- Email: [email protected]
Cardiac disease construction on the borderland
Pergamon
Soc. Sei. Med. Vol. 44, No. 7, pp. 1043-1049, 1997
PII: S0277-9536(96)00239-0
(~ 1997ElsevierScienceLtd All rights reserved. Printed in Great Britain 0277-9536/97$17.00 + 0.00
CARDIAC DISEASE CONSTRUCTION ON THE BORDERLAND J E A N N E DALY ~ and IAN M C D O N A L D ~LSchool of Sociology and Anthropology, La Trobe University, Bundoora, Victoria 3083, Australia and 2Cardiac Investigation Unit, St Vincent's Hospital, Fitzroy, Victoria 3068, Australia Abstract--The diagnosis of possible heart disease in the well patient has undergone remarkable shifts over the past century. The traditional medical view places strong emphasis on the contribution of technological data to the diagnosis of disease. In the case of serious heart disease, cardiac diagnostic technologies can play a defining role but, more often in the clinical context, patients are assessed for heart disease which is minor. The question is whether disease is present at all. In this borderland between health and disease, the interpretation of technological data is inherently uncertain. The diagnosis then depends more heavily on the social utility of particular disease categories. Shifts in diagnostic categorisation are not therefore attributable solely to more extensive forms of cardiac imaging but are socially constructed in an interactive context which involves the technology, the medical profession and the wider social structures which exist at the time of diagnosis. Claims of technological certainty create a social space within which the medical profession generates disease categories. These shifting disease categories may serve the needs of patients but may also be influenced by those of other players. © 1997 Elsevier Science Ltd. All rights reserved Key words--heart disease, diagnosis, medical technology
INTRODUCTION Ryle (1948) pointed out in his classic textbook that diagnosis is the most important task of the physician since it determines both assessment of prognosis and choice of treatment. From the 18th Century onwards, the diagnosis of heart disease has increasingly depended upon technology which is credited with making historic advances in the evolution of knowledge about the heart. In the 19th Century, the stethoscope revolutionised diagnosis. The electrocardiograph, radiograph, cardiac catheterisation, angiography and echocardiography each contributed knowledge about the heart (White, 1951), progressively penetrating the living body to display abnormality and disease. The contribution of cardiac tests to the treatment of organic heart disease is u n d i s p u t e d - - m o d e r n cardiac surgery would be impossible without accurate diagnostic information from cardiac imaging. However, when physicians encounter symptoms such as chest pain, palpitations, breathlessness or even such signs as heart murmurs, these are frequently not manifestations of serious heart disease. Indeed, the diagnostic question is often whether heart disease is present at all. In such cases, diagnosis requires negotiating the borderland between health and disease. With patients in whom heart disease is possible but unlikely, statistical uncertainty is inevitable, doubt is common and there is the risk of diagnostic error. A borderland character-
ised by technical uncertainty presents a prime site for analysing the social subtext of diagnosis. The social construction of disease has been wellargued although sociologists may have emphasised the social context to the partial exclusion of the physical world (Fox, 1993, p. 150). We draw here on Figlio's argument that diseases are socio-clinical entities, arising as a set of symptoms in the context of social change (Figlio, 1982, 1986). The medical profession is authorised to deal with the socioeconomic and political meaning of the symptoms, thus producing a social space within which disease is constructed. Figlio distinguishes between practices which can be precisely defined and those which are less controlled, requiring interpretation through experience. It is the inherent uncertainty and dependence on interpretation which makes the borderland between health and disease significant in identifying the interplay between the physical and social worlds. As a case study we are focusing on the borderland of cardiac diagnosis using imaging technologies. While all diagnostic processes have a borderland, we have selected cardiac imaging because, sociological arguments notwithstanding, it represents the area where clinicians are least inclined to accept the social construction of disease. After all, imaging has proved its worth in planning cardiac surgery and the internal structure and function of the human body is displayed in such vivid detail that it seems counterintuitive to argue that
1043
1044
Jeanne Daly and lan McDonald
these images do not also determine diagnosis on the borderland. Our aim is to argue for a postmodern view of diagnosis which acknowledges ambiguity and which may well benefit clinician and patient alike by an increased emphasis not on technological data but on the care of the patient. However, the patients themselves are absent from our analysis in this case, appearing only as the focus of competing medical views on where, and why, to draw the diagnostic boundary. Our data are drawn from the medical debate on the technological diagnosis of heart disease in the borderland. Such diagnosis has undergone swings of the pendulum in the course of 150 years. Signs and symptoms attributed to heart disease have attracted various labels, including soldiers' heart, the effort syndrome and mild mitral valve prolapse, and these show considerable overlap. Wooley (1976) argues that these are historical manifestations of the same disease. Alternately, the heart could be seen as normal and the symptoms as neurotic. Commonly the change in diagnostic label was attributed to the development of a new diagnostic technology but it is possible to elicit the role of the social context at least in an implicit form. The source of the problem has always been uncertainty over what the data mean. Recognising that technical and social aspects of diagnosis are inextricably entwined, for conceptual purposes we discuss first the technical source of uncertainty in the present time. We then turn to the way in which uncertainty has been interpreted historically to give a changing disease label. Uncertainty has remained, paradoxically perhaps increased, with the availability of diagnostic images of increasing sophistication and detail. Patients in whom heart disease is possible, but unlikely, present in a confusing manner. One group has symptoms like chest pain, breathlessness or palpitations which may indicate organic heart disease or may be manifestations of anxiety; they may even be a healthy cardiac response to fear or stress. In others, doubt has arisen because of a heart murmur heard with a stethoscope or a suspected abnormality in a chest radiograph or an electrocardiogram. Heart murmurs are merely audible blood flow but they are difficult to diagnose. Some murmurs indicate heart disease; more often, especially in the young, they only indicate rapid blood flow. The search is for a technology which can resolve the implications for disease of these overlapping categories. The diagnostic accuracy of modern imaging may now surpass that of direct inspection of the heart at surgery or autopsy. Hence a test such as echocardiography (cardiac ultrasound) is beholden to no "gold standard" against which its pronouncements can be checked. When the "gold standard" for cardiac diagnosis was set by the stethoscope, it was well understood that clinicians had to have con-
siderable skill in interpreting nuances of the heart sounds. Only rarely could they have physical evidence to support the diagnosis. The illusion now is that cardiac imaging, giving a direct visual display of the beating heart, has circumvented the difficulty of interpretation. After all, "seeing is believing". The images are seductive, distracting attention from the fact that diagnostic criteria are still necessary to distinguish normal from abnormal. Thus diagnosis remains a matter of clinical judgement, difficult, prone to variation and thus contentious. A test result of doubtful significance will be called "equivocal" often as a result of lying on the borderline between disease and normality. It is recognised that, with increasing use of the test on well populations, there is an unavoidable epidemiological risk of false diagnosis of disease. What is less well recognised is that there is an even greater risk of an equivocal result. Since there is no other technology to which we can appeal, a false or doubtful result will often stick. If diagnosis is uncertain, so is treatment. We have shown that the best of the noninvasive cardiac imaging techniques, echocardiography actually makes only a minor contribution to the management of well patients (McDonald et al., 1988). If minor disease is diagnosed, antibiotic prophylaxis can be prescribed against bacterial infection of the heart valves but effectiveness is unproven (Anon., 1992). Alternatively, if the heart is found to be normal, anxious patients benefit from reassurance. Unfortunately, the very process of referring the patient for testing raises concern about the heart. A normal test result usually reduces this concern about the heart but it does not necessarily resolve it, nor does it ensure that concern will not recur (Daly, 1989; Daly and McDonald, 1993). In the case of a false or equivocal result, anxiety will be increased. Nevertheless, echocardiography is commonly used on well patients. We found that the reasons were more often social than technical: maintaining professional credibility, satisfying the real or implied request of the referring doctor or satisfying the perceived expectations of insurers. If such social considerations can influence the ordering of the test, might they not also subtly influence the interpretation of the image? We now turn to a more detailed analysis of the overlapping concerns of diagnostic uncertainty and the social construction of disease. Our aim is to identify the contribution made by diagnostic technologies but also to identify the way in which criteria for disease have been responsive to social forces. Our intention is not to present a detailed analysis of the social forces acting at any one time but to establish the need for a more definitive analysis of the social context of diagnosis at the margins of disease.
