Non-Cardiac Chest Pain and Benign Palpitations in the Cardiac Clinic Br Heart J: First Published As 10.1136/Hrt.72.6.548 on 1 December 1994

548 Br HeartJ 1994;72:548-553

Non-cardiac chest pain and benign palpitations in the cardiac clinic Br Heart J: first published as 10.1136/hrt.72.6.548 on 1 December 1994. Downloaded from Richard Mayou, Bridget Bryant, Colin Forfar, David Clark

Abstract Reports of chest pain and palpitations are the Objectives-To assess the characteristics most common reasons for referral from a gen- of consecutive patients referred from eral practice to cardiac clinics. Patients not general practice with the presenting dis- given a cardiac diagnosis are known by clini- order ofchest pain or palpitations, and to cians to be difficult to manage. In particular, determine the outcome at six months and there is consistent evidence that patients with three years. chest pain who are found to have normal Setting-A single consultant teaching coronary angiograms have a poor outcome hospital cardiac clinic receiving new and continue to report persistent symptoms, referrals from a health district. disability, and continuing concern about heart Design-94 consecutive referrals by gen- disease.' Less is known about the much larger eral practitioners to a cardiac clinic with number of patients with non-cardiac chest the presenting disorder of chest pain or pain who are seen in cardiac clinics but do not palpitations were assessed at first atten- undergo angiography,2" and especially about dance (research interview, cardiologists' patients who are reassured that their palpita- ratings, systematic medical case note tions are not medically significant.4 There is information), home interview six months therefore a need for further research directed later, and by a postal questionnaire at to improving the differential diagnosis of non- three years. cardiac disorders, the recognition of those Outcome measures-Physical and who might benefit from extra treatment, and psychological symptoms, limitation of to the development and evaluation of treat- activities, satisfaction with care, and use ments suitable for use with large numbers of ofhealth care resources. patients in routine clinical settings. Results-39 patients were given a cardiac This paper describes the assessment and

diagnosis and 51 patients were not given management of a cohort of new patients http://heart.bmj.com/ a cardiac or other major physical diag- referred to a district hospital cardiac clinic nosis. The non-cardiac group was more and reports their symptomatic and functional likely to be young women, and to report outcome at six months and three years. other physical symptoms and previous Findings for those with and without cardiac psychiatric problems. The cardiac and disorders are compared, and issues relating to non-cardiac groups reported progressive patients thought not to have heart disease are improvement in presenting symptoms discussed: and disability at the six months and three (a) What is the size of the clinical problem? on October 2, 2021 by guest. Protected copyright. year follow up, but little change in mental What proportion of the patients is reassured state. Even so, three quarters of the non- by clinic assessment that their chest pain or cardiac subjects described continuing palpitations are not medically serious? How limitation of activities, concern about the many continue to suffer physical symptoms, cause oftheir symptoms, and dissatisfac- disability, and distress? tion with medical care. (b) What are the clinical or other characteristics University Conclusions-A substantial proportion of of non-cardiac patients compared with those Department of the consecutive referrals continued to diagnosed as having definite or probable heart Psychiatry, describe symptoms and disability disease? Warneford Hospital, Can we either at initial clinic Oxford throughout the three years after clinic (c) identify R Mayou attendance. Outcome was poor for those attendance, or at follow up, a subgroup of B Bryant who had negative investigations and were patients with poor prognosis who might bene- D Clark reassured that they had no cardiac dis- fit from extra intervention? Department of order or other serious physical finding. (d) What are the clinical implications of the Cardiovascular Medicine, John These results have implications for defin- provision of improved, but cost-effective, care Radcliffe Hospital, ing the role of psychological assessment for large numbers of patients? Oxford and for the formulation of cost effective C Forfar clinical measures to (a) minimise dis- Correspondence to: Dr Richard Mayou, ability associated with cardiac disorder; Methods University Department of and (b) prevent and treat handicaps in Consecutive patients referred by general prac- Psychiatry, Wameford Hospital, Oxford OX3 7JX. those without major physical diagnoses. titioners to a consultant cardiologist with pre- Accepted for publication senting disorders of chest pain or palpitations 1 June 1994 (Br Heart J 1994;72:548-553) were identified at their outpatient attendance. Non-cardiac chest pain and benign palpitations 549

