief researcher for the care reform remained very much concern with the problem of committee.5 Falk directed a mor- alive. In 1936, Sydenstricker and Edgar Sydenstricker. chronic disease had intensified bidity study of nearly 9000 fami- Falk collaborated on another Source. Sterling Library. Yale University. during the previous decade, this lies. Enumerators visited families governmental commission, the issue was not yet a major one for at two-month intervals over a full Interdepartmental Committee to most health reformers, including year and noted occurrence and Coordinate Health and Welfare, those initiating the survey. This duration of illness, as well as use which brought together leading orientation gradually emerged and costs of medical services.6 figures from the public health and because of the coming together Soon after the reports of the welfare domains. The overex- during the Depression of reform- committee were completed, tended Sydenstricker turned oriented figures in the public Sydenstricker and Falk began down an offer to become executive health and social welfare sectors. working on a book about health director and died later that year. New ways of thinking epidemio- insurance systems.7 The project A technical committee was formed logically about disease burdens was abandoned in June 1934, in 1937 and included Martha Eliot grew out of this interaction. however, when both were named of the Children’s Bureau; Falk, to the Council on Economic Secu- then working for the Social Secu- THE SURVEY AND ITS rity, which designed the Social rity Board; and St. John Perrott, ORIGINS Security Act of the following year. Joseph Mountin, and Clifford E. The committee took an expansive Waller from the USPHS. This The two men who initiated the approach to its mandate to “pro- committee produced the National survey, Edgar Sydenstricker and vide at once security against Health Program, which served as Isidore S. Falk, are familiar to several of the great disturbing fac- the basis of a National Health historians of American health tors in life”8 and included a study Conference held in 1938. care reform. Sydenstricker of health insurance for which Even before the formation of was undoubtedly the leading Sydenstricker served as director the Interdepartmental Committee,

March 2011, Vol 101, No. 3 | American Journal of Public Health Weisz | Peer Reviewed | Public Health Then and Now | 439 PUBLIC HEALTH THEN AND NOW

the federal government announced insurance.15 Falk suggested that the National Health Program and a new program of job creation, he stepped out of the picture organized the National Health the WPA. Sydenstricker and because he or others believed Conference. The results of the Falk quickly submitted a pro- that his active participation NHS served the needs of this posal for a national health survey would provoke the opposition reform-oriented committee, carried out by WPA workers. of the American Medical Asso- whose argument for health insur- Falk later claimed that he had ciation, for whose leaders he ance and health reform generally played a key role in convincing the embodied socialized medicine. was based on three practical New Deal’s dominant figure on But there may also have been arguments: (1) the health situa- welfare matters, Harry Hopkins— some turf warfare over control of tion of Americans was very bad who supervised the Federal the survey between the USPHS in the wake of the Depression, Emergency Relief Administra- and the Social Security Board for (2) the poor suffered disproportion, the Civil Works Administra- which Falk now worked.16 The tionately from illness, and (3) tion, as well as the WPA—that former prevailed, and manage- the poor had much less access such a survey would be a good ment of the survey fell to Perrott to health care than did the more investment.10 Although the claim and Selwyn Collins, both young affluent. As a consequence of is not implausible, it is worth veterans of Sydenstricker’s earlier inadequate health care, more Depression studies. The original people than necessary were Even before the formation of the Inter- plan for a National Health Inven- sicker longer than necessary tory was extremely ambitious: at enormous social and eco- departmental Committee, the federal government nomic cost.

announced a new program of job creation, First, a house-to-house canvass The problem with this set of “ in ninety-five communities, lo- arguments was that mortality the WPA. Sydenstricker and Falk quickly cated in nineteen states, repre- rates had been falling for submitted a proposal for a national health survey senting the various geographic divisions of the country; second, decades and continued to fall carried out by WPA workers. an inventory of public health during the Depression. This and medical facilities throughout decline suggested to some that the nation; third, a study of mor- 18 bidity and mortality according to no health crisis existed. To noting that Hopkins had a long occupation based upon the re- counteract this perception, career in medical” and welfare cords of sick-benefit associations reformers had to demonstrate in industry; and fourth, commu- administration in New York City nication with every physician at- that (1) national health status before joining the New Deal and tending a case of illness reported could be best understood was deeply concerned with in the house-to-house canvass through morbidity rather than for the purpose of obtaining his health issues; by 1936, he had technical knowledge of the na- mortality rates, (2) morbidity sur- successfully committed about ture of the disabling illness.17 veys presented a somber national $29 million of WPA money for health picture, and (3) morbidity 650 studies in a wide variety Although a great deal of data surveys demonstrated that the of domains.11 was collected, much remained poor, lacking adequate health It is clear, however, that Falk unanalyzed by the time the study care, also suffered disproportion- played a key role in developing was shut down in 1941. It was ately from illnesses that could be the survey. Sydenstricker pro- the first of the tasks in the pre- cured or mitigated. With this set duced a first draft12 and Falk ceding quotation, the morbidity of suppositions, a moral, eco- produced another more elabo- study, that became the core of nomic, and medical argument for rate draft after consultation the NHS’s contribution to US health reform could be made, with Michael Davis and Mountin.13 health policy. including some form of insurance Although the organization of the Although there was no official or public health care for the less survey fell to others, Falk pro- link between the NHS and the affluent. But morbidity studies duced a number of preliminary Interdepartmental Committee, entailed their own problems. tables of results and presented there were many personal ones, Opponents could argue that a plan of analysis.14 Perrott had including Roche’s overall leader- many of the illnesses reported no doubt that Falk viewed the ship of both and Perrott’s move were trivial. More important, survey as part of his wider strat- from the NHS to the committee such surveys provided only weak egy to promote national health in 1937, where he helped write evidence for the link between