Cardiac disease construction on the borderland THE HISTORY
It was the stethoscope which first promised to "display" the internal structure of the beating heart: ... the stethoscope, when applied to the organ of hearing, shall convert it into an organ of vision, enabling the listener to observe, with the clearness of ocular demonstration, the ravages which disease occasionally commits in the very centre of the rib-cased cavity of the body (Anon., 1831-1832, p. 122). The stethoscope facilitated the collection of scientific data on heart disease. Since, "[i]n the pursuit of science every truth, every fact discovered is of value" (Corrigan, 1828, p. 588), benefit to patients was relegated to the sidelines. The stethoscope performed its scientific task so convincingly that it came to symbolise the scientific doctor. Diagnostic technologies also strengthened the social credentials of the emerging medical profession with social institutions like employers, the army and insurers by providing an objective means of evaluating evidence from "malingerers, conscripts, applicants for pensions, insurance or society benefits, plaintiffs claiming damages for i n j u r y . . . " (Jackson, 1984, p. 337). Uncertainty dogged the new science. Heart sounds matched "the diversity of hues in a rainbow" (Anon., 1826, p. 471). Some murmurs and heart irregularities were linked by autopsy to fatal disease but murmurs also occurred in the absence of any organic heart disease, the so-called "functional" murmur. Interpretation was difficult: ...the murmur revealed by auscultation became a bugbear; it was misinterpreted as meaning a damaged valve and as ominously threatening an early death.., only too often the physician spread gloom as he warned the parents of the presence of the murmur that threatened disaster (Herrick, 1942, p. 231). Diagnosis thus posed a dilemma for clinicians, a problem clearly perceived by Austin Flint (who identified Flint's murmur): Without the negative proof afforded by physical exploration the mind of the practitioner must be in doubt as to the diagnosis. If he give a decided opinion, it is a guess which may prove to be either fight or wrong. If he avoid giving a decided opinion, the inference which the patient usually draws is that organic disease exists and that the physician is reluctant to tell the truth (Flint, 1888, p. 354). Harm could be done to patients and Flint concluded: " . . . truly fortunate were they who kept aloof from the stethoscope of the auscultator" (quoted in Baldry, 1971, p. 72). The American Civil War
During the American Civil War, the physician, Da Costa, identified a group of soldiers with palpitations, chest pain, rapid, often irregular pulse, indigestion and diarrhoea. Remarkably, despite the carnage of Civil War battles, Da Costa did not associate these problems with fear. Indeed, he noted that one of these soldiers loved campaigning and
1045
refused to leave his regiment. Da Costa concluded that war affected the heart: "It seems to me the most likely that the heart has become irritable, from its over-action and frequent excitement..." (Da Costa, 1871). Soldiers' heart was distinct from the feigned heart disease of malingerers induced by tightly binding the chest. But how was it to be treated? Da Costa recognised that a disease label would deplete and demoralise the army. The compromise he found helped to establish the role of the doctors in the army: Soldiers' Heart was a legitimate "affection" which could progress unless treated with rest and digitalis; after treatment soldiers could be returned to active service. Preventive action included a reduction in forced marches with heavy packs. Da Costa's compromise benefited the army and strengthened the role of the army doctor. Even the soldiers may have been grateful for a rest. The symptoms characterising Soldiers' Heart were also common in the community and recognised to be of considerable practical importance. In this context, however, the problem was seen as neurotic, presenting in women who were hysterics, menopausal or who had a displaced uterus or irritated ovaries. Men too could suffer the symptoms as a result of sexual excesses, over-indulgence in tobacco or the mental worries of business (Bramwell, 1884). Within the civilian population, and after the war, the failure was in the person rather the heart. Treatment focused on changing debilitating personal habits. Two explanations of patients' problems, cardiac and neurotic, legitimate and deviant, could thus coexist, predicated on assumptions about the characteristics of the social group to which the patient belonged. The invention of the electrocardiograph produced tracings of electrical impulses from heart beats. This swung the pendulum towards identifying heart irregularity as disease. Pioneer cardiologist Thomas Lewis saw it as evidence of a pathological condition with the risk of serious irregularities in later years. However, he cautioned that "palpitations" were also experienced as "subjective phenomena" by nervous, young females and in people who were tired or had over-indulged. Such patients should be regularly monitored (Lewis, 1915, pp. 51-54). The two Worm Wars
The First World War brought a reassessment of the significance of murmurs. James McKenzie's 1915 edition of Diseases of the Heart pointed out that the murmur heard during systole, associated with regurgitating blood from the mitral valve, had become a medical "bogy", causing harm to patients, including applicants for life insurance and recruits to the armed services. He argued that such murmurs were "functional", did not affect prognosis and might be precipitated by anxiety or other problems like physical debility (McKenzie, 1914, pp. 326 327). McKenzie's views gained currency after
1046
Jeanne Daly and Ian McDonald
the Allied army suffered major defeats in France. There was a flood of casualties and the military command battled what it saw as a contagious epidemic of soldiers failing to face danger (Butler, 1943). Among this list were soldiers presenting with chest pain, breathlessness, palpitations and an irregular heart beat. Initially these soldiers were diagnosed as suffering from valvular heart disease and disordered action of the heart (DAH) (the name under which Soldiers' Heart had won official military recognition) (Howell, 1985). Lewis, by then, consultant physician in diseases of the heart to the British Army, now reinterpreted "the borderland between health and disease" in a manner "tinged with a war purpose" (Lewis, 1919a, pp. 4, iv). There had been too many rejections from the army and too many pensions. Thus heart sounds lost their significance in favour of an emphasis on short term prognosis (Lewis, undated). DAH was recast as the Effort Syndrome, an exaggerated cardiac response to physical effort in the puny or nervous. Such soldiers were removed from hospitals, put on exercise programmes and half of them were returned to the front. The immediate economic saving was considerable. Howell (1985) argues that Lewis earned a knighthood by excluding soldiers from pensions for heart disease. This redefinition of heart disease helped establish the credentials of the new specialty of cardiology. What cardiologists learnt in war time, they then put into practice in peace time. Lewis argued that the diagnosis of murmurs, particularly murmurs associated with an incompetent mitral valve, was "a diagnosis by fashion only" which lead to too many unwarranted pensions: It is my confident belief that, had systolic murmurs and modifications of the heart sounds never been discovered, the practice of the profession would have stood on a much higher plane today than it actually does... As these signs are used by medical men in general at the present time, they are productive of infinitely more harm than good (Lewis, 1919b, p. 7). This turn of events benefited cardiology but not the life insurance industry which responded by claiming that the evaluation of murmurs was not a clinical question but a matter for "life insurance medicine". After a review of companies' records showed that functional heart murmurs were associated with at least a fourfold increase in mortality, they concluded that no murmur was innocent (McCrudden, 1931) and applicants with murmurs were penalised (Blumenthal, 1942). A psychosomatic explanation was also current. Extra heart sounds during systole were found to be common and associated with "psychoneurosis". The cause of the sounds was unknown but "delicate strands joining pericardium and pleura" were reported (Johnston, 1938, p. 222). Clinicians again faced a difficult choice. Heart disease was associated with loud murmurs but,
