Family Practice Advance Access originally published online on April 12, 2006
Family Practice 2006 23(4):427-436; doi:10.1093/fampra/cmi125
Variations in cardiac interventions: doctors' practices and views
a Department of Primary Care and Population Sciences, University College London Royal Free Campus, London, UK
b Department of Psychology, University College London Royal Free Campus, London, UK
Correspondence to Professor Ann Bowling, Department of Primary Care and Population Sciences, University College London, Royal Free Campus, Rowland Hill St, London NW3 2PF, UK; Email: a.bowling{at}ucl.ac.uk
Received 1 August 2005; Accepted 24 January 2006.
 |
Abstract |
|---|
 Top Abstract BackgroundAimMethodsResultsDiscussionDeclarationReferences |
|---|
Objectives. To investigate referral rates for cardiac interventions by clinical specialty, to document doctors' reasons for referrals and to explore doctors' perceptions of the factors that influenced their clinical decisions.
Study design. Doctors completed a clinical decision-making exercise involving, in total, 6093 electronic patients with cardiac disease, and subsequently took part in the semi-structured interviews about influences on their decisions. Interviews were audio-recorded, transcribed and coded using a thematic approach, with the coding categories derived from the data.
Study setting. Eighty-eight doctors (GPs, care-of-the-elderly specialists, cardiologists) participated in the full study, in seven areas in southern, central and northern England. Complete interview data were analysed for 76 of these.
Principal findings. Not all patients who were eligible for specific investigations or treatment received these. The extent of variations in clinical decisions differed by type of intervention. Apart from the general reasons for referrals, doctors raised nine main influences on their actual decision making. The most commonly reported influence (‘barrier’) was poor access to equipment for intervention, which increased thresholds for investigation and treatment.
Conclusions. The current emphasis on achieving targets in the British NHS has led to a focus on easily measurable, but crude, process targets such as waiting lists. This study points to the need to include a broader quality assurance element to investigate the cluster of system failures which lead to variations in clinical decisions and thereby to inequitable treatment.
Keywords. barriers, cardiology, clinical, equity, referral rates.
|
Background |
|---|
|
TopAbstractBackgroundAimMethodsResultsDiscussionDeclarationReferences |
|---|
Coronary heart disease is a major cause of disability and mortality in the developed world; the UK has one of the highest rates of this disease along with relatively low intervention rates (e.g. with revascularization).1 The financial burden of current treatment is still considerable.2 The efforts of the UK Government to improve access to publicly funded services, including cardiology, have focused on reducing waiting time and setting minimum clinical standards for investigation and treatment, through national service frameworks.3 Waiting list figures can be an important, if crude, indicator of access to treatment, and a yardstick for service improvement. However, variations in their methods of construction have led to their validity being questioned4; they have been criticized for leading to service ‘rationing’,5 and for being easily distorted.6 A more comprehensive approach to investigating target, system and investigation failures would be beneficial to patient care and prevention of medical error.7–9
This points to the need for research on the situational influences on clinical practice, particularly given research showing wide variations in hospital admission rates by general practice, supporting research which shows variations in out-patient referrals, uptake of screening services, prescribing patterns and night visits.10 Research has indicated that factors other than the nature and characteristics of disease account for variations in cardiac patient management.11–28 And the first comprehensive audit of cardiac surgery the UK revealed that while there was a steady upward trend in the volume and complexity of cardiac surgery, there were also substantial variations in intervention rates,29 raising further concern over equity of access.30 The audit showed that 20% of patients undergoing surgery in some units were aged over 70 years, compared with just 10% in others. And the proportion of patients undergoing surgery and who had severe angina ranged from <10% to 40% between units. Again, such data emphasize the need for investigation of non-clinical influences on medical decision making.
Studies of clinical decision making have shown that, in the absence of adherence to established treatment protocols, treatment decisions are often made in ad hoc ways,31,32 and the choice of procedure partly depends on the clinician's preferred style as an interventionist or as a non-interventionist.33 Failure to take equity of patient access into account when analysing intervention and referral rates leads to false conclusions and wrong policy decisions. One published review of out-patient cardiology records concluded that 42% of patients studied had no interventions or treatment changes performed on clinic visits, and simply deduced from this that a substantial proportion of these patients did not require regular review.34 However, there is little existing data on the extent to which clinical decision making is influenced by situational and social factors.