440 | Public Health Then and Now | Peer Reviewed | Weisz American Journal of Public Health | March 2011, Vol 101, No. 3 PUBLIC HEALTH THEN AND NOW

poverty and illness. The work of correlated with poverty. He could Sydenstricker believed these Sydenstricker and his associates, draw on numerous mortality results. They explained them supplemented by unemployment studies demonstrating this away by suggesting that investi- and welfare surveys occurring point.20 But mortality rates were gators tended to subconsciously simultaneously, overcame these a double-edged sword that also record as illnesses those condi- problems by gradually focusing suggested that things were tions that entailed medical costs on chronic illnesses and longer- improving. It was harder to make and ignored those that did not.25 term disabilities. this link through morbidity stud- Several years later, Sydenstricker’s ies. Sydenstricker could point to a collaborator Collins reanalyzed MORBIDITY STUDIES number of industrial disability the CCMC data without men- BEFORE THE NATIONAL studies, as well as his own early tioning its findings about the link HEALTH SURVEY work among South Carolina cot- between economic status and ton-mill workers, that supported morbidity rates.26 Sydenstricker was throughout a link between disease and pov- Most of the data for the CCMC his career concerned with mor- erty.21 The poverty–morbidity studies were collected in 1926. bidity, which he considered link in his Hagerstown studies, The Depression created new more indicative of health status however, was not strong: needs and opportunities for and medical need than mortality. morbidity studies. In 1933, He experimented with different These differences are not of the Sydenstricker began directing a kinds of data: industrial disability same magnitude as those found survey on behalf of the Milbank previously for infant mortality, data, the results of insurance tuberculosis, or pellagra, for ex- Fund and the USPHS that was medical examinations, and, espe- ample. . . . A somewhat detailed connected to an international study cially, information collected dur- analysis of the data, however, by the League of Nations. This revealed the facts that the asso- ing the morbidity surveys for ciation of illness with poor eco- consisted of a sickness and mor- which he became famous. His nomic status 1) appeared for tality survey that canvassed overriding interest was not in certain causes only, and 2) was about 12 000 wage-earning fami- indicated in adult life and not in kinds of diseases but rather in childhood or adolescence.22 lies in 10 localities, including distinguishing illness incidence eight large cities, a group of coal- among various groups by age, On several occasions, Syden- mining communities in West Vir- sex, ethnicity, and, increasingly, stricker speculated on the reasons ginia, and a group of cotton-mill occupation and income. The for the apparent weakness of this villages in South Carolina. The most significant finding of his link,23 but he never doubted that families were not meant to be and other morbidity studies was his data seriously underestimated representative; mainly poor dis- that morbidity patterns differed this correlation and that poverty tricts (although not slums) were from mortality patterns. In the strongly affected health by shap- canvassed, because the goal was latter case, “general diseases” ing direct causal factors such as to gauge the effect of the Depres- like cancer and cardiovascular nutrition, sanitary conditions, and sion on families that had been disease were taking a leading overcrowding. He would spend self-supporting before the down- role, whereas respiratory and much of the first half of the turn. Wealthy neighborhoods infectious diseases predominated 1930s trying to strengthen the were excluded on the assumption in the former.19 This distinction case for the relationship between that living standards of inhabit- likely explains why he was not poverty and disease. ants had not dropped enough to among those who warned about The Report by the Committee affect health. “Colored” neighbor- a “chronic disease” problem in for the Costs of Medical Care hoods were not canvassed to the 1920s and early 1930s. The convincingly demonstrated that avoid the effect of race.27 implications of his findings great differences in unmet need The survey had an explicitly would be spelled out during the for medical care were linked political rationale. Death rates Depression; despite falling mor- to economic status.24 Much to and reports of communicable dis- tality rates, illness was far more everyone’s surprise, however, the ease had not, it was admitted, widespread than generally report suggested that the more risen during the worst years of believed. affluent suffered slightly more the Depression. According to Sydenstricker was equally con- illness than did the poor. Neither Perrott and Collins, “The com- vinced that disease was closely the authors of the study nor fortable conclusion is drawn by