since even a faint murmur might indicate rheumatic fever which could severely damage the heart, a murmur could only be called benign if all organic cause had been carefully excluded (Levine, 1937, p. 209). Many schoolchildren with murmurs were prescribed prolonged rest in convalescent homes and hospitals (Garrow, 1943). Alternatively, if murmurs were seen as psychosomatic, prognosis was believed to be unaffected but symptoms were resistant to treatment. Clinicians were treading dangerous territory. The strong fear of heart disease in the community predisposed people to worry about their hearts. Thus a doctor uncertain about functional symptoms could create a "cardiac cripple": "[h]aving made the diagnosis of heart disease one should speak with emphasis and assurance" (Levine, 1937, p. 216). The Second World War swung the pendulum against the diagnosis of heart disease. Murmurs were seen as ambiguous, too uncertain in their significance to provide the basis for a disability pension. There was concern about "unwarranted cardiac invalidism" among children and the need to unify the medical view of murmurs (Evans, 1943). Clinicians were confused: Probably every second of every day some doctor somewhere is confronted with a mitral systolic murmur, and he has got to do something, or nothing, about it... I should like to know what causes the murmur and why it is "innocent"... Anybody talking about "innocent" murmurs should.., give his reasons for calling them so by advancing some physical cause for their occurrence (Gaskell, 1943, p. 582). Not so in the armed services. From January, 1940, Paul Wood was physician to the British army's Effort Syndrome Unit (Wood, 1941). He rapidly recognised Da Costa's "condition" and the Effort Syndrome as the same obsolete concept. The problem, he argued, was fear. To Wood we owe the recognition that the diarrhoea noted by da Costa could be attributable to fear before a battle! Wood, however, did not consider the fear response normal. It occurred in people with a family history of nervousness or insanity and an inferior character caused by having been allowed to "cling too long to their mothers' skirts". Psychoneurosis could be diagnosed with a few pertinent questions: "their distaste for this label may prove quite helpful". Cardiac neurosis was also present in the community: " . . . t h e y are merely clothed differently, the former in battle dress, the latter in nylon" (Wood, 1956, p. 202). Its victims tended to be women, from the "emotional" races (Jews and Italians) and "indifferent to higher ideals". Accepting the patient account could turn a patient into "a chronic and incurable psychoneurotic", so the doctor had to persuade the patient that "he has met a doctor who thoroughly understands his case". Since "[i]t is remarkable what little insight these patients have, and disconcertingly how little shame", a normal diagnostic test provided an important means of
Cardiac disease construction on the borderland exposing an hysterical motive for gain (Wood, 1956, pp. 944-946). In Wood's hands the diagnosis of deviance reached its purest form. In contrast, in the United States army, army psychiatrists found malingering to be rare but symptoms did offer an escape mechanism. They recognised that, in war, all personnel could develop psychoneurosis. Psychiatrists were sent into the field to build up morale and the motivation to fight. The lesson they learnt from the war was that the emotional problems of the individual should be considered "within the context of his group and social culture" (Medical Department, U.S. Army, 1966, Chap. 1). The technological explosion in cardiology The post-war years ushered in an heroic period of intervention in cardiology which depended on accurate diagnosis. Attention returned to murmurs. A series of well patients with late systolic murmurs was subjected to angiography, X-ray imaging of a dye dispersing in the heart from a catheter fed into the left ventricle. In some cases they observed regurgitation back into the atrium indicating a faulty mitral valve (Barlow et al., 1963). Sometimes the leaflet supporting the mitral valve "billowed" or domed (Barlow and Bosman, 1966)--hence mitral valve prolapse. Patients with late systolic murmurs, excluded from life insurance in the 1930s, classified as psychoneurotic during the war years, could now be diagnosed as having mild mitral valve prolapse. Angiography was too dangerous to perform on well patients whose murmurs had to be diagnosed according to labile criteria based on "experience": The point of maximum intensity of late systolic murmurs has ranged, in our experience, from the middle of the murmur to very near the end... Segall and Likoff... have regarded the diamond-shaped late systolic murmur, and especially that initiated by a systolic click, as innocent (Barlow and Bosman, 1966, p. 176). The problem for the clinician worsened (Caceres and Perry, 1967, p. 231): which late systolic murmurs were innocent and which were attributable to mitral regurgitation? Echocardiography introduced safe, noninvasive imaging appropriate for well and ill. As a result, mitral valve prolapse proliferated "in almost every country where clinical investigation is performed" (Wooley, 1976, p. 750). The disease was found in up to 21% of healthy young female volunteers without the murmur (Markiewicz et al., 1976). Mild mitral valve prolapse had no cause, treatment or cure and an unknown prognosis (Criley et al., 1966). However, penicillin could now be used to prevent bacterial infection of an incompetent, regurgitating mitral valve. The spread of heart disease in the well study population is worth noting (Markiewicz et al., 1976). Forty nine out of the 100 well volunteers had heart murmurs and, usually as a result, 10 had previously been diagnosed as having a "heart problem". Symptoms included 29 with
1047
chest discomfort, 30 with palpitations and nine with breathlessness. A diagnosis of mild mitral valve prolapse was thought to account for both signs and symptoms. The tide was to turn again in the 1980s. The reasons are not clear. One possibility is that economic recession induced a more critical approach to technological proliferation. There are also signs of the by now familiar clinical backlash against an "epidemic" disease of little benefit and potential harm to patients. The significance of murmurs was reassessed. Nishimura et al. (1985) published a long-term follow-up of echocardiographically diagnosed mitral valve prolapse. It was argued to be benign in the majority of cases, especially in the young and healthy. Only a minority of patients had serious disease which might require valve replacement. An editorial in the New England Journal Of Medicine (Wynne, 1986) identified the same two groups: the high risk group and a group "no more symptomatic than their peers without the diagnosis". The epidemic of mitral valve prolapse in the well was attributed to the belief that echocardiography could "distinguish normal from abnormal". More stringent criteria were needed to exclude "previously silent normal variants of structure and function" displayed by the test. Apart from identifying a culprit, the technology, there was little explanation offered for the change in criteria except to argue that mitral valve prolapse was a "disease of fashion", the result of a test having "its day in the sun" and seductive since it provided a "catchall explanation of a number of frustrating complaints". DISCUSSION AND CONCLUSION
We accept the efficacy of cardiac technology in the diagnosis of manifest heart disease. Mitral valve prolapse, for example, is an anatomically distinct and potentially serious disease in a small minority. Diagnosis, however, occurs along a continuum from serious disease through manifestation of uncertain significance to normality. We have focused on the latter end. While medical discourse has, for 150 years, emphasised the need for technical certainty in diagnosis, the intraprofessional discourse of diagnosis at the margins has been about how best to give meaning to the image. Technological "advance" notwithstanding, it has been impossible to draw an exact line between structurally sound and diseased hearts. For the unwary, Thomas Lewis (1919a) (p. 61) said, "the whole of this ground is filled with craters, craters filled with memories of battles old and recent fought over them". It remains true today. The criteria for diagnosis in the borderland between health and disease are inherently uncertain. Technologies have produced ever more detail. Indeed, given modern imaging potential, it is not clear who can be confident of remaining normal.
