Patient non-compliance: deviance or reasoned decision-making?

See. Sci. Med. Vol. 34, No. 5. pp. K~o7-513. 1992 Printed in Great Britain. All rights reserved PATIENT Copyright NON-COMPLIANCE: DEVIANCE DECISION-MAKING? 0277-9536192 65.00 + 0.00 Q 1992 Pergamon Press plc OR REASONED JENNY L. DONOVAN’ and DAVID R. BLAKE~ ‘Health Care Evaluation Unit, University of Bristol, Canynge Hall, Whiteladies Road, Bristol BS8 2PR, U.K. *Boneand Joint Research Unit, Royal London Hospital Medical College, ARC Building, Ashfield Road, London El 2AD. U.K. large quantity of research concerning issues of patient compliance with medications has been produced in recent years. The assumption in much of this work is that patients have little option but to Abstract-A comply with the advice and instructions they receive. Studies have shown, however, that between one third and one half of all patients are non-compliant, but different authors cite different reasons for this high level of non-compliance. In this paper, the concept of compliance is questioned. It is shown to be largely irrelevant to patients who carry out a ‘cost-benefit’ analysis of each treatment, weighing up the costs/risks of each treatment against the benefits as they perceive them. Their perceptions and the personal and social circumstances within which they live are shown to be crucial to their decision-making. Thus an apparently irrational act of non-compliance (from the doctor’s point of view) may be a very rational action when seen from the patient’s point of view. The solution to the waste of resources inherent in non-compliance lies not in attempting to increase patient compliance per se, but in the development of more open, co-operative doctor-patient relationships. Key w&r-physician-patient relationships, compliance, doctor-patient communication, lay beliefs, paiient participation _ zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA I N T RODU CI T ON Patient compliance with medical instructions has provided the focus for enormous amounts of research work in medicine and social science. It has been estimated that about 4000 English language articles have been published up to 1985 [ 11,and a further 4000 have been listed on Medline alone up to 1990. This proliferation of work has found that non-compliance exists on a large scale (about one half of all patients are thought to be non-compliant [24]), but investigations into the causes of non-compliance have proved inconclusive, even contradictory. It is not the purpose here to review the massive literature about non-compliance-this has been done well elsewhere (see e.g. [S, 61). It is the aim of this paper to show that much of the work about non-compliance, suggesting that patients are too ignorant to understand medical instructions or that they forget large proportions of what they are told, is misplaced. Compliance is, itself, a value-laden term, closely entangled with issues surrounding the dominance of medicine and concerns about costs. For patients, compliance is not an issue: they do not perceive taking drugs entirely in terms of obeying the doctor’s orders. Instead, they weigh up the costs and benefits of taking particular medications as they perceive them within the contexts wd constraints of their everyday lives and needs. The term ‘compliance’ is itself an interesting one. To comply is to obey, submit, defer or accedeto instructions. Much of the work looking at compliance (or non-compliance) carries the implication that patients should comply; indeed that they have little option but to do so, and that in an ideal world, non-compliance would not occur. There is a tendency in much of the work to assume that medical practitioners deliver information clearly and that any failure to comply is thus the fault of the patient or caused by problems with the doctor-patient relationship [4,5,7-10). Compliance is closely tied to the dominance of medicine (1 l-141. An historical analysis has shown that what clinicians now refer to as compliance used to be presented more overtly as physician control [1] and this author goes on to describe compliance as an ideology, “a system of shared beliefs that legitimate particular behavioural norms and values at the same time that they claim and appear to be based in empirical truths” [1, p. 13001. This ideology portrays non-compliance as deviant behaviour and ensures that the blame for it is directed largely towards patients. It is patients who fail to comply, intentionally or unintentionally, because they are ignorant or forgetful (see e.g. [4, 151). Some researchers have looked more closely at the doctorpatient relationship and its effect on compliance. There have been very many articles suggesting changes to it that might be employed to improve doctor-patient communication and thus increase compliance: by increasing doctors’ friendliness and approachability 19, 161; encouraging patients to ask more questions [ 171; encouraging co-operation between doctors and patients [18, 191; encouraging doctors to be more patient centred [20-221 or empathetic [23]; or improving doctors’ teaching skills 171. 