March 2011, Vol 101, No. 3 | American Journal of Public Health Weisz | Peer Reviewed | Public Health Then and Now | 441 PUBLIC HEALTH THEN AND NOW

what was emphasized instead was that families on public or private relief experienced more illness than any other group. And within this welfare population, the highest rates were among wage earners in 1929 that only later went on relief.31 This finding suggested to the authors “that the highest illness rates were observed among those who had suffered the greatest change in standard of living.”32 Aside from showing that the poor, who had the greatest need for health care, had the least access to it, the overrepresenta- tion of the relief population among the ill suggested that it might make good economic sense to increase the availability of medical care and even to find ways to raise living standards to more healthful levels to reduce welfare spending. In September 1935, Roche From left to right: Assistant Surgeon many that the physical well-being The preceding analysis was made direct political use of this General Lewis Thompson, Selwyn of the American people not only based on “disabling” illnesses, survey. In a New York Times article, Collins, Josephine Roche and George St. John Perrott examining survey has not suffered but, in view of conditions that prevented people she insisted that declining mortal- data. the continued low death rate, from working or otherwise func- ity rates were very poor indicators Source. National Library of Medicine. may have been benefited by the tioning normally for some of the nation’s physical condi- economic catastrophe. Such a period30; this choice emphasized tion and that illness rates were a conclusion, based upon mortality more serious conditions and far better gauge. These rates statistics alone, is open to ques- may also have been an attempt demonstrated that the Depres- tion.”28 Morbidity studies, it was to avoid repeating the results of sion had a serious effect on the claimed, constituted a more reli- the CCMC survey, which were “rate of acute and chronic dis- able indicator of health status based on all reported illness. eases and serious physical and medical need. Perrott and The rate of disabling disease impairments among families on Collins reported that with the among families classified as poor relief rolls.” She specifically cited exception of only a few localities, was 23% higher by per capita Sydenstricker’s study as having income and 30% higher by total found “distressing conditions.”33 the disabling illness rate of fami- family income than it was in After presenting a few statistics lies having no employed workers is consistently higher in each the grouping classified as com- and emphasizing that sickness city than that of families having fortable. Illnesses considered among the “new poor” was part-time or full-time workers. largely chronic that began prior most prevalent, she concluded Inasmuch as most of the families having no employed work- to the Depression showed even in language that Irving Fisher ers in 1932 had one or more higher rates of excess among the had popularized: “Obviously employed workers in 1929, poor than among the comfort- facts such as these reveal not these data are striking evidence of the association between a rel- able—50% by per capita income only conditions of human suf- atively high rate of disabling ill- and 80% by total family income. fering and wretchedness but ness and loss of employment This fact, which might have economic waste, and challenge during the depression, with ac- companying loss of income and directed attention to the chroni- us to a swift-moving program reduced standard of living.29 cally ill, was largely ignored; of conservation of one of our

442 | Public Health Then and Now | Peer Reviewed | Weisz American Journal of Public Health | March 2011, Vol 101, No. 3 PUBLIC HEALTH THEN AND NOW