1048
Jeanne Daly and lan McDonald
But who has disease? A diagnostic label derives from the image. However, it also involves clinical judgment about where to draw the line between disease and health in the case of particular patients. This is where the social context of diagnosis comes powerfully to the fore. Historically the interests of the medical profession itself and of life insurers have promoted the identification of disease. The interests of the army, pension schemes and perhaps the state as health insurer have been served by denying disease or by promoting alternative explanations such as cardiac neurosis. Recognition of the clinicians' responsibility in caring for patients has often been referred to, more as a focus of concern than as a central issue of debate. The way in which these conflicting forces are resolved in practice is obscured by the smoke screen of technological certainty. In part, we would argue, there is the need for better science. Scientifically, we need more rigorous criteria for the categorisation of heart disease according to the prognosis associated with particular signs or symptoms. However, especially in the case of a steady increase in imaging capacity, uncertainty about the interpretation of the image is inevitable. Scientifically, this carries certain implications for practice. Firstly, on the borderland of disease, the search for a definitive image is misplaced. Secondly, the emphasis on certainty obscures the substantial contribution to diagnosis made by the social context of patient and doctor. While patients may be malingering, it is equally true that doctors may be acting out of personal interest or under the influence of powerful social institutions. In the absence of clear knowledge of the consequences of diagnosis based on such interests, they provide a decidedly insecure scientific foundation for the diagnosis of disease or deviance. Scientifically, we need to subject diagnostic practice to critical scrutiny, including the social implications for the patients so diagnosed. Otherwise, there is no option but to have recourse to the unsatisfactory, unscientific explanation of blaming a prevailing "fashion". It is also surely bad science to harm patients. Labelling patients with disease may cause harm (Haynes et al., 1978) especially when there is no treatment for the condition diagnosed; patients whose distressing symptoms are attributed to their own deviance are clearly not helped. Such problems of diagnosis on the borderland remain unresolved and they will not be resolved while the only response is further incantation of the diagnostic capacity of imaging technologies. On the contrary, we could turn to an alternative direction. The uncertainty of diagnostic data in the borderland presents clinicians with a dilemma. They worry when they take action based on uncertain knowledge which may involve doing harm to their patients. The discourse which carries these concerns is subjugated by the warning that they must appear
certain or risk turning their patients into "cardiac cripples"; instead they must take responsibility for exerting control over patients' problems. A public acknowledgment of the inherent uncertainty of diagnostic decisions in the borderland would weaken this control over patients' problems, it is true, but it would open a new social space within which a different medical discourse could grow. It would emphasise ambiguity instead of certainty, negotiation instead of control and provide better opportunity for doctor and patient together to negotiate the meaning of the patient's problem. We end by echoing words from what Fox (1993) (p. 160) calls a "postmodern prayer": Where there was truth May we celebrate ambiguity Where there was control May we be generous.
REFERENCES
Anon. (1992) Chemoprophylaxis for infective endocarditis: faith, hope, and charity challenged. Lancet 339, 525526. Anon. (1831-2) Review. Lancet 2, 122. Anon. (1826) Review from the West. Lancet 9, 471. Baldry, P. E. (1971) The Battle against Heart Disease. Cambridge University Press, Cambridge. Barlow, J. B. and Bosman, C. K. (1966) Aneurysman protrusion of the posterior leaflet of the mitral valve. American Heart Journal 71(2), 166-178. Barlow, J. B., Pocock, W. A., Marchand, P. and Denny, M. (1963) The significance of late systolic murmurs. American Heart Journal 66(4), 443-452. Blumenthal, J. S. (1942) The systolic murmur. Annals of Internal Medicine 17, 637-641. Bramwell, B. (1884) Diseases of the Heart and Thoracic Aorta. Young, J. Pentland, Edinburgh. Butler, A. G. (1943) The Australian Army Medical Services in the War of 1914-1918, Vol. III. Australian War Memorial, Canberra. Caceres, C. A. and Perry, L. W. (1967) The Innocent Murmur." A Problem in Clinical Practice. Little and Brown, Boston. Corrigan, D. J. (1828) Aneurism of the aorta. Lancet 1, 588. Criley, J. M., Lewis, K. B., O'Neal Humphries, J. and Ross, R. S. (1966) Prolapse of the mitral valve: clinical and cine-angiographic findings. British Heart Journal 28, 488-496. Da Costa, J. M. (1871) On irritable heart: a clinical study of a form of functional cardiac disorder and its consequences. American Journal of the Medical Sciences. 33, 40. Daly, J. (1989) Innocent murmurs: echocardiography and the diagnosis of cardiac normality. Sociology of Health and Illness 11(2), 99-116. Daly, J. and McDonald, I. (1993) The Social Impact of Echocardiography: Opening Pandora's Box. Australian Government Publishing Service, Canberra. Evans, W. (1943) Mitral systolic murmurs. British Medical Journal January 2, 8-9. Figlio, K. (1982) How does medicine mediate social relations? Workmen's Compensation Practices and medico-legal practices, 1890-1940. In The Problem of Medical Knowledge: Examining the Social Construction of Medicine, eds P. Wright and A. Treacher, pp. 174-
224. Edinburgh University Press, Edinburgh.
Cardiac disease construction on the borderland Figlio, K. (1986) Medical diagnosis, class dynamics, social stability. In Science, Technology and the Labour Process, eds L. Levidow and B. Young, pp. 129-165. Free Association Books, London. Flint, A. (1888) A Treatise on the Principles and Practice of Medicine. Designed for the Use of Practitioners and Students of Medicine, 6th Edn. J. and A. Churchill, London. Fox, N. J. (1993) Postmodernism, Sociology and Health. Open University Press, Buckingham. Garrow, R. P. (1943) Unwarranted cardiac invalidism in children. British Medical Journal 23 January, 114--115. Gaskell, H. S. (1943) Mitral systolic murmurs. British Medical Journal g, 582. Haynes, R. B., Sackett, D. L., Taylor, D. W., Gibson, E. S. and Johnson, A. L. (1978) Increased absenteeism from work after detection and labelling of hypertensive patients. New England Journal of Medicine 299, 741. Herrick, J. B. (1942) A Short History of Cardiology. Charles C. Thomas, Springfield. Howell, J. D. (1985) "Soldier's Heart": the redefinition of heart disease and specialty formation in early twentiethcentury Great Britain. In The Emergence of Modern Cardiology, eds W. F. Bynum, C. Lawrence and V. Nutton, pp. 34-52. Wellcome Institute for the History of Medicine, London. Jackson, E. (1984) Report on the values of objective tests for the determination of Ametropia, Ophthalmoscopy, Ophthalmometry, Skiascopy. Journal of the American Medical Association 88, 337. Johnston, F. D. (1938) Extra sounds occurring during cardiac systole. American Heart Journal 15, 222. Levine, S. T. (1937) Clinical Heart Disease. W. B. Saunders Company, Philadelphia. Lewis, T. (1915) Clinical Disorders of the Heart Beat." A Handbook fi,r Practitioners and Students. Shaw and Sons, London. Lewis. T. (1919a) The Soldier's Heart and the Effort Syndrome. Shaw and Sons, London. Lewis, T. (1919b) On cardinal principles in cardiological practice. British Medical Journal November 15, p. 7.
1049
Lewis, T. (undated) Prognosis in cardio-vascular affections: prognosis with special reference to the assessment for pension purposes and to the capacity of the worker. British Medical Association. Markiewicz, W., Stoner, J., Lomdon, E., Hunt, S. and Popp, R. (1976) Mitral valve prolapse in one hundred presumably healthy young females. Circulation 53(3), 464473. McCrudden, F. H. (1931) Heart murmurs and insurance. New England Journal of Medicine 204(12), 598-600. McDonald, I. G., Guyatt, G. H., Gutman, J. M., Jelinek, V. M., Fox, P. and Daly, J. (1988) The contribution of a non-invasive test to clinical care: the impact of echocardiography on diagnosis, management and patient anxiety. Journal of Clinical Epidemiology 41(2), 151-162. McKenzie, J. (1914) Diseases of the Heart, 3rd Edn. Henry Frowde, London. Medical Department, United States Army (1966) Neuropsychiatrv in Worm War 11." Volume I, Zone of the Interior. Offices of the Surgeon General, Washington. Nishimura, R. A., McGoon, M. D., Shub, C., Miller, F. A., Ilstrip, D. M. and Tajik, A. J. (1985) Echocardiographically documented mitral-valve prolapse: long-term follow-up of 237 patients. New England Journal of Medicine 313(21), 1305-1309. Ryle, J. A. (1948) The Natural History of Disease, 2nd Edn. Oxford University Press, Oxford. White, P. D. (1951) Heart Disease, 4th Edn. Macmillan, New York. Wood, P. (1941) Aetiology of Da Costa's syndrome. British Medical Journal 7 June, 845-851. Wood, P. (1956) Diseases of the Heart and Circulation, 2nd. Edn Eyre and Spottiswoode, London. Wooley, C. F. (1976) Where are the diseases of yesteryear? Da Costa's syndrome, soldier's heart, the effort syndrome, neurocirculatory asthenia--and the mitral valve prolapse syndrome. Circulation 53(50), 749-751. Wynne, J. (1986) Mitral-valve prolapse. New England Journal of Medicine 314 (9), 577-578.
