Chest pain is one of the commonest reasons for consulting health professionals,1 with up to a quarter of the population presenting with symptoms over a lifetime.2 Initial care is principally directed at swift identification, and management, of potentially life-threatening cardiac conditions.3 The National Service Framework for coronary heart disease advocated the establishment of rapid access clinics for chest pain.4 Dedicated rapid access chest pain clinics have therefore been established in acute settings both to reduce waiting times for assessment, and to enhance risk stratification.5 However, while cardiac and non-cardiac pain can be differentiated more rapidly in these clinics, an understandable focus on potential cardiac pathology has neglected the majority of patients attending. Up to 70% of patients present with pain of non-cardiac origin6 and often receive neither definitive diagnosis nor conclusive information, leaving scope for potential development of further health concerns.

The management of patients whose chest pain remains unexplained is challenging. Many studies have noted that prevalence of anxiety states, such as panic disorder, is higher for such populations,7^–9 and that, in comparison with patients with cardiac conditions, those with non-cardiac pain are more sensitive to body sensation and prone to health anxiety.10 Even for those whose distress does not reach that of a psychiatric patient, a substantial majority will continue to experience distress and disability11 12 despite no increased risk of cardiac-related mortality. Over one-third of patients with non-cardiac chest pain report their physical and emotional health as only fair or poor and they frequently oscillate between primary and secondary care, undergoing costly investigations and consuming medicine inappropriate to their needs.13

Rapid access chest pain clinics do not directly attend to psychological distress, but guidelines direct staff to provide prompt reassurance to patients whose symptoms are non-cardiac and to minimise adverse effects of diagnosis and treatment.3 This usually comprises feedback on test outcomes and/or referral for further investigations, and appears founded on the assumption that reassurance offered by early clinic attendance, and a non-cardiac diagnosis, will alleviate distress.

However, there has been little consideration and examination of the impact of rapid access chest pain clinic attendance on psychological morbidity or beliefs about health, particularly over the longer term. Given the potency of patients’ beliefs and attributions in predicting and explaining health-related behaviour in cardiac populations,14 and their commitment to rehabilitation,15 it seems essential to understand them from the perspective of the chest pain clinic. Thus this study aimed at examining the psychological morbidity of patients referred to a rapid access chest pain clinic. We also sought to examine the potential differences in psychological morbidity and illness beliefs, between those with cardiac and unexplained chest pain over time, and to explore the impact of clinic attendance on distress and appraisal of symptoms.

METHODS

Design

We undertook a prospective, questionnaire survey of new attenders to the rapid access chest pain clinic at a tertiary referral centre in Leicester, UK. All patients, referred to the clinics between January and March 2004 were invited to participate in the study. Patients referred to these clinics were seen within either 48 hours or 2 weeks, dependent upon the general practitioner’s assessment of urgency. Participants completed self-administered questionnaires on three occasions: after an initial appointment had been offered but before clinic attendance, 1 week after clinic attendance, and 2 months after initial clinic attendance.

Sample

The teaching hospital in which this clinic is based sees about 2600 patients annually. Four hundred and thirty-one patients were referred to the clinic during the sampling period, 371 of whom were eligible and approached to participate by a postal invitation. Letters of invitation were issued from the eligible patient’s consultant cardiologist accompanied by a patient information sheet and initial questionnaire. Those patients who wished to participate were asked to complete the questionnaire before the initial appointment and bring it with them on the day of clinic attendance, where they met a researcher (NJ), could discuss concerns or queries and were asked to complete a consent form. Eligibility was dictated by admission criteria for clinic attendance (that presentation of chest pain was new and suggestive of angina; that the pain was a new presentation for patients with known ischaemic heart disease; that symptom duration was <6 months; and that male patients were aged ⩾35, female patients ⩾40).

Main outcome measurements

A binary and ordinal response questionnaire was developed covering psychological morbidity and illness appraisal, and was complemented by demographic data, details of medical history, and ultimate clinic diagnosis gleaned from routine clinic records.

Psychological measures

Psychological distress

Anxiety and depression were assessed using the Hospital Anxiety and Depression Scale (HADS), which has robust psychometric properties,16 17 and is brief and easy to administer. The measure comprises 14 items, each rated on a four-point Likert scale (0–3). Scores from each of the seven-item anxiety and depression subscales are summed to a total ranging from 0 to 21. Higher scores indicate greater levels of anxiety and depression. Optimal balance between sensitivity and specificity for populations with atypical chest pain is suggested to be a cut-off point of ⩾8,18 and we therefore applied this criterion.

Health anxiety was assessed using the Health Anxiety Inventory (HAI)—short version,19 a self-rated measure with good reliability and validity, and able to distinguish normal levels of health concern from frank hypochondriasis. Respondents are presented with 18 items (14 dealing with general health anxiety and a further four focusing on consequences of serious illness), each comprising four statements from which the most applicable is selected. Statements are rated on a four-point Likert scale (0–3) and responses are summed to provide scores for health anxiety and negative consequences. Higher scores reflect greater anxiety.