most valued national resources— this discussion. The first was the to deal with this problem, and the health and vitality of our movement to expand the scope they had a major impact on hos- people.”34 of public health—beyond infec- pital policy in New York City. As Chronic diseases played almost tious diseases to include degener- a result of the New Deal’s efforts no role in these studies. This fact ative diseases and mental health, to cope with the Depression, the is surprising because chronic ill- and beyond prevention to issue of welfare chronicity moved ness during these years became include diagnostic and curative beyond such local arenas and increasingly framed as a major services. In the latter case, there took on national dimensions. By health and welfare problem in was some ambiguity about the mid-1930s, welfare surveys the works of Ernst Boas and whether public health agencies merged with morbidity surveys to George Bigelow (on behalf of the should provide such services or generate new sorts of questions. Massachusetts Department of coordinate the work of private While Sydenstricker’s group Health)35 as well as numerous practitioners, but the principle of was doing its Depression study, lesser-known figures. Concern public leadership was clear. national welfare institutions were with cancer in particular was Charles E. Winslow of Yale Uni- collecting data for their own pur- spawning an impressive institu- versity was the leading spokes- poses, and these were available tional apparatus and consider- man for this position.37 Some to Sydenstricker’s team at the able public support, which was successes were achieved during USPHS. In 1936, Perrott and translated into political pressure the 1920s, notably the creation H. C. Griffin published an article to create special hospitals and in 1926 of a state cancer hospi- based on a survey undertaken in research institutions. Much of this tal managed by the Massachu- 1934 by the Federal Emergency interest was based on mortality setts Department of Health and Relief Administration—also under statistics that showed rising mor- the endorsement the following the authority of Hopkins—of the tality for cancers, cardiovascular year by the American Public occupational characteristics of diseases, and other chronic con- Health Association of public more than 165 000 relief fami- ditions. The data were highly action for the diagnosis and care lies in 79 cities. This study was controversial, however, because of cancer.38 But it remained a in some ways similar to earlier of radical changes in disease cat- minority opinion until the local welfare surveys but done egories, diagnostic procedures, Depression and New Deal made on an immeasurably greater recourse to the health care sys- it the official position of the scale. This survey defined serious tem, and the way death certifi- national public health leadership. disability as “physical or mental cates were filled out. None of the The second source involved handicaps of a serious and per- principals involved in the survey the recognition by numerous manent nature that impeded abil- had devoted much time to local welfare agencies from the ity to work.”40 These disabilities chronic disease or invalidity. early 1920s that a large propor- could include conditions like Sydenstricker mentioned it only tion of the indigent welfare popu- infantile paralysis, loss of limb, in passing in his wide-ranging lation was in fact sick, infirm, or mental defect or nervous condi- book of 1933, Health and Envi- disabled and that little was being tion, or diseases like tuberculo- ronment. The only discussion of it done to cure or rehabilitate them, sis, heart disease, or epilepsy. in the unpublished report for the which would have allowed at Twenty-one percent of those Committee on Economic Security least some of them to work and older than 16 years reported was a paragraph in the conclu- get off the welfare rolls. Studies some handicap, similar to results sion noting that little was known documenting this situation were of a separate study based on about the incidence of perma- produced in cities like Philadel- medical examinations in Chi- nent disability and recommend- phia, Boston, and New York as cago.41 Most serious by far ing “that provision should be well as the state of New Jersey.39 were orthopedic problems made for the further study of the Characteristically, they quantified (37%), followed by heart and occurrence of permanent disabil- cases of chronic morbidity and circulatory issues (33.2%), rheu- ity and of measures to furnish disability within local welfare matism (20.2%), and senility protection against this risk.”36 populations and found them to (20.9%). Here was an old wel- The emergence of chronic dis- be high. Ernst Boas’s numerous fare concern—chronic disease ease as a social problem had sev- articles and his 1929 book were and disabilities among the relief eral sources. Two are central to influential in suggesting strategies population—now recorded by a

March 2011, Vol 101, No. 3 | American Journal of Public Health Weisz | Peer Reviewed | Public Health Then and Now | 443 PUBLIC HEALTH THEN AND NOW

diseases increases regularly would thoroughly analyze dura- from the lowest rate in profes- 47 sional, proprietary, and clerical tion. Somewhat inexplicably, classes to the highest among no article on this subject ever 43 unskilled laborers; appeared, and neither the Hagerstown study nor his later Point A reinforced the link Depression studies dealt with dis- between disability and chronic ease duration. One might plausibly disease on one hand and welfare surmise that such information spending on the other. Point B was not considered reliable introduced a relatively new in retrospective accounts to notion: chronic diseases were surveyors.48 not just linked to welfare; like The CCMC survey also col- diseases in general, they were lected such data, which the far more prevalent among the report’s authors did not bother to poor than among the comfort- discuss. In his 1936 article, how- able, with increases directly pro- ever, Perrott reanalyzed the old portional to income levels. CCMC morbidity data. Viewed Original table maquette from Depression Survey, 1934-1935, reproduced in In an article published that by annual case rates, respiratory several articles. same year, Perrott made another diseases including tuberculosis Source. Sterling Library, Yale University. intellectual leap by introducing predominated, with degenerative “days of disability.”44 Determining diseases being not very significant. national welfare agency and ana- the number of days of disabling But when Perrott measured sick- lyzed by public health experts in illness in surveys was hardly new. ness disability “expressed in terms morbidity studies. Studies of industrial disability of total duration, time lost through Sydenstricker reacted immedi- usually included such informa- disabling illness and days of con- ately to this report by changing tion. Sydenstricker himself, a finement in bed,”49 everything gears. That same year, he pub- labor economist by training, used changed. Chronic disease was lished an article based largely on such data in his early studies with transformed into a central cause a section of the surgeon gener- Goldberger on South Carolina of morbidity as well as mortality: al’s annual report that had been cotton-mill workers. In fact, the written by Selwyn Collins. working definition of illness for The average total duration, the Sydenstricker’s paper brought to average amount of disability, these studies was inability to the fore what had been a minor and the bed days per case for work. Such data, moreover, were this group, represented by the point in the 1935 paper by Perrott used to confirm the link between degenerative diseases, rheuma- and Collins: the relief population tism, and nervous conditions, poverty and illness.45 In his not only suffered from higher are of a definitely higher order Hagerstown studies, however, of magnitude than those for the rates of illness, it also “contains a Sydenstricker chose another mea- typically acute illnesses caused disproportionately large number by the minor respiratory and sure of illness: “The measure of the of persons who have chronic dis- communicable disease. . . . A incidence of any specific disease study of the relative severity of eases or physical defects or who was the extent to which it mani- the various disease groups in are susceptible to frequent terms of duration thus intro- fested itself in visible illness.”46 attacks of acute illness.”42 He duces a new basis for their This measure presumably did evaluation. . . . [I]llness due to then went on to cite the data on away with the ambiguity involved chronic disease, although rela- physical impairment revealed by tively low in incidence, becomes in interpreting why someone the Federal Emergency Relief of major importance when the stayed away from work, and it severity of the average case is Administration survey: 50 was applicable to nonindustrial considered. These data indicated that (a) a populations. Still, information much higher proportion of per- about the duration of disease Perrott closed the circle by sons on relief had serious physical defects or chronic diseases was collected, and it was speci- returning to the question of than those of the same occupa- fied that 60% of the illnesses illness and economic status. tional class who were not on re- recorded lasted eight days or Al though frequency of illness in lief; (b) in both the relief and nonrelief populations the pro- longer. Sydenstricker promised the CCMC study was more or portion with impairments and in a footnote that a future article less the same or slightly greater