Soc. Sei. Med. Vol. 44, No. 7, pp. 1043-1049, 1997
PII: S0277-9536(96)00239-0
(~ 1997ElsevierScienceLtd All rights reserved. Printed in Great Britain 0277-9536/97$17.00 + 0.00
CARDIAC DISEASE CONSTRUCTION ON THE BORDERLAND J E A N N E DALY ~ and IAN M C D O N A L D ~LSchool of Sociology and Anthropology, La Trobe University, Bundoora, Victoria 3083, Australia and 2Cardiac Investigation Unit, St Vincent's Hospital, Fitzroy, Victoria 3068, Australia Abstract--The diagnosis of possible heart disease in the well patient has undergone remarkable shifts over the past century. The traditional medical view places strong emphasis on the contribution of technological data to the diagnosis of disease. In the case of serious heart disease, cardiac diagnostic technologies can play a defining role but, more often in the clinical context, patients are assessed for heart disease which is minor. The question is whether disease is present at all. In this borderland between health and disease, the interpretation of technological data is inherently uncertain. The diagnosis then depends more heavily on the social utility of particular disease categories. Shifts in diagnostic categorisation are not therefore attributable solely to more extensive forms of cardiac imaging but are socially constructed in an interactive context which involves the technology, the medical profession and the wider social structures which exist at the time of diagnosis. Claims of technological certainty create a social space within which the medical profession generates disease categories. These shifting disease categories may serve the needs of patients but may also be influenced by those of other players. © 1997 Elsevier Science Ltd. All rights reserved Key words--heart disease, diagnosis, medical technology
INTRODUCTION Ryle (1948) pointed out in his classic textbook that diagnosis is the most important task of the physician since it determines both assessment of prognosis and choice of treatment. From the 18th Century onwards, the diagnosis of heart disease has increasingly depended upon technology which is credited with making historic advances in the evolution of knowledge about the heart. In the 19th Century, the stethoscope revolutionised diagnosis. The electrocardiograph, radiograph, cardiac catheterisation, angiography and echocardiography each contributed knowledge about the heart (White, 1951), progressively penetrating the living body to display abnormality and disease. The contribution of cardiac tests to the treatment of organic heart disease is u n d i s p u t e d - - m o d e r n cardiac surgery would be impossible without accurate diagnostic information from cardiac imaging. However, when physicians encounter symptoms such as chest pain, palpitations, breathlessness or even such signs as heart murmurs, these are frequently not manifestations of serious heart disease. Indeed, the diagnostic question is often whether heart disease is present at all. In such cases, diagnosis requires negotiating the borderland between health and disease. With patients in whom heart disease is possible but unlikely, statistical uncertainty is inevitable, doubt is common and there is the risk of diagnostic error. A borderland character-
ised by technical uncertainty presents a prime site for analysing the social subtext of diagnosis. The social construction of disease has been wellargued although sociologists may have emphasised the social context to the partial exclusion of the physical world (Fox, 1993, p. 150). We draw here on Figlio's argument that diseases are socio-clinical entities, arising as a set of symptoms in the context of social change (Figlio, 1982, 1986). The medical profession is authorised to deal with the socioeconomic and political meaning of the symptoms, thus producing a social space within which disease is constructed. Figlio distinguishes between practices which can be precisely defined and those which are less controlled, requiring interpretation through experience. It is the inherent uncertainty and dependence on interpretation which makes the borderland between health and disease significant in identifying the interplay between the physical and social worlds. As a case study we are focusing on the borderland of cardiac diagnosis using imaging technologies. While all diagnostic processes have a borderland, we have selected cardiac imaging because, sociological arguments notwithstanding, it represents the area where clinicians are least inclined to accept the social construction of disease. After all, imaging has proved its worth in planning cardiac surgery and the internal structure and function of the human body is displayed in such vivid detail that it seems counterintuitive to argue that
1043
1044
Jeanne Daly and lan McDonald
these images do not also determine diagnosis on the borderland. Our aim is to argue for a postmodern view of diagnosis which acknowledges ambiguity and which may well benefit clinician and patient alike by an increased emphasis not on technological data but on the care of the patient. However, the patients themselves are absent from our analysis in this case, appearing only as the focus of competing medical views on where, and why, to draw the diagnostic boundary. Our data are drawn from the medical debate on the technological diagnosis of heart disease in the borderland. Such diagnosis has undergone swings of the pendulum in the course of 150 years. Signs and symptoms attributed to heart disease have attracted various labels, including soldiers' heart, the effort syndrome and mild mitral valve prolapse, and these show considerable overlap. Wooley (1976) argues that these are historical manifestations of the same disease. Alternately, the heart could be seen as normal and the symptoms as neurotic. Commonly the change in diagnostic label was attributed to the development of a new diagnostic technology but it is possible to elicit the role of the social context at least in an implicit form. The source of the problem has always been uncertainty over what the data mean. Recognising that technical and social aspects of diagnosis are inextricably entwined, for conceptual purposes we discuss first the technical source of uncertainty in the present time. We then turn to the way in which uncertainty has been interpreted historically to give a changing disease label. Uncertainty has remained, paradoxically perhaps increased, with the availability of diagnostic images of increasing sophistication and detail. Patients in whom heart disease is possible, but unlikely, present in a confusing manner. One group has symptoms like chest pain, breathlessness or palpitations which may indicate organic heart disease or may be manifestations of anxiety; they may even be a healthy cardiac response to fear or stress. In others, doubt has arisen because of a heart murmur heard with a stethoscope or a suspected abnormality in a chest radiograph or an electrocardiogram. Heart murmurs are merely audible blood flow but they are difficult to diagnose. Some murmurs indicate heart disease; more often, especially in the young, they only indicate rapid blood flow. The search is for a technology which can resolve the implications for disease of these overlapping categories. The diagnostic accuracy of modern imaging may now surpass that of direct inspection of the heart at surgery or autopsy. Hence a test such as echocardiography (cardiac ultrasound) is beholden to no "gold standard" against which its pronouncements can be checked. When the "gold standard" for cardiac diagnosis was set by the stethoscope, it was well understood that clinicians had to have con-
siderable skill in interpreting nuances of the heart sounds. Only rarely could they have physical evidence to support the diagnosis. The illusion now is that cardiac imaging, giving a direct visual display of the beating heart, has circumvented the difficulty of interpretation. After all, "seeing is believing". The images are seductive, distracting attention from the fact that diagnostic criteria are still necessary to distinguish normal from abnormal. Thus diagnosis remains a matter of clinical judgement, difficult, prone to variation and thus contentious. A test result of doubtful significance will be called "equivocal" often as a result of lying on the borderline between disease and normality. It is recognised that, with increasing use of the test on well populations, there is an unavoidable epidemiological risk of false diagnosis of disease. What is less well recognised is that there is an even greater risk of an equivocal result. Since there is no other technology to which we can appeal, a false or doubtful result will often stick. If diagnosis is uncertain, so is treatment. We have shown that the best of the noninvasive cardiac imaging techniques, echocardiography actually makes only a minor contribution to the management of well patients (McDonald et al., 1988). If minor disease is diagnosed, antibiotic prophylaxis can be prescribed against bacterial infection of the heart valves but effectiveness is unproven (Anon., 1992). Alternatively, if the heart is found to be normal, anxious patients benefit from reassurance. Unfortunately, the very process of referring the patient for testing raises concern about the heart. A normal test result usually reduces this concern about the heart but it does not necessarily resolve it, nor does it ensure that concern will not recur (Daly, 1989; Daly and McDonald, 1993). In the case of a false or equivocal result, anxiety will be increased. Nevertheless, echocardiography is commonly used on well patients. We found that the reasons were more often social than technical: maintaining professional credibility, satisfying the real or implied request of the referring doctor or satisfying the perceived expectations of insurers. If such social considerations can influence the ordering of the test, might they not also subtly influence the interpretation of the image? We now turn to a more detailed analysis of the overlapping concerns of diagnostic uncertainty and the social construction of disease. Our aim is to identify the contribution made by diagnostic technologies but also to identify the way in which criteria for disease have been responsive to social forces. Our intention is not to present a detailed analysis of the social forces acting at any one time but to establish the need for a more definitive analysis of the social context of diagnosis at the margins of disease.