507 JESNY L. DONOVANand DAVIDR. BLAKE 508 The 1970s and 1980s have witnessed an explosion in consumer demands for information about medical treatments. Today’s patients want more information and greater opportunities to be active in their dealings with medical practitioners (see e.g. [24, 271). This view is at odds with the approach of much of the research in patient compliance, where patients are seen to be passive and powerless. A small number of studies have attempted to look at compliance from the patient’s perspective. They have found, for example, that some patients choose not to comply with medical instructions as a way of expressing their attempts to cope with their disease[2]; as a reaction to the way they have been treated by doctors [3]; or as a way of fighting the system by breaking its symbolic rules [28]. Some papers thus acknowledge that patients are able to choose deliberately not to comply with (or adhere to Refs [7, IO]) medical advice, but what is missing from much of the work is an understanding of the ways in which patients think and feel about their illnesses and treatments and how these impact on their behaviour. For although doctors remain the gatekeepers for drug provision in terms of legal power and knowledge, patients do retain, and always have, the ability ultimately to decide what happens to the doctors’ orders: whether or not they take the drugs prescribed and in what ways and quantities. It is the aim of this paper to examine patients’ reactions to advice and medications prescribed by doctors in rheumatology clinics. Although the work is confined to rheumatology patients, it seems likely that the findings will be applicable to other specialties. The study seeks to show that patients are not, on the whole, passive or powerless. Indeed, the patients in this study indicate that they are quite capable of making choices about treatments and lifestyles rationally within the contexts of their beliefs, responsibilities and preferences. These choices draw on the information at their disposal which can be quite different from that available to medical staff. Non-compliance may thus not be deviance, but reasoned decision-making. Firstly, the methodology of the study will be outlined, followed by an assessment of the levels of non-compliance, and finally there is a discussion of the issues that surround the concept of patient compliance and its relevance to the treatment of patients today. METHODS Consultant rheumatologists in three rheumatology units examined their letters of referral from general practitioners (GPs) over a six month qriod and selected patients with a suspected inflammatory arthropathy. The aim of the study was to uncover the patients’ views of the clinic process and so a-qualitative methodology, drawn from anthropology and interpretative sociology, was employed. This involved intensive interviews with patients, and the obser- vation and tape-recording of consultations with rheumatologists. Patients were interviewed in their own homes before their first appointments in order to obtain detailed base-line data about their perceptions of their illness and treatments and their expectations of the clinics. Interviews were semi-structured with openended questions, allowing patients to discuss matters important to themselves, but also including a checklist to ensure that relevant issues were covered. These included such matters as the onset and aetiology of the joint problems, experience of symptoms, the effects of arthritis on life (employment, housework, childcare, hobbies etc.), experience of previous treatments (conventional and alternative), patients’ expectations of the clinic, and their future health. Fifty-four patients make up the bulk of the study, 39 of whom were diagnosed finally as having an inflammatory arthropathy such as rheumatoid arthritis or ankylosing spondylitis, and the remainder having osteo-arthritis or more minor rheumatological complaints. Consultations between patients and doctors were observed and tape-recorded, and patients were interviewed again at home after each consultation. In these interviews, patients were asked about the treatment they had received, what complaints or praise they had for the clinic and/or staff, how much they could recall of what the doctor had told them, whether or not they intended to take the