Table I Characteristics ofpatients with cardiac diagnoses (definite or probable) and those with no cardiac or other major physical diagnosis. Values are No (%) unless stated othenvise Non-cardiac patients (n = 51) Cardiac patients (n = 39)

Mean (range) age (years) 43-5 (21-65) 53-4 (39-69) Br Heart J: first published as 10.1136/hrt.72.6.548 on 1 December 1994. Downloaded from Male 22 (43) 23 (59) Previous hospital care for current episode Outpatient 10 (20) 2 (5) Inpatient 2 (4) 4 (10) Emergency 3 (6) 3 (8) General practice care Cardiac drugs 20 (39) 22 (56) Psychotropic drugs 4 (8) - (-) Diagnosis of heart disease 6 (12) 11 (28) Time before referral (months) <6 24 (47) 17 (44) 6-12 8 (16) 7 (20) >12 19 (37) 15 (40) Symptoms Mean (SD) Beck depression inventory 4-96 (4 33) 4.59 (4 08) Mean (SD) Spielberger trait anxiety 3900 (11-16) 37-10 (10-74) Mean (SD) hyperventilation score 12-24 (10-56) 9-72 (7-26) DSMIIIR psychiatric diagnosis 22 (42) 9 (24) Cardiologists' assessment and notes Mental state (moderate/severe) 20 (39) 7 (18) Hyperventilation (possible/probable) 13 (26) 4 (10) Psychiatric contribution to symptoms (probable/definite) 17 (33) 5 (13) Current psychiatric problem 22 (42) 6 (16) Social problem 11 (21) 4 (11) Investigations Exercise test 25 (49) 26 (67) 24 hour tape 8 (16) 5 (13) Angiography 1 (2) 10 (25)

Subjects were interviewed using a brief semi- Fifty one (54%) were given no physical diag- structured interview derived from previous nosis (non-cardiac group). Most of those with work with angina5 and non-cardiac chest pain6 cardiac diagnoses, but few of the non-cardiac and completed three standard self report patients, were offered continuing clinic follow questionnaires: the Beck depression inven- up. tory,7 the Spielberger trait anxiety,8 and a hyperventilation check list.9 The cardiac CARDIAC GROUP assessment by a single consultant (CF) fol- The cardiologist diagnosed heart disease in 39 lowed normal clinical practice but, in addi- patients, 31 (79%) with angina (24 definite tion, the cardiologist completed a single page and seven probable) and eight (21%) with standard assessment sheet at the end of the paroxysmal arrhythmias (six definite and two consultation covering the physical diagnosis, probable). Table 1 gives their characteristics. aetiology ofthe present symptoms, hyperventi- Physical symptoms were usually marked (fig- lation, and role of psychological factors. ure): 17 (43%) rated their symptoms as sub- http://heart.bmj.com/ Further information was abstracted from jectively "severe", 11 (28%) as "very medical case notes. RM subsequently made distressing", and 8 (21%) as "very limiting". Severity stan- 100 psychiatric diagnoses according to the The means of the self report anxiety and 90 U diagnostic and statistical man- 80 Non cardiac dard American depression scores were within the upper nor- 70 IiCardi cdaual IIIR diagnostic criteria'° for psychiatric mal range for the general population,78 and 50 disorders. we made a research diagnosis of psychiatric 40 at disorder major depression and five anxi- 30 Six months later subjects were reassessed (four on October 2, 2021 by guest. Protected copyright. 20 10 home. The self report questionnaires were ety disorders) in a quarter of patients. 0 repeated and a semistructured interview cov- Reactions to the diagnosis of heart disease 100 Distress ered symptoms, beliefs, consultation, and dis- varied. Five patients found the definite diag- 90 80 ability. At three years subjects were sent nosis of heart disease a relief after months of 70 m postal questionnaires covering the same areas uncertainty. Ten patients described consider- o 60 .1 50 as assessments, with the able, but transient, distress. Two, both of 40 the previous together 30 Beck and Spielberger scales. Non-responders whom had long histories of chronic psycho- 20 10 were contacted by telephone. logical problems, reported continuing severe Statistical analysis was by the x2 test and by worry. Limitation generalised linear interactive modelling. 90 80 Outcome 70 Patients with cardiac disorders continued to Results use cardiological and general practice Most patients had been managed in general resources throughout the three year follow up 230 practice for several months before referral period. Outcome was generally encouraging, 0 6/12 3 years (table 1) and were eventually referred to but a small minority had disproportionate dis- Proportions ofsubjects establish a definite diagnosis for persistent ability and appeared to pose management reporting symptoms as symptoms. Clinical diagnosis was usually problems for other than strictly physical being subjectively severe, established within two clinic visits. Thirty reasons. distressing, and limiting at one had been clinic attendance, three nine of the 94 (41%) consecutive referrals At six months patient rediag- months, and three yearsfor were given either a definite or a probable car- nosed as "non-cardiac" having had a normal cardiac (n = 39, 29, 31 diac diagnosis (cardiac group). Four further coronary angiogram. Twenty nine patients non- respectively), and were other (74%) were still attending the clinic and 31 cardiac (n = 51, 46, 40 patients given specific physical respectively) subjects. diagnoses (two oesophagitis, two pneumonia). (79%) were receiving regular cardiac drugs. 550 Mayou, Bryant, Forfar, Clark