|
Aim |
|---|
|
TopAbstractBackgroundAimMethodsResultsDiscussionDeclarationReferences |
|---|
Objective
To investigate referral rates for cardiac interventions by clinical specialty, including doctors' reasons for referrals, to explore doctors' perceptions of the factors that influenced their clinical decisions.
|
Methods |
|---|
|
TopAbstractBackgroundAimMethodsResultsDiscussionDeclarationReferences |
|---|
The methodology was multi-method, incorporating quantitative clinical judgement exercises35 and semi-structured interview methods. For the initial decision-making exercises, each doctor was presented with the clinical details of sets of 72 fictional electronic patients with cardiac disease (total across all doctors: 6093 patients), via a computerized simulation. They were asked to collect information about patients, and make decisions about their investigation and treatment as they would in a usual patient encounter. (As 32 variables were used in the judgement exercises, a large number of patients would have been required for inclusion, based on a usual judgment analysis design.35 But since not all information is likely to be searched for each case, it was possible to include fewer than is theoretically required. On the basis of previous research and experience it was estimated that, on average, participants would search for a large amount of information, but their judgements would be influenced by a small amount of this (4–7 pieces or ‘cues’). Thus after allowing for factorial combination of age, sex and ethnicity, 72 cases were presented to each doctor.)
Sex, age, occupational grouping and ethnicity were varied factorially across each set of patients. Other clinical information which was available for each case was varied at random across patients [e.g. symptom typicality and severity; risk factors (smoking, alcohol intake, body mass index, family history of disease, cholesterol level, blood pressure), current medication, co-morbidity, history of and results for (where available), investigations including levels for angiography, 12-lead ECG, thallium scan, exercise test, echocardiogram, chest X-ray, abdominal ultrasound, barium swallow and CT scan]. These covered normal-to-severe test results, and were based on anonymized samples obtained from local hospitals (see Annex 1, available online at http://fampra.oxfordjournals.org/).
A typical electronic scenario would include a visual image of the patient, with their details. For example, a 61-year-old Afro-Caribbean man presented with a burning feeling that lasted a few minutes. The doctor would need to probe for further information, producing a patient report of the pain starting on walking 200 yards, on quick exertion or sometimes on eating (without radiating anywhere else). The doctor then had the option of clicking on the screen to open windows on the patient's medical history (e.g. in this case eight different medical conditions were recorded), risk factors (e.g. ex-smoker; blood pressure 155/95; no family history of heart disease), current medication (e.g. Isosorbide dinitrate, 20 mg once daily), availability of test results, previous treatment and so on, before making a clinical decision about (further) investigation and/or treatment. Each test result option opened up a visual image (e.g. of a chest X-ray and report or a blank screen indicating none available).
Doctors' decisions were compared with those derived from internationally agreed, deliberately tight criteria for clinical intervention, and which were used by one of the authors (AB) in previous studies,15,16,21 and then analysed by the characteristics of the doctor and the patient. In semi-structured interviews after completion of the judgement exercises, the doctors were asked for their reasons for such referral decisions (to a cardiologist, cholesterol testing, exercise tolerance testing, angiography, revascularization) in their real-life settings. They were next asked about their perceptions of the clinical and non-clinical (e.g. situational) influences on their decision making, again in their real-life settings. The interviews were audio-recorded, transcribed, coded using a thematic approach, derived from the data and analysed. DF carried out the interviews; the audio tape recordings were transcribed by one of three transcribers. One of the investigators (AB) read the transcripts of each interview several times, identified the broad and detailed themes, which emerged, and developed the coding frame for their categorization. The process was checked independently by two other investigators (CH and NH). The themes were analysed using SPSS to produce frequency and contingency tables, means and percentages; t-tests and chi-square tests of statistical significance were conducted to test for differences in referrals by specialty, doctors' characteristics and by the themes mentioned. Unless otherwise stated, differences are reported where they achieved at least the P < 0.05 level of significance.
The sample
The aim of the sampling was to recruit equal numbers of cardiologists, specialists in care of the elderly and GPs in England. The intention was to provide insightful information on the process of decision making, rather than population estimates.