444 | Public Health Then and Now | Peer Reviewed | Weisz American Journal of Public Health | March 2011, Vol 101, No. 3 PUBLIC HEALTH THEN AND NOW

for the well-off than for the poor, all the ills of the flesh must be classified as ill in the study actu- things looked very different if one the ultimate goal of the health ally suffered from chronic disease department. . . . It is becoming focused on duration of disability. widely recognized that physi- or disability. The one-in-six figure This study showed, according to cians and hospitals cannot be nonetheless stuck and would be Perrott, that the poor suffered far expected to render service to cited for the next 20 years as the the indigent without remunera- more days of disability than the tion and that there must be pub- incidence rate of chronic disease rich. Those earning under $1200 lic responsibility for the medical and disability. It would remain a annually had more than twice the care of these unfortunates who key argument in favor of some otherwise must depend upon disability days of those earning the charity of physicians.53 form of health reform, although $3000 or more.51 All the pieces the precise nature of that reform were now in place. By quantify- And the point was indeed would remain contentious. ing chronic diseases and disabili- demonstrated—with the help, it ties as days of disability, one must be said, of some sleight of CONCLUSIONS could demonstrate how pervasive hand. The phenomenon mea- these were, and how much more sured by the survey was The original NHS, conducted the poor, with less access to in the winter of 1935 and 1936, health care, suffered from dis- Disabling illness which had kept may well have been the last large- persons away from work for ease. Elementary social justice scale publicly financed epidemio- seven consecutive days or lon- and economic interest would dic- ger during the 12 months pre- logical survey in the United States tate a reform of health care to ceding the day of the canvass; originated and organized to sup- and other handicapping disease correct this situation. In this way, port radical health care reform. Its or condition including orthope- chronic disease and disability dic impairment, blindness and successor, the permanent NHS became the focus of the NHS. deafness.54 established in 1956, was some- thing very different. In the view A SURVEY OF CHRONIC Through use of such broad and of one of its architects, none other DISEASE AND DISABILITY largely unprecedented criteria, it than St. John Perrott, the Republi- was not hard to provide devastat- can administration needed a Although it was shut down ing statistics regarding the health health bill to satisfy public opin- before most of its data were ana- status of the nation. The New York ion. The least controversial lyzed, the NHS produced an Times blared on its front page option, and the one least likely to Illustration in M. Ross. How Healthy that six million people daily were enormous amount of information. lead to “socialized medicine,” was Are We? 1937. In the 20 years that followed, incapacitated in the United States. the creation of a permanent data- Source. Survey Graphics. Internet Archive: The survey’s Preliminary Report 56 more than 200 reports, articles, gathering agency that would be Digital Library. Available at: http://www. and comparative studies based on framed it differently: innocuous enough to gain biparti- archive.org/stream/surveygraphic26survrich #page/372/mode/2up. Accessed January 4, this survey were published.52 The san support and, although hardly 2011. immediate impact of the study, [I]t is estimated that reactionary, would primarily however, and the one reported 23,000,000 persons, or more serve to produce information use- than one person in six in the ful to health policy planners of all on most widely focused on the United States have some 57 morbidity situation in the United chronic disease, orthopedic im- political stripes. States, which was purportedly pairment or serious defect of Nonetheless, the first NHS had hearing or vision. By reason of major long-term consequences. serious enough to demand major these disorders almost a billion reform of the health care system. days annually are lost from On the technical level, the com- Well before its completion, Roche work or other usual pursuits bination of public health morbid- and a minimum of 1,500,000 ity and welfare disability studies described its conclusions and persons are disabled for such consequences: long periods of time (12 months introduced—or more correctly, or more) that they can be con- reintroduced—a powerful new sidered permanent invalids.55 The survey provides national tool, “days of disability,” that recognition of the fact that the quantified and standardized the health service of the future will probably be expanded to cover A closer look at the reports amorphous notion of serious dis- other fields than control and makes it clear that true pro- ease. This tool was flexible as well. prevention of the communicable longed illness was defined as three Depending on context and moti- diseases. With the cooperation of the medical profession, the months or more of disability and vation, one could define serious control, prevention, and cure of that only about 45% of those disease as seven days, three