Table 2 Outcome at six months for cardiac and non-cardiac patients. V4alues are No The twenty nine subjects who were reinter- (%) unless stated otherwise. Percentages arefor all those interviewed orforr total number of viewed reported that their physical symptoms thosefor whom applicable had improved (figure), although effects on Non-cardiac patients Cardiac patients everyday activities and social life remained = (n 46) = Br Heart J: first published as 10.1136/hrt.72.6.548 on 1 December 1994. Downloaded from (n 29) marked (table 2). Most patients were satisfied Symptoms Improved (better/much better) 27 (59) 14 (48) with their medical care, but five reported side Symptoms in last month 35 (76) 23 (79) effects of/I blockers. There was little change Effects on social activities (moderate/great) Walking two miles 11/44 (26) 17OR (Al) in anxiety, depression, and hyperventilation Household tasks 5/46 (11) 11/29 (38) scores. Social life 4/40 (9) 7/29 (24) Sex 3/33 (9) 8/25 (32) At three years there had been no deaths Driving 7/36 (19) 6/24 (25) and further overall clinical improvement Family life 8/41 (20) 4/26 (15) Work 9/37 (24) 8/19 (42) (table 3). Four patients had had angiography Mean (SD) Beck depression inventory 4-74 (4-61) 5 31 (4-41) since the six month assessment, and six had Mean (SD) Spielberger trait anxiety 38-13 (11-62) 37-59 (10-39) Mean hyperventilation 11-0 (9 72) 9-97 (8-21) had an operation; one patient was awaiting Medical care coronary artery surgery. Seventy per cent were Dissatisfaction General practitioner 25 (54) 7 (24) taking regular drugs for angina. Hospital 16 (35) 6 (21) A fifth of the thirty one patients (78%) who Reassured by clinic Not reassured 18 (39) 12 (41) completed follow up questionnaires were Partially reassured 21 (46) 13 (45) symptom free and two fifths were not limited. Very reassured 7 (15) 4 (14) Although most were satisfied with their Attending clinic 5/51 (10) 29/39 (74) progress and medical care, a fifth (six patients) described considerable distress and frustration at the effects of their symptoms on Table 3 Outcome at three years for cardiac and non-cardiac patients. Values are No their everyday lives (figure). (%) unless stated otherwise. Percentages are for all those interviewed orfor total number of thosefor whom applicable NON-CARDIAC GROUP Non-cardiac patients Cardiac patients Fifty one patients were given no major physical (n = 40) (n = 31) diagnosis. Although all were new referrals to Symptoms last three months Chestpain 14 (35) 18 (58) the clinic, many had already received consid- Palpitations 20 (50) 15 (48) erable medical care for their symptoms (table Breathlessness 17 (43) 16 (52) Mean (SD) Beck depression 5-33 (4 99) 5-06 (4 84) 1). Thirteen (25%) had previously attended Mean (SD) Spielberger anxiety 39 97 (13-70) 38-10 (16-54) hospitals during the current episode, six Medical care Further angiography (last two years) 0 (0) 4 (13) (12%) had been told by their general practi- Would have liked more tests 10 (25) 5 (16) tioners that they had heart disease, and 20 Further cardiac treatment (last two years) 8 (20) 22 (71) Currently attending oupatient clinic 0 (0) 9 (29) (40%) were being treated with cardiac drugs. Attended general practice in last three months 26 (65) 22 (71) Five patients had had previous prolonged Cardiac surgery of chest or Angioplasty 3 periods pain palpitations, during Coronary artery bypass grafting 1 which four had definitely been told they did ASD repair 1 not have heart disease. http://heart.bmj.com/ Pacemaker 1 We rated the presenting symptoms as less ASD, atrial septal defect. limiting than those of the cardiac group (figure), but as no less subjectively severe or distressing. Overall the psychological and social Three patients had posed major non-medical characteristics were similar to those of the car- management difficulties; one had refused diac group, but the non-cardiac patients were further appointments, one was angry at being more likely to be women (p < 0 05), young told that an operation was not required, and (p < 0 05), and have no family history of heart on October 2, 2021 by guest. Protected copyright. the third had multiple physical, psychological, disease (p < 0 05). They were more likely to and social difficulties. report multiple physical disorders and to be aware of heart disease in either relatives or other people. Although mean anxiety and Table 4 Non-cardiac group: comparison of those presenting primarily with chest pain depression scores were similar to those of the andpalpitations. Values are No(%) unless stated otherwise cardiac group, general practitioners and cardi- < Pain (n = 33) Palpitations (n = 18) ologists recognised significantly (p 0 05) more psychological problems among the non- Mean (range) age (years) 49-6 (22-69) 41-9 (21-65) Male 20 (61) 2 (11) cardiac patients. Twenty one patients (42%) were given research diagnoses of psychiatric Previous history Symptom duration (months) disorders (anxiety and depression). 0-6 15 (46) 9 (50) 7-12 6 (18) 2 (11) >12 12 (36) 7 (39) Chest pain and palpitations Symptom characteristics These two subgroups were similar (table 4). Severity (severe/very severe) 7 (21) 10 (55) were Distress (moderate/severe) 19 (58) 13 (72) The patients with "benign" palpitations Limitation (moderate/severe) 13 (39) 5 (28) more likely than those with non-cardiac pain Cardiologist's assessment < < 0 Hyperventilation (possible/definite) 7 (21) 6 (33) to be women (p 0-05), younger (p 05), Mental state (moderate/severe) 12 (36) 8 (44) and to report previous (p < 0 05) psychiatric Psychiatric contribution (probable/definite) 10 (30) 9 (50) Psychological assessment problems. Previous consultation for "nerves" 5 (15) 10 (56) The cardiologist rated 20 (60)% of those Psychological factors (general practitioner) 12 (36) 10 (55) with the disorder of chest as Mean (SD) Beck depression 4-79 (4-11) 4-89 (4 48) primary pain Mean (SD) Spielberger anxiety 37-96 (10-88) 39-67 (11-12) having minor physical explanations of their DSMIIIR diagnosis 13 (38) 9 (50) symptom: chest wall problems (1 1 patients); Non-cardiac chest pain and benign palpitations 551