advice and treatments offered (and why), and other issues of importance to individual patients. Patients were interviewed and observed until they were discharged or until the end of the study (between 3 months and 3 years). The interviews and consultations produced a large amount of taped conversation (over 100 hr), all of which was fully transcribed. The transcripts were then repeatedly scrutinised for shared themes which formed the basis of categorisation and theoretical development (see e.g. [29-311). The material was analysed in several ways: as casestudies of individual patients, longitudinally, and comparatively across patients and events. The results are largely qualitative. Some quantification is possible, principally to illustrate the strength of descriptions, but the main aims of the study, to understand the perceptions, experiences and behaviours of patients within the contexts of their lives are not easily quantifiable and so are presented descriptively. Lesels of non -compliance Compliance is not a simple matter of obeying Non-compliance exists on different instructions. levels and is expressed in different ways. It is also not just concerned with drugs, although this is the most commonly studied form. In this study, it was possible, occasionally, to observe patients talking with nurses and therapists, and it was clear from the interviews that these patients received copious amounts of ad- Patient noncompliance 509 vice (sometimes conflicting) about rest, exercise, gadgets and diet from health professionals, alternative therapists, friends and the media. The primary focus in this study was, however, on the information exchanged between patients and doctors, and so a large quantity of the data concerns compliance with medications and braces/supports. More than four-fifths of all the patients spontaneously expressed their dislike at having to take drugs at all: I don’t like taking drugs. If there was something else I could take or do instead of taking drugs, then I would. (woman aged 49 years with sero-positive rheumatoid arthritis.) ing the rheumatic disease process, but which have to be taken for at least six weeks for effects to be noticed. Four patients failed to comply with their second-line prescriptions: two altered the doses of their drugs and then one of these and three others stopped taking them altogether. At no time did they discuss their altering of doses with doctors, nor did they inform anyone when they had stopped taking the drugs. Their reasons for not complying were because of side effects: nausea, vomiting, feeling “grotty”, and a “dirty” tongue: Almost every patient said that they would like more information about the drugs they were taking and any others that might be suitable. Nearly one half of the patients (25) could be considered to be noncompliant in that they admitted to failing to take drugs according to their prescriptions (similar figures have been found elsewhere [24]). Patients were asked about the drugs they had been prescribed and most produced the bottles they had been given so that actual prescriptions could be noted and compared with what they said they actually took. There were several levels of non-compliance within the group of 25 non-compliers. Ten patients no longer took the drugs they had been prescribed. Three of these had taken the drugs on a lower dose for a while before giving them up; the others gave them up after a period of time at the prescribed dose. The most common level of non-compliance (by 13 patients), was to take fewer tablets than prescribed, usually by reducing doses by an arbitrary amount (often by one half) or taking tablets fewer times per day. A much less common type of non-compliance involved taking more tablets than prescribed-one woman took two or three Brufen tablets when she needed them rather than the one, three times per day prescribed; and another women took Co-Proxamol prescribed for her husband. The different levels of non-compliance indicate that these patients were thinking about their drug-taking. It was not just a matter of obeying the instructions or not: they experimented with dosages and timing. The most common reason given for not taking the drugs or dosages prescribed was the fear of side effects, cited by nearly two thirds (16) as their primary reason for failing to comply. During the study, 41 patients had taken NSAIDs (non-steroidal anti-inflammatory drugs), and of these 15 complained of side effects, particularly gastro-intestinal problems. Ten were non-compliers: four reduced their doses; six gave up the drugs altogether: Feldene worked OK but they make me feel sick and diz& I just stopped taking them hecause I couldn’t do anything when I was feeling giddy.. . I’d