March 2011, Vol 101, No. 3 | American Journal of Public Health Weisz | Peer Reviewed | Public Health Then and Now | 445 PUBLIC HEALTH THEN AND NOW

months, or a year of disability. It about the problem of “chronic aroused violent opposition. Compare also had consequences for the disease” would expand exponen- “Doctors Will Not Reply; The Census Disease Queries to Be Ignored,” New York field of epidemiology. Although it tially after World War II, turning Times, May 28, 1890, p. 8 (ProQuest is difficult to demonstrate causal- chronic disease into a major Historical Newspapers); “Disease in the ity, the first NHS—along with the health policy issue, buttressed Census,” Washington Post, May 24, 1890, p. 4 (ProQuest Historical News- next major survey, the East Balti- by the now-dated statistics of papers). more Longitudinal Study, which the NHS. 3. Some of its complex organizational was organized before World War In recent years, chronic dis- features are described in P. J. Funigiello, II by the USPHS and the Mil- ease has continued to be near Chronic Politics: Health Care Security From FDR to George W. Bush (Lawrence: bank Memorial Fund and was the center of health policy dis- University Press of Kansas, 2005), 24– specifically devoted to chronic cussions—overshadowed by the 29. disease58—likely played a signifi- more urgent, ongoing, and con- 4. H. S. Cumming, “Chronic Disease as a cant role in US epidemiology’s troversial health insurance Public Health Problem,” Milbank Memo- rial Fund Quarterly 14, no. 2 (1936): postwar turn to chronic disease, debate but never far from the 125–131, 127. an orientation that has been only surface. During the last presiden- 5. A. Derickson, Health Security for All: slightly modified by AIDS and tial election, a pressure group Dreams of Universal Health Care in severe acute respiratory syn- was formed to push presidential America (Baltimore: Johns Hopkins Uni- versity Press, 2005), 42–56; M. Terris, drome (SARS). candidates to develop policies for “Introduction,” in C. E. A. Winslow, Evo- 62 The survey also had major dealing with chronic diseases, lution and Significance of the Modern political implications. It became and now-President Barack Obama Public Health Campaign (South Burling- ton, VT: Journal of Public Health Pol- the main data source on which mentioned it in one of the tele- icy, 1984); reprint of Winslow’s book the government framed its health vised debates. Our current and initially published in 1923 by Yale Uni- proposals.59 Fox has argued that by now long-standing anxiety versity Press. NHS data played a key role in about the effects of chronic dis- 6. J. Eyler, “Health Statistics in Histori- cal Perspective,” in Health Statistics: policies to plan and construct ease is perhaps the most endur- Shaping Policy and Practice to Improve hospitals, fund biomedical ing legacy of the NHS. Q the Population’s Health, ed. D. J. Fried- research, and expand education man, E. L. Hunter, and R. G. Parrish (Ox- ford: Oxford University Press, 2005), for the health professions.60 The About the Author 42. result most emphasized initially George Weisz is with the Department of 7. This and other material is based on was that low-income groups suf- Social Studies of Medicine, McGill Univer- documents in Collection of Isidore S. fer disproportionately from dis- sity, Montreal, Quebec. Falk Papers, Yale University, Sterling Correspondence should be sent to Memorial Library, Manuscripts and Ar- eases and disabilities and are George Weisz, Social Studies of Medicine, chives, Manuscript Group Number least able to afford health care. McGill University, 3647 Peel Street, Mon- 1039 of the Contemporary Medical This outcome supported argu- treal, Quebec H3A 1X1, Canada (e-mail: Care and Health Policy Collection. Falk [email protected]). Reprints can be eventually published a book on this ments for some form of public ordered at http://www.ajph.org by clicking subject himself. See I. S. Falk, Security health care for the poor, argu- the “Reprints/Eprints” link. Against Sickness: A Study of Health In- ments that led eventually to This article was accepted May 26, surance (Garden City, NY: Doubleday, 2010. Doran, 1936). Medicaid. In the longer term, the survey established in the public 8. Social Security Online, “The Commit- Acknowledgments tee on Economic Security,” available at consciousness that chronic dis- Research for this article was made possi- http://www.ssa.gov/history/reports/ces/ ease was a major public health ble by the Social Sciences and Research cesbasic.html (accessed December 1, problem. A section of the Council of Canada (grant 410-2008- 2009). 0619). National Health Program of 9. Social Security Online, “Members of I am grateful to the archival staff of the Committee, Advisory Boards and 1938, written by the ubiquitous the National Library of Medicine, Na- Staff,” available at http://www.ssa.gov/ Perrott, was devoted to the tional Archives, and Sterling Library of history/reports/ces/ces6.html (accessed Yale University for invaluable help in December 1, 2009). subject. In 1940, the American tracking down sources. Hospital Association and the 10. Falk papers, transcript, oral history interview conducted by Peter A. Corn- American Public Welfare Asso- Endnotes ing, part 1 (copyright, Columbia Univer- ciation published a statement 1. I. Fisher, National Conservation Com- sity), 1965. Series 5: Box 172; File about the need to improve insti- mission. A Report on National Vitality, Its 2634, pp. 196–197. Wastes and Conservation (Washington, 11. Associated Press, “New Deal Sur- tutional care for the chronically DC: Government Printing Office, 1909). 61 veys Life on 700 Fronts,” New York ill. This document was only the 2. An effort by Billings in 1890 to get Times, January 12, 1936, p. N1 (Pro- beginning. Interest in and anxiety doctors to report chronic illnesses Quest Historical Newspapers).