oesophageal reflux or spasm (three patients); Discussion cervical spine disorders (one patient); infec- We report a prospective and consecutive tions (two patients). They also often rated series of referrals from general practitioners to hyperventilation (26%) and psychological a single cardiologist providing a mainly dis- factors (33%) as probable contributing trict service. The findings have implications Br Heart J: first published as 10.1136/hrt.72.6.548 on 1 December 1994. Downloaded from factors. for the systematic development and evalua- All 18 patients with the primary disorder of tion of improved services. Our study has the palpitations were rated by the cardiologist as advantage of reporting outcome for patients being abnormally aware of cardiac rate or receiving usual care in a routine cardiac clinic, rhythm, six of sinus tachycardia, and 12 of the only change from normal clinical proce- ventricular ectopics. They rated psychological dure being the one page research assessment factors as important in seven (39%). Thirteen form on which the cardiologist recorded clini- (33%) of those with chest pain were given cal diagnosis, possible aetiological factors, and research psychiatric diagnoses (nine major the role of hyperventilation and psychological depression and four anxiety disorders). Nine factors. The main disadvantages are the sub- (50%) of those presenting with palpitations jective nature of the cardiological assessment were given research psychiatric diagnoses (and of the "reassurance" given) and the lack (eight with panic disorder and one major of precise diagnostic criteria or systematic depression). investigation, and uncertainty about whether the findings are representative. We believe our Outcome findings are consistent with published data, Outcome over the three year follow up was typical of the hospital's four consultant clin- disappointing, with many patients reporting ics, and also accord with wider clinical experi- continuing symptoms, emotional stress, limi- ence, but they will need to be repeated and tations of everyday activities, and concern extended in further systematic multicentre about heart disease (tables 2 and 3). At 6 research. months only five (10%) were still attending the clinic; four were still receiving cardiac CLINICAL IMPLICATIONS FOR THOSE WITH drugs. The 46 (90%) patients who were re- CARDIAC DIAGNOSES interviewed were no more likely than the The clinical outcome of these patients was cardiac patients to report improvement in encouraging. Patients felt reassured, even their presenting symptoms (figure) or in their relieved, to have been given a definite diagno- mood or hyperventilation scores, and the sis and authoritative advice and treatment. effects on everyday life remained considerable Two potentially avoidable problems were (table 2). Many patients were markedly dis- apparent (a) lengthy waiting lists for angio- satisfied with medical care in general practice plasty and on operation caused distress and (n = 25, 54%) and hospital (n = 16, 35%), resulted in serious work and other practical particularly reporting that they had not been difficulties; (b) a minority either did not com-