rather just cope with_the pain than the pain and the side effects as well. (Woman aged 36 than feel like that. That’s not to say [the Sulphasalazine] didn’t ease [the arthritis] because it did, but to feel the way I did was useless, it wasn’t worth it. (woman aged SOyears with sero-positive rheumatoid arthritis.) I’ve stopped taking the Sulphasalazine completely. I felt really ‘drastic’ and I thought I would sooner have the pain years with Reynaud’s phenomenon.) There is evidence from this study and another [32] that patients comply more fully with second-line therapies than others. This is chiefly because doctors spend a lot of time explaining these therapies and gaining patients’ co-operation. Patients are able to understand and recall this information, and then decide whether or not to participate. Those that do are then willing to take the medications as prescribed because they understand the reasoning behind the regimen. With NSAIDs, information given out by GPs and Rheumatologists is very limited, because, for doctors, such information is taken to be obvious. For patients, however, details about the optimum times to take tablets, the advantages of taking them regularly, the risks of gastro-intestinal side effects, and the improvements that can come from changing brands are largely unknown. Many patients who did not comply fully with their medications did so to reduce their fears of side effects. They reasoned that there would be fewer side effects with fewer tablets, and so kept their doses to the minimum they could. In some cases, this meant them putting up with considerable amounts of pain and discomfort, or, in the case of second-line treatments, could lead to the drugs being ineffective. Most of those who did comply were also afraid of possible side effects, but these fears were outweighed by their perceived need for relief from pain and stiffness. A quarter of all patients said they were not compliant because they were afraid that they would become dependent on drugs, or that their bodies would become accustomed to drugs which would eventually lose their effectiveness. Fourteen patients were given braces and nine resting splints to support their wrist joints. Of these, only three complied with the advice to wear braces, and only one the resting splints. Those who did not comply complained that the supports aggravated problems in other joints, were tatty, ugly and cumbersome, and many were unsure of the reasons for wearing them: Nineteen patients were prescribed second-line drugs that are thought to have an effect in suppress- They are a nuisance in some ways. They do keep [your hands] in the right position, but it makes it inconvenient 510 zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA JENSY L. DOF~OVANand DAVID R. BLAKE because it is only your fingers you can move-you can’t use This man was unwilling to follow the programme your hand properly. (Woman aged 49 years with sero-posiof exercises devised for him by clinic staff because his tive rheumatoid arthritis.) zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA Non -compliance.. deuiance or reasoned decision - making? As indicated above, the implication in much of the research on compliance is that patients ought to comply, but do not because of ignorance, forgetfulness, or as a reaction to their disease and treatments. The actual rates of non-compliance are reported to be high: about one third to one half of all patients are reputed to be non-compliant according to this study and others [2-4]. If this is true for all patients in all specialties, then non-compliance may be at least as, if not more, common than compliance, and it is difficult then to view such behaviour as deviant or even abnormal. In fact, rates of non-compliance are high because patients do not perceive compliance to be an issue. The advice from doctors has to compete with other advice and perceptions before patients decide for themselves what to do about their treatments. Once out of the consulting room, patients find that many other things impinge on their decisions about treatments. The organised settings for recommending treatment are far remote from where the treatments themselves are implemented [33]. The patients in this study were active in their non-compliance (see also Refs [2, 31). They did not forget or misunderstand what the doctor had said or what the prescription itself indicated, but they chose to ignore advice or alter doses: When [the arthritis) is bad, I take two [Feldene] because that’s how it’s directed, but otherwise, if I get a twinge, I only take the one. There is days I don’t take any. I’d rather not take them if I can. (Woman aged 76 years with sero-positive rheumatoid arthritis.) In doing