Cardiac disease construction on the borderland THE HISTORY
It was the stethoscope which first promised to "display" the internal structure of the beating heart: ... the stethoscope, when applied to the organ of hearing, shall convert it into an organ of vision, enabling the listener to observe, with the clearness of ocular demonstration, the ravages which disease occasionally commits in the very centre of the rib-cased cavity of the body (Anon., 1831-1832, p. 122). The stethoscope facilitated the collection of scientific data on heart disease. Since, "[i]n the pursuit of science every truth, every fact discovered is of value" (Corrigan, 1828, p. 588), benefit to patients was relegated to the sidelines. The stethoscope performed its scientific task so convincingly that it came to symbolise the scientific doctor. Diagnostic technologies also strengthened the social credentials of the emerging medical profession with social institutions like employers, the army and insurers by providing an objective means of evaluating evidence from "malingerers, conscripts, applicants for pensions, insurance or society benefits, plaintiffs claiming damages for i n j u r y . . . " (Jackson, 1984, p. 337). Uncertainty dogged the new science. Heart sounds matched "the diversity of hues in a rainbow" (Anon., 1826, p. 471). Some murmurs and heart irregularities were linked by autopsy to fatal disease but murmurs also occurred in the absence of any organic heart disease, the so-called "functional" murmur. Interpretation was difficult: ...the murmur revealed by auscultation became a bugbear; it was misinterpreted as meaning a damaged valve and as ominously threatening an early death.., only too often the physician spread gloom as he warned the parents of the presence of the murmur that threatened disaster (Herrick, 1942, p. 231). Diagnosis thus posed a dilemma for clinicians, a problem clearly perceived by Austin Flint (who identified Flint's murmur): Without the negative proof afforded by physical exploration the mind of the practitioner must be in doubt as to the diagnosis. If he give a decided opinion, it is a guess which may prove to be either fight or wrong. If he avoid giving a decided opinion, the inference which the patient usually draws is that organic disease exists and that the physician is reluctant to tell the truth (Flint, 1888, p. 354). Harm could be done to patients and Flint concluded: " . . . truly fortunate were they who kept aloof from the stethoscope of the auscultator" (quoted in Baldry, 1971, p. 72). The American Civil War
During the American Civil War, the physician, Da Costa, identified a group of soldiers with palpitations, chest pain, rapid, often irregular pulse, indigestion and diarrhoea. Remarkably, despite the carnage of Civil War battles, Da Costa did not associate these problems with fear. Indeed, he noted that one of these soldiers loved campaigning and
1045
refused to leave his regiment. Da Costa concluded that war affected the heart: "It seems to me the most likely that the heart has become irritable, from its over-action and frequent excitement..." (Da Costa, 1871). Soldiers' heart was distinct from the feigned heart disease of malingerers induced by tightly binding the chest. But how was it to be treated? Da Costa recognised that a disease label would deplete and demoralise the army. The compromise he found helped to establish the role of the doctors in the army: Soldiers' Heart was a legitimate "affection" which could progress unless treated with rest and digitalis; after treatment soldiers could be returned to active service. Preventive action included a reduction in forced marches with heavy packs. Da Costa's compromise benefited the army and strengthened the role of the army doctor. Even the soldiers may have been grateful for a rest. The symptoms characterising Soldiers' Heart were also common in the community and recognised to be of considerable practical importance. In this context, however, the problem was seen as neurotic, presenting in women who were hysterics, menopausal or who had a displaced uterus or irritated ovaries. Men too could suffer the symptoms as a result of sexual excesses, over-indulgence in tobacco or the mental worries of business (Bramwell, 1884). Within the civilian population, and after the war, the failure was in the person rather the heart. Treatment focused on changing debilitating personal habits. Two explanations of patients' problems, cardiac and neurotic, legitimate and deviant, could thus coexist, predicated on assumptions about the characteristics of the social group to which the patient belonged. The invention of the electrocardiograph produced tracings of electrical impulses from heart beats. This swung the pendulum towards identifying heart irregularity as disease. Pioneer cardiologist Thomas Lewis saw it as evidence of a pathological condition with the risk of serious irregularities in later years. However, he cautioned that "palpitations" were also experienced as "subjective phenomena" by nervous, young females and in people who were tired or had over-indulged. Such patients should be regularly monitored (Lewis, 1915, pp. 51-54). The two Worm Wars
The First World War brought a reassessment of the significance of murmurs. James McKenzie's 1915 edition of Diseases of the Heart pointed out that the murmur heard during systole, associated with regurgitating blood from the mitral valve, had become a medical "bogy", causing harm to patients, including applicants for life insurance and recruits to the armed services. He argued that such murmurs were "functional", did not affect prognosis and might be precipitated by anxiety or other problems like physical debility (McKenzie, 1914, pp. 326 327). McKenzie's views gained currency after
1046
Jeanne Daly and Ian McDonald
the Allied army suffered major defeats in France. There was a flood of casualties and the military command battled what it saw as a contagious epidemic of soldiers failing to face danger (Butler, 1943). Among this list were soldiers presenting with chest pain, breathlessness, palpitations and an irregular heart beat. Initially these soldiers were diagnosed as suffering from valvular heart disease and disordered action of the heart (DAH) (the name under which Soldiers' Heart had won official military recognition) (Howell, 1985). Lewis, by then, consultant physician in diseases of the heart to the British Army, now reinterpreted "the borderland between health and disease" in a manner "tinged with a war purpose" (Lewis, 1919a, pp. 4, iv). There had been too many rejections from the army and too many pensions. Thus heart sounds lost their significance in favour of an emphasis on short term prognosis (Lewis, undated). DAH was recast as the Effort Syndrome, an exaggerated cardiac response to physical effort in the puny or nervous. Such soldiers were removed from hospitals, put on exercise programmes and half of them were returned to the front. The immediate economic saving was considerable. Howell (1985) argues that Lewis earned a knighthood by excluding soldiers from pensions for heart disease. This redefinition of heart disease helped establish the credentials of the new specialty of cardiology. What cardiologists learnt in war time, they then put into practice in peace time. Lewis argued that the diagnosis of murmurs, particularly murmurs associated with an incompetent mitral valve, was "a diagnosis by fashion only" which lead to too many unwarranted pensions: It is my confident belief that, had systolic murmurs and modifications of the heart sounds never been discovered, the practice of the profession would have stood on a much higher plane today than it actually does... As these signs are used by medical men in general at the present time, they are productive of infinitely more harm than good (Lewis, 1919b, p. 7). This turn of events benefited cardiology but not the life insurance industry which responded by claiming that the evaluation of murmurs was not a clinical question but a matter for "life insurance medicine". After a review of companies' records showed that functional heart murmurs were associated with at least a fourfold increase in mortality, they concluded that no murmur was innocent (McCrudden, 1931) and applicants with murmurs were penalised (Blumenthal, 1942). A psychosomatic explanation was also current. Extra heart sounds during systole were found to be common and associated with "psychoneurosis". The cause of the sounds was unknown but "delicate strands joining pericardium and pleura" were reported (Johnston, 1938, p. 222). Clinicians again faced a difficult choice. Heart disease was associated with loud murmurs but,