given adequate explanations for their continu- ply with medical advice or described "dispro- http://heart.bmj.com/ ing symptoms. They remained worried that portionate" limitations of their everyday lives. the symptoms were due to cardiac or other Such difficulties might be prevented or serious physical causes. There were few differ- reduced by early recognition and the provi- ences between those who had presented with sion of extra help (information, advice, exer- chest pain and those with palpitations, cise training, or psychological intervention) in although the former reported more social the cardiac clinic, rehabilitation unit, or in disability. primary care. At three years there had been no deaths on October 2, 2021 by guest. Protected copyright. and a further improvement in symptoms and SIZE OF THE PROBLEM OF NON-CARDIAC functional capacity (table 3), but there was a DISORDERS considerable continuing use of hospital and More than half of the new referrals presenting general practice resources. Most of the 40 to a cardiac clinic with either chest pain or patients (80%) who completed questionnaires palpitations did not receive a major physical reported persistent symptoms, although there diagnosis. Despite reassurance, most of these had usually been an improvement in the patients described continuing symptoms, dis- severity of emotional distress and in everyday ability and distress six months and three years activities. Individual patterns of outcome were later. These findings for consecutive attenders little changed from six months. All those who are consistent with other evidence for more had a generally unsatisfactory outcome at selected patients with non-cardiac chest pain' I' three years had reported similar problems in and provide evidence of a similarly poor out- the earlier follow up. Four patients remained come for those with palpitations. The continu- convinced that heart disease was the cause of ing management of the patients poses a their continuing symptoms and a further eight considerable problem in primary care. believed it might be a cause. Other causes mentioned were: stress (18 probable, nine CHARACTERISTICS OF PATIENTS WITH possible), indigestion (five), muscle pain NON-CARDIAC PAIN OR BENIGN PALPITATIONS (two), and infections (two). Ten patients said We have previously argued that functional they did not know what had been the cause of somatic symptoms, such as persistent non- their symptoms and several found continuing cardiac chest pain and benign palpitations, are uncertainty unsatisfactory and worrying. of multicausal aetiology and that patients 5-52 Mayou, Bryant, Forfar, Clark