so, they draw on the information they have at their disposal, based most importantly on their lay beliefs and experiences, but also on information gleaned from GPs, the media, and clinic and pharmacy staff. In another example, a man aged 18 years had decided that he had gout, as described in a family medical book. His GP had diagnosed rheumatoid arthritis (RA), but the man was unable to accept this as he felt his symptoms to be quite different from his grandmother who had advanced RA. When he attended the rheumatology unit, the registrar told him that he had ankylosing spondylitis, that he should take NSAIDs regularly and do exercises for his back. The young man’s response a week later was: The doctors say I have got to do exercises, but I don’t want to do a load of boring exercises for the rest of my life.. I take the tablets when I feel like it. I don’t like having to carry a bottle of pills round with me all the time.. .I’m not sure I’ve got this spondylitis thing, anyway. All the symptoms I’ve had seem lo point to gout, so I’ll stick lo that one. (Man aged 18 years with ankylosing spondylitis.) lay beliefs precluded the diagnosis of ankylosing spondylitis. Clinic staff commented that he was lazy and unto-operative-they did not elicit his reasons for non-compliance. Lay beliefs have recently become a topic of academic study (see Refs [3&36]). Lay beliefs are derived from an individual’s history, background and culture, and are firmly grounded in the social and economic contexts in which people live. They contain many common-sense theories for coping with everyday life, including the onset of illness and appropriate treatments. Within lay beliefs are many ideas about health and illness which are usually internally consistent, but may be quite different from modern biomedicine (35, 361. Most patients had theories about their joint problems and treatments before they came to the specialist clinic, revealed in their first interviews. When they were offered treatment and advice in the clinics, patients carried out what amounted to a cost-benefit analysis of each item. The costs mentioned explicitly included the unpleasantness or stigma of having to take drugs or wear supports, the necessity of having to attend a clinic regularly for blood and urine tests, and the perceived risks of side effects and dependence on drugs. Benefits included immediate improvements in symptoms or the promise of relief in the longer term. The most popular treatments were those with easily discernible benefits and fewest costs, such as ‘one-off’ applications of heat (hot wax or flannels) or cold (ice), or steroid injections. The least popular were drugs which had to be taken regularly or for indefinite periods, and that often resulted in few, if any, symptomatic improvements, such as second-line therapies (e.g. Sulphasalazine, Penicillamine). Patients carried out their cost-benefit analyses in theory and many put them into practice. Some would decide that the risks outweighed the benefits of drugs so clearly that they would not even try the medications, disposing of scripts or made-up drugs: I had Naprosyn and they made not much, so 1 stopped them. from the tablets didn’t balance risk was higher than the result. osteo-arthritis of the knees.) [the arthritis] a bit To me, the odds, the risks of taking (Woman aged 43 easier, but the results them. The years with In most cases, however, drugs would be taken ‘on trial’ for whatever period seemed most appropriate to patients. Drugs such as NSAIDs would be tried for a few days, and if clear beneficial effects were not perceived or side effects were experienced, they would be stopped or the doses reduced: [the doctor] gave me anti-inflammatory tablets. I tried them for a week and they didn’t seem to help at all and I don’t particularly like taking tablets anyway, so I stopped taking them a week after I started. (Man aged 39 years with an early inflammatory arthropathy.) In many cient cases, chance patients to work, did not but most give were the drugs unaware suffiof this. Patient noncompliance Patients taking second-line drugs were more likely to be compliant for longer because they were made aware by clinic staff of the need to take them for extended periods, and they were usually non-compliant because of the experience of side effects. Patients complied with medical advice when it made sense to them and seemed effective. It made sense if it accorded with their own lay beliefs, and was possible to carry out within the constraints of their everyday lives. For some patients, taking tablets regularly did not fit into their lifestyles or the ways that they thought about themselves. The two youngest patients, for example, aged 18 and 23 years, did