With the support of local health authorities (regarding list of GPs), the British Cardiac Society and the British Geriatric Society, who enabled the researchers to mail the doctors on their lists, more than 250 doctors responded positively to mail-shots asking for participants. The respondents were spread across southern, central and northern England, and were matched by geographical area, in order to minimize variance (e.g. due to differing health budgets per area).
|
Results |
|---|
|
TopAbstractBackgroundAimMethodsResultsDiscussionDeclarationReferences |
|---|
Respondents' characteristics
Eighty-eight of the 250 respondents to the mail-shots were selected to participate in the full decision-making study, based on the ability to match a complete trio (GP, cardiologist, care-of-the-elderly specialist) to an area (see above). They comprised 30 cardiologists (CA), 30 care-of-the-elderly specialists (CE) and 28 GPs, and were based in six areas of London, as well as Northamptonshire, Sussex, Nottinghamshire, Leicestershire, Hampshire and Kent. Complete data were available for the decision-making (or, ‘judgement’) analysis for 85 doctors (29 cardiologists, 28 care-of-the-elderly specialists and 28 GPs). Between them they saw and made decisions for 6093 electronic patients.
Complete interview data for analysis were obtained for 76 of these 88 participating doctors; the remaining tape recordings were of insufficient high quality to transcribe or to code. This interview sample included 29 cardiologists, 21 care-of-the-elderly specialists and 26 GPs. Sixty-four per cent (18) of cardiologists were specialist registrars and the rest were consultant grade; in contrast, 39% (11) of the care-of-the-elderly specialists were specialist registrars and the rest were consultant grade (
2, P < 0.001). The GPs had been in general practice for a mean of 10.50 years (95% CI: 4.9–16.11), the cardiologists had been working within their specialty for a mean of 9.57 years (6.6–12.06), and the care-of-the-elderly specialists had been in their speciality for a mean of 8.57 years (5.5–12.03) (not significant with t-test for independent samples). Forty-three of the 76 doctors whose tapes were analysed were male. Thirty-nine were aged under 40; 14 were aged 40–50, 9 were aged 50–60 and the remainder were aged 60 or over. Fifty four doctors were white, eight were from India or Bangladesh and the remainder were in other ethnic groups.
Variations in clinical decisions
The mean number of electronic patients seen by cardiologists, care-of-the-elderly physicians and GPs, who had indications, but not contraindications, for treatments did not differ between these three specialist groups [mean number of patients with indications for exercise testing: 69.47; cholesterol testing: 40.58; angiography: 66.54 (cardiologists and care-of-the-elderly specialists only); revascularization: 46.63 (cardiologists and care-of-the-elderly specialists only)].
Figure 1 shows the percentages of patients, with indications for referral who were referred by type of doctor (note: only eligible doctors shown, thus GPs are not shown in Figure 1e and f as they do not refer directly for angiography or revascularization; cardiologists are not shown in Figures 1c and d as they are the source referred to).
|
Although most patients with indications for a cholesterol test were given one, the proportions referred for other investigations and treatment interventions were far less; there was variation within and between specialties in the proportion of patients referred. The trends in Figure 1 illustrate these variations in referrals both within and between specialties.
Just over three-quarters (76%, 65) of doctors referred 80–100% of their patients for cholesterol testing where they had indications for this. There were differences between specialties in their referral rates for these patients, with care-of-the-elderly specialists making the least referrals (2, P < 0.05; caution as less than 5 expected counts in some cells). Figure 1a illustrates these differences.
Fewer, approximately one-third (32%, 27) of all doctors referred between 80 and 100% of indicated patients for exercise tolerance testing. Figure 1b shows that GPs were less likely than either group of specialists to refer indicated patients for exercise testing: (2, P < 0.001; caution as less than 5 expected counts in some cells). It is possible that some GPs and care-of-the-elderly specialists simply preferred to refer their patients with indications for exercise tolerance testing directly to the cardiologist for further assessment, rather than investigate them themselves (and this might have reflected their lack of direct/easy access to testing facilities—see next section). However, relatively few doctors (14 out of 56 GPs and care-of-the-elderly specialists) referred even 60% or more of these patients to a cardiologist (no significant differences between types of doctor). Similarly, few11 referred even 60% or more of their patients with uncontrolled, unstable angina to the cardiologist (no significant differences between types of doctor) (Figure 1d).
Cardiologists and care-of-the-elderly specialists could choose to refer their patients for angiography or revascularization. Over half of the cardiologists (55%, 16) and none of the care-of-the-elderly specialists referred 80–100% of their indicated patients for angiography (2, P < 0001; caution as less than 5 expected counts in some cells). And just over half, of cardiologists (52%, 15), and none of the care-of-the-elderly specialists referred between 80 and 100% of indicated patients for revascularization (2, P < 0.001; caution as less than 5 expected counts in some cells).