misinterpret minor physical pathology or the are likely to be crucial. Those who continue to physiological consequences of autonomic experience symptoms say that they want an arousal as evidence of heart disease.' 11 explanation which reassures them that they do Misinterpretation is most likely in those who not have a serious medical problem, and are predisposed by a hypochondriacal disposi- advice about how to return to a full life. We Br Heart J: first published as 10.1136/hrt.72.6.548 on 1 December 1994. Downloaded from tion, are worried about life stresses, have per- believe this can best be provided by: (a) sim- sonal knowledge of heart disease in their own ple measures feasible within the cardiac clinic family or in other people, or have a psychiatric together with reinforcement by the general disorder. Patients fears are likely to be exacer- practitioner; and (b) the early recognition and bated and maintained by the delay in estab- selective more intensive treatment of persis- lishing a diagnosis. By the time they are given a tent problems. specialist opinion, a patient may have been given a provisional diagnosis of heart disease Routine care in general practice and been prescribed car- Resources and time in cardiac clinics are lim- diac drugs during a long wait for a specialist ited, especially at a time of increasing refer- appointment and any subsequent investiga- rals. Even so, small changes in clinical tion. Such delay and uncertainty inevitably practice and the use of self help materials increases anxiety (and its associated physical would be possible and valuable (handouts, symptoms) and reinforces concern about cassettes, audiotapes). Information should heart disease. include: (a) reassurance that the presenting It is apparent that the cardiologists implic- symptoms are a common and familiar clinical itly used this model as minor physical factors, problem, that specialist assessment is appro- excessive awareness of cardiac rhythm, hyper- priate, and that outcome is excellent; (b) an ventilation, and psychological factors were explanation of the alternative causes; and (c) rated as aetiological factors in most patients. advice about how to cope with any further The application of this diagnostic formulation symptoms and how gradually to resume nor- was clearly more straightforward in those with mal full activities. palpitations, where the association between There may be a role for appropriately anxiety and excessive awareness of rhythm is trained specialist nurses who would work with often clearer, than for chest pain where the supervision from cardiologists and a behav- role of physical factors over a period of ioural psychologist. We are therefore currently months or even years is more complex and evaluating an improved single intervention for clinically more difficult to establish. patients with chest pain and negative coronary Despite the different aetiology, there were angiograms which is administered by a spe- marked similarities between patients diag- cially trained ward nurse. Our initial impres- nosed as having heart disease and the non- sion is that standardised and systematic cardiac group. In particular, there was no discussion, handouts, cassettes, and follow up systematic difference in mental state. This telephone calls are popular with patients and