not want to be seen to be taking tablets regularly and so often reduced doses or skipped taking them. An older woman with severe rheumatoid arthritis was too busy at work to eat regular meals, and as she had been told she must take her tablets with food, did not even have her several prescriptions made up. She also found it impossible to wear wrist braces because they contravened health and safety regulations. There was a general belief that to rely on drugs was a sign of weakness, and most only took drugs regularly as a last resort, when their symptoms were more unacceptable than the risks of side effects and possible dependence: .. I don’t like taking [tablets]. I’d much rather not, but I’m not prepared to just give up and stop working and stop living just because I’ve got something wrong with me. If taking painkillers is a way of living a normal life, I want to live it to the full and so I take them. (Woman aged 26 years with a post-viral inflammatory arthropathy.) I have two tablets at night and I put two by the bed if I need them. But I don’t take them unless it is absolutely necessary. It’s no good taking tablets when you’re not in pain because when you are, your body will be used to them. (Woman aged 64 years with sero-positive rheumatoid arthritis.) Many of their fears about drugs sprung not from medical advice, but from their own or others’ experiences of side effects, or from the publicity about withdrawn drugs such as Opren. For those on second-line therapy, a reduction in dosage could diminish the long term effectiveness of the drug. Some of those with severe rheumatoid arthritis who reduced their doses of NSAIDs experienced considerable amounts of pain and stiffness as a result. For these patients, better information about the drugs might encourage them to take more or to alter the timing of their drug-taking, thus obtaining greater control over their pain and mobility. A small number of patients did eventually do this, either on the basis of advice from others or according to their own theories: Sister told me to take another tablet later in the evening, just before I go to bed and it has made a real difference. Pm a-going to take it regular. (Woman aged 57 years with sero-positive rheumatoid arthritis.) I’m supposed to take these Naprosyn tablets in the morning and evening. Well, I delay the morning one for as long as I can so that I’ll be all right at work in the afternoon. If I take it first thing, I feel like I need one as soon as I get there, 511 but if I wait I can just about last out ‘til tea. (Woman aged 49 years with sero-positive rheumatoid arthritis.) Many patients wanted to feel that they were helping, or at least doing something for themselves. Altering doses of drugs, or choosing whether or not to take them or wear splints, represented ways of regaining some control over their illness. Approximately one half of the patients also admitted to using some sort of alternative remedy, sometimes instead of, sometimes as well as, medications. Alternatives, such as kelp, cod liver oil, homeopathy were perceived to be harmless and natural: Even if [feverfew] doesn’t do me any good, it won’t do me any harm. It’s a natural thing. (Woman aged 58 years with rheumatoid arthritis.) a third of patients expressed an interest in altering their diets as another way to help themselves, although only a small number (4) were able to put this into practice. Almost all of these patients expressed a desire for more information at some time in their interviews. There were only three patients (all elderly) who wanted the doctor to take control of their illness; the remainder wanting more information about, for example, their disease (what was happening inside their joints, its aetiology), available treatments, side effects of drugs, their prognosis. In the interviews, many of these patients were able to express their information requirements succinctly: Over I would like to know what is actually wrong with me and some more about it, like (a) what caused it, and (b) what the future is-whether it is treatment or how it is likely to progress, or if it is going to fade away. (Woman aged 46 with inflammatory symptoms.) In the unfamiliar surroundings of the clinic, however, these questions are rarely asked, but they are not forgotten: I used to think of so many questions that I wanted to ask, and when I got to the hospital, but the time you get there and are doing what they say, you don’t get round to asking. (Woman aged 49 with sero-positive rheumatoid arthritis.) Concern about waste and its financial consequences has led to further interest in patient compliance. Non-compliance is, after all, an archetypal example of wasted resources: thousands of prescriptions