Illness perceptions

The Illness Perceptions Questionnaire—revised (IPQ-R),20 which has shown robust psychometric properties,21 assessed participants’ cognitive and emotional representations of their chest pain. The questionnaire is derived from the Common-Sense Self-Regulation model (CS-SRM),22 which suggests that when patients are confronted by a threat to their health, they construct mental representations of their condition and these intrinsic processes dictate both behaviours and emotional reactions. The representations comprise five factors: identity, referring to the symptoms experienced (chest pain) and label attached to a condition (such as heart attack); beliefs about the consequences of a change in health (such as physical, social and financial sequelae); beliefs about cause (such as worry or smoking); timeline (the duration of ill health); and perceived scope for cure or control of the condition.

Participants affirmed whether or not they had experienced listed symptoms, to produce an identity score. Representations of timeline (acute/chronic and cyclical), emotional dimensions, consequences, control (personal and treatment), and illness coherence (understanding of the pain) were assessed from responses to 50 statements on a five-point Likert scale (1–5). Total scores for each subscale were obtained by summing item scores and dividing by number of items, with lower scores indicating better adjustment (such as fewer consequences, pain viewed as more treatable) for all except consequences and illness coherence (reversed scores). Causal beliefs were explored separately with the same Likert scale, but were not summed as each item reflected a distinct causal belief.

Statistical analysis

Data were analysed using SPSS for Windows 11.5 statistical software package. Items of the psychological measures (HADS, HAI) and multiscore items of the IPQ-R, were treated as continuous data and interrogated parametrically. Changes over time for these data were assessed using a two-way analysis of variance, with time as a within-subjects and diagnosis as a between-subjects factor, and age as a covariate. All significant interactions between time and these measures were further explored using independent t tests. Frequency data were analysed using χ2 test.

RESULTS

Of 371 eligible patients, 246 (66%) completed the questionnaires at the first time point. One hundred and seventy-six (47%) participants responded at time point 2 and 140 (38%) at time point 3. Not all respondents completed all questions at the second time point. Median waiting time for first attendance at clinic was 12 days. Table 1 summarises patient baseline characteristics, showing few differences between the groups except that those whose pain was diagnosed as cardiac were older and more likely to be retired. Comparison with local demographic data showed that our sample comprised a greater proportion of patients who defined themselves as white British, χ2 = 8.20, p<0.02. Potential bias from sample attrition was dealt with by comparing demographic, HADS and IPQ-R data for those who remained in the study with those who dropped out at the second and third time point. Significant demographic and clinical differences were few—at time point 2 those who remained in the study appeared older than those dropping out (t = 2.25, p<0.05) with no differences evident at time point 3.

HADS data showed that the extent of clinically notable anxiety across study participants was higher than community norms23 before clinic attendance (table 2) and did not significantly differ between the two diagnostic groups. This lack of difference between the groups at the study outset is confirmed by responses to the Health Anxiety Inventory, where no significant difference was found in total scores (table 3), nor was the current sample more health anxious than outpatient attenders comprising the norm data set on which the measure is validated.

After clinic attendance, levels of anxiety for those with non-cardiac pain significantly exceeded those of patients diagnosed with cardiac pain, as measured both by HADS and the HAI focusing on health anxiety specifically. In neither diagnostic group were mean anxiety scores significantly reduced subsequent to clinic attendance (non-cardiac, F = 1.39, NS; cardiac, F = 2.04, NS). For those with non-cardiac pain, levels of anxiety remained in excess of community norms 2 months after clinic attendance, while levels of depression for those with a cardiac diagnosis also significantly increased to above community norms after clinic attendance.

Examination of illness beliefs showed that a number of significant group differences were sustained over time (table 4). Those diagnosed with non-cardiac pain construed the condition as less subject to control by treatment, perceived themselves to have less personal control over their condition, and reported less illness coherence (less understanding of their symptoms) than the cardiac group at all time points. Respondents in both diagnostic categories reported less emotion associated with their condition over time, but this was significant only for the cardiac group: those with a cardiac diagnosis showed a significant decrease in emotion associated with their condition over time (F = 7.06, p<0.05).

DISCUSSION

The purpose of the present study was to assess psychological morbidity in a cohort of patients referred with chest pain to a rapid access clinic. Unsurprisingly, given that participants were experiencing symptoms that might presage serious illness, levels of anxiety across the study population were raised before clinic attendance, irrespective of subsequent diagnosis. However, despite the speedy access and assessment of chest pain there appeared no significant reduction in anxiety in the immediate aftermath of attendance. For both diagnostic groups, levels of anxiety exceeded community norms both before clinic attendance, and at 1 week after clinic attendance. For patients whose pain was non-cardiac almost two-thirds of the group remained clinically anxious 2 months after their clinic visit, while anxiety levels for those with cardiac diagnoses returned to levels close to normal. Mean HADS scores for those with non-cardiac pain were also markedly higher than the range of 4–6 usually expected in primary care.24

Depression was also increased for participants over the period of the study. Those with cardiac diagnoses showed an increase from community levels (11.8%) before clinic attendance to around 40% at later time points. This trajectory is not atypical for cardiac patients as the impact and ramifications of diagnosis are processed. However, given the independent role depression may play in subsequent cardiac events,25 and that affective response may not be fully evident at initial clinic attendance, continuous monitoring of psychological status may be necessary to assess whether further support or active interventions are needed. Non-cardiac participants also showed increased levels of depression but these remained constant throughout and appear to reflect depressive symptomatology before clinic attendance.