since even a faint murmur might indicate rheumatic fever which could severely damage the heart, a murmur could only be called benign if all organic cause had been carefully excluded (Levine, 1937, p. 209). Many schoolchildren with murmurs were prescribed prolonged rest in convalescent homes and hospitals (Garrow, 1943). Alternatively, if murmurs were seen as psychosomatic, prognosis was believed to be unaffected but symptoms were resistant to treatment. Clinicians were treading dangerous territory. The strong fear of heart disease in the community predisposed people to worry about their hearts. Thus a doctor uncertain about functional symptoms could create a "cardiac cripple": "[h]aving made the diagnosis of heart disease one should speak with emphasis and assurance" (Levine, 1937, p. 216). The Second World War swung the pendulum against the diagnosis of heart disease. Murmurs were seen as ambiguous, too uncertain in their significance to provide the basis for a disability pension. There was concern about "unwarranted cardiac invalidism" among children and the need to unify the medical view of murmurs (Evans, 1943). Clinicians were confused: Probably every second of every day some doctor somewhere is confronted with a mitral systolic murmur, and he has got to do something, or nothing, about it... I should like to know what causes the murmur and why it is "innocent"... Anybody talking about "innocent" murmurs should.., give his reasons for calling them so by advancing some physical cause for their occurrence (Gaskell, 1943, p. 582). Not so in the armed services. From January, 1940, Paul Wood was physician to the British army's Effort Syndrome Unit (Wood, 1941). He rapidly recognised Da Costa's "condition" and the Effort Syndrome as the same obsolete concept. The problem, he argued, was fear. To Wood we owe the recognition that the diarrhoea noted by da Costa could be attributable to fear before a battle! Wood, however, did not consider the fear response normal. It occurred in people with a family history of nervousness or insanity and an inferior character caused by having been allowed to "cling too long to their mothers' skirts". Psychoneurosis could be diagnosed with a few pertinent questions: "their distaste for this label may prove quite helpful". Cardiac neurosis was also present in the community: " . . . t h e y are merely clothed differently, the former in battle dress, the latter in nylon" (Wood, 1956, p. 202). Its victims tended to be women, from the "emotional" races (Jews and Italians) and "indifferent to higher ideals". Accepting the patient account could turn a patient into "a chronic and incurable psychoneurotic", so the doctor had to persuade the patient that "he has met a doctor who thoroughly understands his case". Since "[i]t is remarkable what little insight these patients have, and disconcertingly how little shame", a normal diagnostic test provided an important means of
Cardiac disease construction on the borderland exposing an hysterical motive for gain (Wood, 1956, pp. 944-946). In Wood's hands the diagnosis of deviance reached its purest form. In contrast, in the United States army, army psychiatrists found malingering to be rare but symptoms did offer an escape mechanism. They recognised that, in war, all personnel could develop psychoneurosis. Psychiatrists were sent into the field to build up morale and the motivation to fight. The lesson they learnt from the war was that the emotional problems of the individual should be considered "within the context of his group and social culture" (Medical Department, U.S. Army, 1966, Chap. 1). The technological explosion in cardiology The post-war years ushered in an heroic period of intervention in cardiology which depended on accurate diagnosis. Attention returned to murmurs. A series of well patients with late systolic murmurs was subjected to angiography, X-ray imaging of a dye dispersing in the heart from a catheter fed into the left ventricle. In some cases they observed regurgitation back into the atrium indicating a faulty mitral valve (Barlow et al., 1963). Sometimes the leaflet supporting the mitral valve "billowed" or domed (Barlow and Bosman, 1966)--hence mitral valve prolapse. Patients with late systolic murmurs, excluded from life insurance in the 1930s, classified as psychoneurotic during the war years, could now be diagnosed as having mild mitral valve prolapse. Angiography was too dangerous to perform on well patients whose murmurs had to be diagnosed according to labile criteria based on "experience": The point of maximum intensity of late systolic murmurs has ranged, in our experience, from the middle of the murmur to very near the end... Segall and Likoff... have regarded the diamond-shaped late systolic murmur, and especially that initiated by a systolic click, as innocent (Barlow and Bosman, 1966, p. 176). The problem for the clinician worsened (Caceres and Perry, 1967, p. 231): which late systolic murmurs were innocent and which were attributable to mitral regurgitation? Echocardiography introduced safe, noninvasive imaging appropriate for well and ill. As a result, mitral valve prolapse proliferated "in almost every country where clinical investigation is performed" (Wooley, 1976, p. 750). The disease was found in up to 21% of healthy young female volunteers without the murmur (Markiewicz et al., 1976). Mild mitral valve prolapse had no cause, treatment or cure and an unknown prognosis (Criley et al., 1966). However, penicillin could now be used to prevent bacterial infection of an incompetent, regurgitating mitral valve. The spread of heart disease in the well study population is worth noting (Markiewicz et al., 1976). Forty nine out of the 100 well volunteers had heart murmurs and, usually as a result, 10 had previously been diagnosed as having a "heart problem". Symptoms included 29 with
1047
chest discomfort, 30 with palpitations and nine with breathlessness. A diagnosis of mild mitral valve prolapse was thought to account for both signs and symptoms. The tide was to turn again in the 1980s. The reasons are not clear. One possibility is that economic recession induced a more critical approach to technological proliferation. There are also signs of the by now familiar clinical backlash against an "epidemic" disease of little benefit and potential harm to patients. The significance of murmurs was reassessed. Nishimura et al. (1985) published a long-term follow-up of echocardiographically diagnosed mitral valve prolapse. It was argued to be benign in the majority of cases, especially in the young and healthy. Only a minority of patients had serious disease which might require valve replacement. An editorial in the New England Journal Of Medicine (Wynne, 1986) identified the same two groups: the high risk group and a group "no more symptomatic than their peers without the diagnosis". The epidemic of mitral valve prolapse in the well was attributed to the belief that echocardiography could "distinguish normal from abnormal". More stringent criteria were needed to exclude "previously silent normal variants of structure and function" displayed by the test. Apart from identifying a culprit, the technology, there was little explanation offered for the change in criteria except to argue that mitral valve prolapse was a "disease of fashion", the result of a test having "its day in the sun" and seductive since it provided a "catchall explanation of a number of frustrating complaints". DISCUSSION AND CONCLUSION
We accept the efficacy of cardiac technology in the diagnosis of manifest heart disease. Mitral valve prolapse, for example, is an anatomically distinct and potentially serious disease in a small minority. Diagnosis, however, occurs along a continuum from serious disease through manifestation of uncertain significance to normality. We have focused on the latter end. While medical discourse has, for 150 years, emphasised the need for technical certainty in diagnosis, the intraprofessional discourse of diagnosis at the margins has been about how best to give meaning to the image. Technological "advance" notwithstanding, it has been impossible to draw an exact line between structurally sound and diseased hearts. For the unwary, Thomas Lewis (1919a) (p. 61) said, "the whole of this ground is filled with craters, craters filled with memories of battles old and recent fought over them". It remains true today. The criteria for diagnosis in the borderland between health and disease are inherently uncertain. Technologies have produced ever more detail. Indeed, given modern imaging potential, it is not clear who can be confident of remaining normal.