being made up but then disposed of or taken incorrectly, resulting in enormous financial costs to the health service (perhaps f300 million per annum in 1986 (371)and little patient benefit in terms of cure or alleviated pain and discomfort [38]. The argument follows that the ability to reduce non-compliance (or increase compliance) would thus save the NHS many millions of pounds, or at least ensure that money was well spent. CONCLUSION It is clear that non-compliance with prescriptions and medical advice is commonplace. The major JENSY L. zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA DON OV AN and DAVID R. BLAKE 512 zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA reason for this is that patients, particularly those with chronic, incurable diseases, make the best decisions they can without worrying about compliance. In effect, patients carry out their own cost-benefit analyses for each treatment they are offered. They weigh up the expected benefits (usually symptomatic relief) against the severity of their symptoms and the perceived risks of treatment (side effects, dependence, time and effort involved, stigma etc) according to their lay beliefs and the information at their disposal. They are not “blank sheets” when they arrive at clinics [39]. They have many beliefs and theories which suggest courses of action, and these are moderated by information from others, particularly family members, medical staff and the media. The key to improving rates of compliance, (although effectively doing away with the concept), is the development of active, co-operative relationships between patients and doctors. For this to be successful, doctors will need to recognise patients’ decisionmaking abilities, to try to understand patients’ needs and constraints, and to work with patients in the development of treatment regimes. For their part, patients will need to make more explicit their needs and expectations, and particularly how they reach their decisions about treatments. Most patients crave more information about their disease and treatments; information that would enable them to make informed decisions about treatments They take decisions now, but often do so with inaccurate knowledge about side effects and the most effective ways to take drugs. The provision of simple information would enable patients to make choices that both fit into their lives and beliefs, and also take into account current medical views about the risks and benefits of drugs. Such information is often not made available at specialist clinics because it is assumed that patients either do not want to know it or know it already [27], or because it seems too obvious to clinic staff. In many cases, it is not at all obvious to patients. The sorts of things patients want to know include the aetiology, symptoms, methods of diagnosis, and likely prognoses of the disease itself; the nature, effects on symptoms and/or disease, and the rates of side effects of all prescribed drugs (including the most common, least explained ones); and the various treatment options and self-help techniques that are available. Such information can be provided simply and cheaply within the normal consultation (as information about second-line therapies was in this study), in the form of leaflets, by a nurse or non-medical provider, or most effectively by a combination of these (401. With the increasing emphasis on patients as consumers and medical staff as providers in the White Paper, and continual worries about costs, concerns about compliance/non-compliance should be laid to rest. There are many potential opportunities for improving patient care, increasing patient and doctor satisfaction with that care. and increasing cost savings through providing the information that patients need to be able to work with doctors in deciding on their treatment regimens. Perhaps the issue now should not be compliance, but how medical staff can understand and participate in the decisions that patients already take about their medications! Acknowledgements-The project, from which this paper is taken, was funded by the Arthritis and Rheumatism Council. I would like to thank the ARC, and the medical and clinical staff who participated in the study, and, of course, the patients who made it possible and enjoyable. Thanks are also due to Bill Fleming, and to Jo Coast and the anonymous reviewers of this journal for their helpful comments. REFERENCES 4. 5. 6. 7. 8. 9. IO. II. 12. 