means it will not be possible to devise a simple their families, and can be integrated into ward http://heart.bmj.com/ psychological assessment which could identify nursing routine. patients with non-cardiac disorders. Greater concentration on the psychiatric history and Selective extra help psychological symptoms might enable more A small proportion of patients may benefit positive diagnoses of psychiatric disorders, from specific treatments, such as referral to a such as depressive illness and panic disorder. gastroenterologist or antidepressant drugs. More often treatment will need to be directed CAN NON-CARDIAC PATIENTS AT RISK OF POOR to correcting patients' misunderstanding and on October 2, 2021 by guest. Protected copyright. OUTCOME BE PREDICTED? misinterpretations of their symptoms and to Our analyses show that it is difficult to predict promoting management of anxiety symptoms symptomatic outcome at clinic attendance, and a graded return to normal activities. apart from a small proportion of patients with We have previously shown that a systematic long histories of multiple unexplained somatic psychological treatment can be effective with symptoms, or who have not been reassured by patients referred from general practice with previous negative investigations. We therefore persistent non-cardiac chest pain,6 and our believe that it is clinically more sensible and current research has replicated these findings cost effective to concentrate on the early in a cardiac clinic. Such intensive treatment is detection of persistent problems during follow only required by the most disabled patients, up after clinic assessment. Patients who and briefer and simpler help could usually be described difficulties at three months almost provided by a general practitioner or specialist invariably described similar difficulties at nurse. three years, whereas those who were improved This prospective study has shown the con- at three months continued to have an excel- siderable size of the clinical problem in hospi- lent outcome. tal and primary care of patients with chest pain and palpitation not due to heart disease. It WHAT ARE THE CLINICAL IMPLICATIONS FOR is likely that simple improvements in care NON-CARDIAC DISORDERS? would be effective in most patients, but a pro- Negative clinical investigation and authorita- portion will require specialist help. Although tive reassurance by a cardiologist is only effec- it would be inappropriate to argue that cardi- tive for a minority of patients. The content of ologists should be responsible for longer term this reassurance and the way it is presented management, they have a central role in Non-cardiac chest pain and benign palpitations 553

establishing the diagnosis and initiating the 2 Mayou RA. Chest pain in the cardiac clinic. Psychosom Res 1973;17:353-7. treatment. There is a need for collaboration 3 Channer KS, James MA, Papouchado M, Russell Rees J. with primary care providers to ensure that Failure of a negative exercise test to reassure patients with chest pain. Q Med 1987;240:315-22. treatment is continued. We believe that fur- 4 Barsky AJ. Palpitations, cardiac awareness, and panic dis- Br Heart J: first published as 10.1136/hrt.72.6.548 on 1 December 1994. Downloaded from ther research should be directed towards eval- order. Am Med 1992;92:31-4. 5 Mayou RA, Bryant B. Quality of life after coronary artery uating cost effective interventions which could surgery. QJ7Med 1991;239:239-48. be delivered in general practice or by appro- 6 Klimes I, Mayou RA, Pearce MJ, Coles L, Fagg JR. Psychological treatment for atypical non-cardiac chest priately trained experienced nurses. It will be pain: a controlled evaluation. Psychol Med 1990;20: essential to establish training programmes and 605-11. 7 Beck AT. Depression inventory. Philadelphia: Center for ensure regular skilled supervision. Improved Cognitive Therapy, 1978. care is likely to lead to a better outcome and 8 Spielberger CD, Gorsuch R, Lushene R. State-trait anxiety inventory manual. Palo Alto: Consulting Psychologist to a reduction in the use of primary care and Press, 1970. hospital resources. 9 Clark DM, Hemsley DR. The effects of hyperventilation: individual variability and its relation to personality. This study was supported by a grant from the Oxford Behav TherExp Psychiatry 1982;13:41-7. University Research and Equipment 10 American Psychiatry Association. Diagnostic and statistical Committee. manual of mental disorder. 3rd ed. Washington DC: 1 Chambers JB, Bass C. Chest pain and normal coronary American Psychiatric Association, 1987. anatomy: review of natural history and possible 11 Mayou RA. Invited review: atypical chest pain. aetiologic factors. Prog Cardiovasc Dis 1990;33:161-84. Psychosom Res 1989;33:373-406. http://heart.bmj.com/ on October 2, 2021 by guest. Protected copyright.