1048
Jeanne Daly and lan McDonald
But who has disease? A diagnostic label derives from the image. However, it also involves clinical judgment about where to draw the line between disease and health in the case of particular patients. This is where the social context of diagnosis comes powerfully to the fore. Historically the interests of the medical profession itself and of life insurers have promoted the identification of disease. The interests of the army, pension schemes and perhaps the state as health insurer have been served by denying disease or by promoting alternative explanations such as cardiac neurosis. Recognition of the clinicians' responsibility in caring for patients has often been referred to, more as a focus of concern than as a central issue of debate. The way in which these conflicting forces are resolved in practice is obscured by the smoke screen of technological certainty. In part, we would argue, there is the need for better science. Scientifically, we need more rigorous criteria for the categorisation of heart disease according to the prognosis associated with particular signs or symptoms. However, especially in the case of a steady increase in imaging capacity, uncertainty about the interpretation of the image is inevitable. Scientifically, this carries certain implications for practice. Firstly, on the borderland of disease, the search for a definitive image is misplaced. Secondly, the emphasis on certainty obscures the substantial contribution to diagnosis made by the social context of patient and doctor. While patients may be malingering, it is equally true that doctors may be acting out of personal interest or under the influence of powerful social institutions. In the absence of clear knowledge of the consequences of diagnosis based on such interests, they provide a decidedly insecure scientific foundation for the diagnosis of disease or deviance. Scientifically, we need to subject diagnostic practice to critical scrutiny, including the social implications for the patients so diagnosed. Otherwise, there is no option but to have recourse to the unsatisfactory, unscientific explanation of blaming a prevailing "fashion". It is also surely bad science to harm patients. Labelling patients with disease may cause harm (Haynes et al., 1978) especially when there is no treatment for the condition diagnosed; patients whose distressing symptoms are attributed to their own deviance are clearly not helped. Such problems of diagnosis on the borderland remain unresolved and they will not be resolved while the only response is further incantation of the diagnostic capacity of imaging technologies. On the contrary, we could turn to an alternative direction. The uncertainty of diagnostic data in the borderland presents clinicians with a dilemma. They worry when they take action based on uncertain knowledge which may involve doing harm to their patients. The discourse which carries these concerns is subjugated by the warning that they must appear
certain or risk turning their patients into "cardiac cripples"; instead they must take responsibility for exerting control over patients' problems. A public acknowledgment of the inherent uncertainty of diagnostic decisions in the borderland would weaken this control over patients' problems, it is true, but it would open a new social space within which a different medical discourse could grow. It would emphasise ambiguity instead of certainty, negotiation instead of control and provide better opportunity for doctor and patient together to negotiate the meaning of the patient's problem. We end by echoing words from what Fox (1993) (p. 160) calls a "postmodern prayer": Where there was truth May we celebrate ambiguity Where there was control May we be generous.
REFERENCES
Anon. (1992) Chemoprophylaxis for infective endocarditis: faith, hope, and charity challenged. Lancet 339, 525526. Anon. (1831-2) Review. Lancet 2, 122. Anon. (1826) Review from the West. Lancet 9, 471. Baldry, P. E. (1971) The Battle against Heart Disease. Cambridge University Press, Cambridge. Barlow, J. B. and Bosman, C. K. (1966) Aneurysman protrusion of the posterior leaflet of the mitral valve. American Heart Journal 71(2), 166-178. Barlow, J. B., Pocock, W. A., Marchand, P. and Denny, M. (1963) The significance of late systolic murmurs. American Heart Journal 66(4), 443-452. Blumenthal, J. S. (1942) The systolic murmur. Annals of Internal Medicine 17, 637-641. Bramwell, B. (1884) Diseases of the Heart and Thoracic Aorta. Young, J. Pentland, Edinburgh. Butler, A. G. (1943) The Australian Army Medical Services in the War of 1914-1918, Vol. III. Australian War Memorial, Canberra. Caceres, C. A. and Perry, L. W. (1967) The Innocent Murmur." A Problem in Clinical Practice. Little and Brown, Boston. Corrigan, D. J. (1828) Aneurism of the aorta. Lancet 1, 588. Criley, J. M., Lewis, K. B., O'Neal Humphries, J. and Ross, R. S. (1966) Prolapse of the mitral valve: clinical and cine-angiographic findings. British Heart Journal 28, 488-496. Da Costa, J. M. (1871) On irritable heart: a clinical study of a form of functional cardiac disorder and its consequences. American Journal of the Medical Sciences. 33, 40. Daly, J. (1989) Innocent murmurs: echocardiography and the diagnosis of cardiac normality. Sociology of Health and Illness 11(2), 99-116. Daly, J. and McDonald, I. (1993) The Social Impact of Echocardiography: Opening Pandora's Box. Australian Government Publishing Service, Canberra. Evans, W. (1943) Mitral systolic murmurs. British Medical Journal January 2, 8-9. Figlio, K. (1982) How does medicine mediate social relations? Workmen's Compensation Practices and medico-legal practices, 1890-1940. In The Problem of Medical Knowledge: Examining the Social Construction of Medicine, eds P. Wright and A. Treacher, pp. 174-
224. Edinburgh University Press, Edinburgh.
Cardiac disease construction on the borderland Figlio, K. (1986) Medical diagnosis, class dynamics, social stability. In Science, Technology and the Labour Process, eds L. Levidow and B. Young, pp. 129-165. Free Association Books, London. Flint, A. (1888) A Treatise on the Principles and Practice of Medicine. Designed for the Use of Practitioners and Students of Medicine, 6th Edn. J. and A. Churchill, London. Fox, N. J. (1993) Postmodernism, Sociology and Health. Open University Press, Buckingham. Garrow, R. P. (1943) Unwarranted cardiac invalidism in children. British Medical Journal 23 January, 114--115. Gaskell, H. S. (1943) Mitral systolic murmurs. British Medical Journal g, 582. Haynes, R. B., Sackett, D. L., Taylor, D. W., Gibson, E. S. and Johnson, A. L. (1978) Increased absenteeism from work after detection and labelling of hypertensive patients. New England Journal of Medicine 299, 741. Herrick, J. B. (1942) A Short History of Cardiology. Charles C. Thomas, Springfield. Howell, J. D. (1985) "Soldier's Heart": the redefinition of heart disease and specialty formation in early twentiethcentury Great Britain. In The Emergence of Modern Cardiology, eds W. F. Bynum, C. Lawrence and V. Nutton, pp. 34-52. Wellcome Institute for the History of Medicine, London. Jackson, E. (1984) Report on the values of objective tests for the determination of Ametropia, Ophthalmoscopy, Ophthalmometry, Skiascopy. Journal of the American Medical Association 88, 337. Johnston, F. D. (1938) Extra sounds occurring during cardiac systole. American Heart Journal 15, 222. Levine, S. T. (1937) Clinical Heart Disease. W. B. Saunders Company, Philadelphia. Lewis, T. (1915) Clinical Disorders of the Heart Beat." A Handbook fi,r Practitioners and Students. Shaw and Sons, London. Lewis. T. (1919a) The Soldier's Heart and the Effort Syndrome. Shaw and Sons, London. Lewis, T. (1919b) On cardinal principles in cardiological practice. British Medical Journal November 15, p. 7.
1049
Lewis, T. (undated) Prognosis in cardio-vascular affections: prognosis with special reference to the assessment for pension purposes and to the capacity of the worker. British Medical Association. Markiewicz, W., Stoner, J., Lomdon, E., Hunt, S. and Popp, R. (1976) Mitral valve prolapse in one hundred presumably healthy young females. Circulation 53(3), 464473. McCrudden, F. H. (1931) Heart murmurs and insurance. New England Journal of Medicine 204(12), 598-600. McDonald, I. G., Guyatt, G. H., Gutman, J. M., Jelinek, V. M., Fox, P. and Daly, J. (1988) The contribution of a non-invasive test to clinical care: the impact of echocardiography on diagnosis, management and patient anxiety. Journal of Clinical Epidemiology 41(2), 151-162. McKenzie, J. (1914) Diseases of the Heart, 3rd Edn. Henry Frowde, London. Medical Department, United States Army (1966) Neuropsychiatrv in Worm War 11." Volume I, Zone of the Interior. Offices of the Surgeon General, Washington. Nishimura, R. A., McGoon, M. D., Shub, C., Miller, F. A., Ilstrip, D. M. and Tajik, A. J. (1985) Echocardiographically documented mitral-valve prolapse: long-term follow-up of 237 patients. New England Journal of Medicine 313(21), 1305-1309. Ryle, J. A. (1948) The Natural History of Disease, 2nd Edn. Oxford University Press, Oxford. White, P. D. (1951) Heart Disease, 4th Edn. Macmillan, New York. Wood, P. (1941) Aetiology of Da Costa's syndrome. British Medical Journal 7 June, 845-851. Wood, P. (1956) Diseases of the Heart and Circulation, 2nd. Edn Eyre and Spottiswoode, London. Wooley, C. F. (1976) Where are the diseases of yesteryear? Da Costa's syndrome, soldier's heart, the effort syndrome, neurocirculatory asthenia--and the mitral valve prolapse syndrome. Circulation 53(50), 749-751. Wynne, J. (1986) Mitral-valve prolapse. New England Journal of Medicine 314 (9), 577-578.