Strengths and limitations of the study

As far as we know, this is the first longitudinal study to have applied reliable and well-validated measures of psychological morbidity to examine a representative sample of patients in such settings. Despite expected attrition, those who remained in the study were demographically and clinically largely indistinguishable from those who dropped out. However, it is possible that those who remained in the study reflect a selection bias and differ psychologically from those who did not complete follow-up questionnaires. Differences in response rate and response bias in the norm populations for HADS and HAI, and those for the present study may also make comparisons less reliable. Potential methodological bias, however, does not diminish the clinical imperative to examine the psychological needs of substantial numbers of patients who report themselves to be anxious or depressed.

The circumscribed menu of questionnaires adopted permitted a balance between achieving adequate participant engagement and addressing the key research questions. However, complementary data on subjective experience of chest pain (intensity, relief-seeking behaviour and functional limitations) may have been useful to prime participants’ appraisals of their condition as well as providing a correlate to psychological morbidity. Data on healthcare utilisation, measured by subsequent consultations with primary and secondary providers and medication use, would also enhance the study.

Comparison with other studies

Previous studies have suggested that a significant number of patients who receive a non-cardiac diagnosis have comorbid anxiety states or panic disorder. Unexplained chest pain is also recognised as a presentation of hypochondriacal disturbance,26 and may reflect “somatised” emotional distress.27 Yet current indices of health anxiety, revealed in HAI scores, were no higher than those obtained for other outpatients at the study outset. Health anxiety did increase 1 week after the clinic for non-cardiac patients, significantly exceeding scores for cardiac patients and remaining significantly higher 2 months later, but still remained within normal parameters. Those with non-cardiac chest pain therefore appeared not to be unduly anxious about their health, and the increase in their scores may well reflect increased focus on health as a consequence both of the investigative procedures and remaining uncertainty about a definitive diagnosis.

In addition to the significant differences in psychological morbidity between groups, appraisals of symptoms also differed. From before clinic attendance those with non-cardiac pain felt that their symptoms were less amenable to their personal control, less likely to improve with treatment, and were less clear about the nature of their condition—a consistent difference sustained at follow-up. Attendance at a chest pain clinic offers an opportunity to shape patients’ appraisals through definitive messages so that symptoms experienced are less likely to be construed as heart disease. There is some tentative evidence from this study that cardiac patients, although anxious, report less emotion-laden appraisals after clear diagnosis is provided.

The role and influence of illness appraisals is increasingly documented in cardiac care. Patients who represent their chest pain as having severe consequences are least likely to delay attendance at emergency departments.28 For patients with a diagnosed myocardial infarction, beliefs about lesser consequences predict reduced illness-related disability and faster return to work, independent of actual pathology.29 Brief hospital intervention, designed specifically to alter patients’ maladaptive beliefs about the consequences of myocardial infarction, have significantly enhanced functional outcomes such as chest pain symptoms30 and underline the value of examining psychological problems in cardiac settings.

Yet in this study patient appraisals changed little as a consequence of clinic attendance. Up to 2 months after referral, patients with non-cardiac chest pain reported no increased sense of control or greater understanding of their condition. The current study cannot tease out reasons underpinning the lack of improvement, but it seems that any reassuring health messages offered during the clinic visit have little positive effect. Patient appraisals may not be sufficiently, or explicitly, dealt with by staff who are focused primarily on a cardiac diagnosis. Equally given the increased levels of anxiety displayed by participants, and their impact on attention and memory, patients may be unable to fully comprehend the health messages delivered in a busy diagnostic clinic. This is disappointing since the clinic offers an unrivalled opportunity to capitalise on a consultation with patients who are seeking plausible explanations for their pain.

Rapid access chest pain clinics, as they are currently set up therefore may not deal with the needs of the majority of patients attending. They may also be iatrogenic, and engender rather than alleviate distress. Extensive diagnostic procedures, even when revealing no cardiac problems, may increase erroneous beliefs about health status, and the advantages inherent in early cardiac diagnosis may be offset by an increasing number of chronically disabled patients with pain that remains unexplained.31

CONCLUSIONS

Identification of those patients who may require more detailed consultations and more intensive intervention to deal with concerns, fears and inappropriate beliefs about their condition, has been recommended to improve patients’ understanding of their condition, with benefits of reduced symptoms and more appropriate use of healthcare resources in other outpatient settings. Rapid access chest pain services should also undertake basic psychological screening, focusing particularly on anxiety and patient appraisals, as part of a comprehensive package of care, especially for patients diagnosed as having non-cardiac chest pain. This information would be valuable for subsequent management of patients in primary care. Further research should incorporate measures of pain intensity, behaviours and coping, as well as subsequent consultations and medication use, for a comprehensive analysis of psychosocial outcomes.