446 | Public Health Then and Now | Peer Reviewed | Weisz American Journal of Public Health | March 2011, Vol 101, No. 3 PUBLIC HEALTH THEN AND NOW

12. Falk Papers, Series 2: Box 39; Work of Edgar Sydenstricker,” Journal Academy of Political and Social Science 62. J. Pindell, “Push to Make Chronic File 133, letter of August 22, 1935, of the History of Medicine and Allied Sci- 188 (1936): 131–143. Disease Part of 2008 Discussion,” Bos- Sydenstricker to Falk. ences 58, no. 1 (2003): 34–55. ton Globe, September 25, 2007, p. A8. 45. E. Sydenstricker, The Challenge of 13. Falk Papers, Series 2: Box 39; File 28. Perrott and Collins, “Relation of Facts; Selected Public Health Papers of 133. Sickness to Income,” 595. Edgar Sydenstricker (New York: Pub- 14. “A Progress Report on a Plan of 29. Ibid, 607. lished for the Milbank Memorial Fund by PRODIST, 1974), 169. Analysis for the Survey of Chronic Dis- 30. Ibid. ease in the United States,” Falk Papers, 46. E. Sydenstricker, “A Study of Ill- Series 2: Box 39; File 139. 31. Ibid, 607, 619. ness in a General Population Group: Hagerstown Morbidity Studies No. I: 15. National Library of Medicine, 32. Ibid, 622. The Method of Study and General George St. John Perrott Papers: MS 370, 33. J. Roche, “Cost of Depression in Results,” Public Health Reports 41, no. Box 1 folder 4, Interview with Perrott Health Revealed,” New York Times, Sep- 39 (1926): 2069–2088, 2074. by Peter Corner, 1966, pp. 21–22. tember 15, 1935, p. E10. 47. Ibid. 16. Falk Papers, Series 2: Box 43; File 34. Ibid. 321; Letter of May 28, 1935 from Falk 48. E. Sydenstricker, “The Incidence of 35. E .P. Boas and N. Michelsohn, The to L. Thomson. Various Diseases According to Age: Challenge of Chronic Diseases (New York: Hagerstown Morbidity Studies No. 17. Cumming, “Chronic Disease,” 128. Macmillan Company, 1929); G. H. VIII,” Public Health Reports 20 (1928): Bigelow and H. L. Lombard, Cancer and 18. N. Krieger, Embodying Inequality 1124–1156, 1125. Epidemiologic Perspectives (New York: Other Chronic Diseases in Massachusetts 49. Perrott, “The State of the Nation’s Baywood Publishing Company, 2005), (Boston: Houghton Mifflin Company, Health,” 138. 39–42. 1933). 50. Ibid, 139–140. 19. E. Sydenstricker, “A Study of Illness in 36. Social Security Online, “CES Report a General Population Group: Hagerstown on Health Insurance: The Unpublished 51. Ibid, 140–141. Morbidity Studies No. I: The Method of 1935 Report on Health insurance & Disability,” March 7, 1935, Committee 52. F. E. Linder, “National Health Survey,” Study and General Results,” Public Science 127 (1958): 1275–1280, 1277. Health Reports 41, no. 39 (1926): on Economic Security, available at 2069–2088, 2083–2084. http://www.socialsecurity.gov/history/ 53. J. Roche, “Economic Health and reports/health.html (accessed December Public Health Objectives,” American 20. E. Sydenstricker, Health and Envi- 2, 2009). Journal of Public Health and the Nation’s ronment (New York: Arno Press and the Health 25 (1935): 1181–1185, 1183. New York Times, 1972; originally pub- 37. C. E. A. Winslow, “Public Health at lished in 1933). the Crossroads,” American Journal of 54. Preliminary Reports, Bulletin 6: The Public Health 16, no. 11 (1926): 1075– Magnitude of the Chronic Disease Problem 21. Ibid, 86–88. 1085; A.