13. Trestle J. A. Medical compliance as an ideology. Sot. Sci. zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPO M ed. 27, 1299-1308, 1988. Conrad P. The meaning of medications: another look at compliance. Sot. Sci. M ed. 20, 29-37, 1985. Zola 1. K. Structural constraints in the doctor-patient relationship: the case of non-compliance. In The Relevance of Social Science for M edicine (Edited by Eisenberg L. and Kleinman A.), pp. 241-252. D. Reidel, Dordrecht, 198 1. Ley P. and Spelman M. S. Communicaling with rhe Patient. Staples Press, London, 1967. Garrity T. F. Medical compliance and the clinician-patient relationship: a review. Sot. Sci. M ed. 15E, 215-222, 1981. DiMatteo M. R. and DiNicola D. D. Achiming Pafienl Compliance. Pergamon Press, New York, 1982. Bartlett E. E., Grayson M., Barker R., Levine D. M., Golden A. and Libber S. The effects of physician communications skills on patient satisfaction, recall and adherence. J. Chron. Dis. 37, (9), 755- 164, 1984. Hughes D. Control in the medical consultation. Sociology 16, (3) 359-316, 1982. Svarstad B. The doctor-patient encounter: an observational study of communication and outcome. Unpublished dissertation, University of Wisconsin, 1974. Geersten H. R.. Grav R. M. and Ward J. R. Patient non-compliance’ within the context of seeking medical care for arthritis. J. Chron. Dir. 26, 689- 698, 1973. Freidson E. Profissional Dominance. Dodd Mead and Co., New York, 1970. Mishler E. G. The Dominance o/M edicine. Abler, New Jersey, 1984. Waitzkin H. A critical theory of medical discourse. J. Hllh Sot. Behac. 30, 22G239, 1989. control and 14. Waitzkin H. and Stoeckle J. D. Information the micro-politics of health care. Sot. Sci. M ed. 10, 263- 276, 1976. 15. Todd J. Cruel absurdity of telling your patients everything. Docmr 15, (12). 37, 1983. 16. Mechanic D. M edical Sociology _.. Free Press, New York, 1978. of 17. Rost K., Carter W. and Inui T. S. Introduction information during the initial medical visit. Sot. Sci. M ed. 28, 315- 321, 1989. I. and Lapinska E. Doctor-patient 18. Heszen-Klemens interaction, patients’ health behaviour and effects of treatment. Sot. Sci. ,bfed. 19, 9- 18, 1984. U. Qualitative research on chronic illness; 19. Gerhardt the issue and the story. Sot. Sci. M ed. 30, 1149-l 159, 1990. inter20. Stewart M. What is a successful doctor-patient view? Sot. Sri. M ed. 19, 167- 175, 1984. Patient non -compliance 513 32. Deyo R. A., Inui T. S. and Sullivan B. Non-compliance 2 I. Byrne P. S. and Long B. E. zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA Doctors Talking to Patients. with arthritis drugs. /. Rheum. 8, (6) 931-936,.1981. HMSO, Exeter, 1976. 33. Frankel R. M. The lavina on of hands. In The Social 22. Henbest R. J. and Stewart M. Patient-centredness in the Organisarion of Doctor-P% ienr Communication (Edited consultation: dies it really make a difference? Fam. by Fisher S. and Todd A. D.), pp. 19-54. Center For Pracr. 7, (I), 28-33, 1990. Applied Linguistics, Washington, 1983. 23. Squier R. W. A model of empathic understanding and 34. Cornwell J. Hard Earned Lives. Tavistock, London, adherence to treatment regimens in practitioner-patient 1984. relationships. Sot. Sci. M ed. 30, 325- 339, 1990. 35. Donovan J. L. “W e Don’t Buy Sickness, II Just 24. Caughey D. E. (Ed.) Patient communication and effecComes”: Health and Illness in the Lives of Black People tive rheumatology. J. Rheum. 16, 5-6, 1989. in London. Gower, Aldershot, 1986. 25. Waitzkin H. The micro-politics of medicine. In!. J. Hlth 36. Evles J. D. and Donovan J. L. The Social Edicts of Serv. 14, (3), 397- 412, 1984. health Policy . Avebury, Aldershot, 1990. __ 26. Kaplan S., Greenfield S. and Ware J. Assessing the 37. Moll J. M. H. Medical communication: recent aspects effects of physician-patient interactions on outcomes of and the relevance to rheumatology. In Recent Advances chronic disease. M ed. Care 27, (3) Suppl, SllO-S127, in Rheumatology 4 (Edited by Mall J. M. H. and 1989. Sturrock R. D.), pp. 73-102. Churchill Livingstone. 27. Silvers I. J., Hovel1 M. F., Weisman M. H. and Mueller Bath, 1986. M. R. Assessing physician-patient perceptions in rheu38. Burgess M. M. Ethical and economic aspects of nonmatoid arthritis. Arth. Rheum. 28, (3), 3&305, 1985. compliance and overtreatment. CM AJ 141, 777- 779, 28. Wright A. L. and Morgan W. J. Gn the creation of 1989. problem patients. Sot. Sci. M ed. 30, 951-959, 1990. 39. Donovan J. L., Blake D. R. and Fleming W. G. The 29. Charmaz K. Discovering chronic illness: using patient is not a blank sheet: lay beliefs and their grounded theory. Sot. Sci. M ed. 30, 1161-1172, 1990. relevance to patient education. B. J. Rheum. 28, 5841, 30. Glaser B. G. and Strauss A. L. The Discovery of 1989. Grounded Theory. Aldine, Chicago, 1967. . _ 40. Donovan J. L. Patient education and the consultation: 31. Schutz A. The Phenomenology of the Social W orld. the importance of lay beliefs. Ann. Rheum. Dis. In Press. Heinemann, London, 1972.