|J. Viseltear, “Emergence of the in the United States (Washington, DC: 22. E. Sydenstricker, “Economic Status Medical Care Section of the American National Health Survey, 1938 [revised and the Incidence of Illness: Hagerstown Public Health Association, 1926– 1939]), 2. 1948,” American Journal of Public Morbidity Studies No. X: Gross and 55. “6 Million Found Ill in Day by Sur- Health 63, no 11 (1973): 986–1007. Specific Illness Rates by Age and Cause vey of National Health,” New York Among Persons Classified According to 38. Bigelow and Lombard, Cancer and Times, January 17, 1938, p. 1 (ProQuest Family Economic Status,” Public Health Other Chronic Diseases; G. A. Soper, Historical Newspapers); Preliminary Re- Reports 44, no. 30 (1929): 1821–1833, G. H. Bigelow, and H. F. Vaughan, “What ports, Bulletin 6, 12. 1827. Official Public Health Agencies Should 56. Interview with Perrott by Peter 23. Ibid, 1827. Do About Cancer,” American Journal of Corner, 1966, p. 39. 24. I. S. Falk, M. C. Klem, and N. Sinai, Public Health 17, no. 11 (1927): 1135– The Incidence of Illness and the Receipt 1141. 57. A. Haywood, “The National Health Survey—In the Beginning,” Public Health and Costs of Medical Care Among Repre- 39. For example, see M. C. Jarrett, Reports 96, no. (1981): 195–199. sentative Families (New York: Arno Chronic Illness in New York City (New Press, 1976). York: Published for the Welfare Council 58. J. Downes, “Findings of the Study of 25. Sydenstricker, Health and Environ- of New York City by Columbia Univer- Chronic Disease in the Eastern Health ment, 38. sity Press, 1933). Also compare Report District of Baltimore,” Milbank Memorial on Chronic Disease in New Jersey (Tren- Fund Quarterly 22 (1944): 337–351. 26. S. D. Collins, “The Incidence and ton: New Jersey Department of Institu- Causes of Illness at Specific Ages,” Mil- 59. E. D. Berkowitz, Rehabilitation: The tions and Agencies, 1932). bank Memorial Fund Quarterly 13, no. 4 Federal Government’s Response to Dis- (1935): 320–338. 40. G. S. J. Perrott and H. C. Griffin, “An ability, 1935–1954 (New York: Arno Inventory of the Serious Disabilities of Press, 1980), 43. 27. G. S. J. Perrott and S. D. Collins, “Re- the Urban Relief Population,” Milbank lation of Sickness to Income and In- 60. D. M. Fox, review of P.|J. Funigiello’s Memorial Fund Quarterly 14, no. 3 come Change in 10 Surveyed Commu- Chronic Politics, American Historical Re- (1936): 213–241, 216. nities: Health and Depression Studies view 112 (2007): 248–249. No. 1: Method of Study and General 41. Ibid, 218. 61. Association American Hospitals and Results for Each Locality,” Public Health 42. E. Sydenstricker, “Surgeon General’s American Public Welfare Association, In- Reports 50, no. 18 (1935): 595–622. Report on Health and Depression stitutional Care of the Chronically Ill: A Re- On Sydenstricker’s exclusion of African Study,” Milbank Memorial Fund Quar- port of the Joint Committee on Hospital Care Americans from his health studies, and terly 14, no. 3 (1936): 205–208, 207. of American Hospital Association and on his early interest in the economic American Public Welfare Association / 43. Ibid, 207–208. burden of illness, see H. M. Marks, “Epi- Michael M. Davis, Chairman (Chicago: demiologists Explain Pellagra: Gender, 44. G. S. J. Perrott, “The State of the Na- American Public Welfare Association, Race, and Political Economy in the tion’s Health,” Annals of the American 1940).

March 2011, Vol 101, No. 3 | American Journal of Public Health Weisz | Peer Reviewed | Public Health Then and Now | 447