The characteristics of the doctors, including grade of specialist (consultant, specialist registrar) and number of years within the specialty were analysed and there were no significant differences with referrals once the type of specialty was controlled for.
The clinical judgement analyses provided the context for the next part of the research. In the interviews following the decision-making exercises, doctors were asked, for their reasons for, and against, referring patients to a cardiologist, for cholesterol testing, for exercise tolerance testing, for angiography and for revascularization (in their real-life settings). Open-ended questioning was used. Following this, they were asked, again using open-ended questions, about the clinical and non-clinical influences on their decision making in their real-life settings.
Doctors' reasons for referrals
The reasons doctors gave for and against each referral decision are shown in Table 1. Doctors indicated that if they were uncertain about a diagnosis or treatment, they would be more likely to refer to a cardiologist, refer for an angiogram or order an exercise tolerance test. In relation to the latter, the GPs and care-of-the-elderly doctors mentioned the need to refer not just because of the uncertainty of diagnosis, but also the presence of symptoms that were difficult or inappropriate for them to treat. The need to ‘do something’, and the potential benefit from doing something, was also mentioned in association with referral to a cardiologist, angiography and revascularization. Patient's wishes, attitudes and life circumstances were mentioned as reasons for, and against, referral to a cardiologist, and for more invasive referrals (angiography and revascularization). Socio-demographic factors such as ethnicity, age, and job were mentioned both as reasons for, and reasons against all types of referral. Doctors emphasized coronary disease risk factors and symptoms when discussing cholesterol tests, and most patients with indications for this test were referred.
|
Doctors' perceptions of factors that influenced their clinical decisions
When asked about the clinical and non-clinical influences to their decision making in their real-life settings, the following nine themes emerged (‘barriers’): lack of equipment, or lack of direct or easy access to equipment, for interventions; insufficient time; budgetary constraints; staffing shortages; the effects of hospital policies or consultant policies/clinical guidelines on criteria for investigation and referral; long waiting lists for referrals; the personal characteristics of consultants referred to; doctors' own characteristics and interests; and patient characteristics. Illustrative quotes for these are shown in Boxes 1 and 2.
| BOX 1 Examples of doctors' views of influences on decision making: structural factors Access to equipment ‘Equipment resources impact upon your decision making. You try to select the patients that you feel would benefit more from these tests. Not always young people, but people who you feel that the severity of their chest pains is making their quality of life worse. Whereas if someone is elderly and sedentary then sadly, you have to sometimes forget these people.’ (CE167) ‘Exercise tests are easy to get here, but less accessible on (my) other site as I have to make a referral through the main (hospital) site - so I ... refer less (there)’ (CE125) ‘We have no on-site angioplasty ... . so I refer less freely for angiography and revascularisation than I would if I had this facility.’ (CA120) Time, staffing and budget ‘There are not enough technicians to do tests, so tests are reserved for the most deserving ones.’ (CA101) ‘We only have one cardiologist and one who comes in from X Hospital once a week, so we tend not to overload our cardiologists.’ (CE102) ‘Budgets lead to cheaper and smaller dose statin prescriptions.’ (GP147) ‘Budgets lead to rationing statins to people with a certain level of risk of having an ischaemic event.’ (GP183) Waiting lists ‘Availability of investigation is one thing that affects our decision making. In this hospital there is about a 3 month waiting time for tests. So this tends to affect your level of referral as the more you ask for the longer you have to wait. Therefore your criteria go up a bit really.’ (CE167) ‘Requesting angiography when I'm not working in a teaching hospital ... is difficult . (long waiting lists for external referrals). I can see that resources ... affect your decision. You have to do the best you can with what you have for the patient ... I'm sure that treatment...outside the big teaching centres ... (is) ... very different.’ (CA146) ‘ ... - lack of angio resources limits their use, as putting a patient on the (waiting) list clogs it for others. My patients are managed personally in different ways on different sites depending on resources—district general hospital patients with unstable angina don't get referred to (the) specialist centre due to bottle necks (on the waiting list), in contrast to my own specialist site where they do get things done.’ (CA145)
|
| BOX 2 Examples of doctors' views of influences on decision making: policies and personal characteristics Hospital and Consultant policies ‘ ... it works very well in anyone that's younger who would probably get diverted reasonably quickly to the cardiologists, anyone that's older may well go through various other things ... some people get a good service and some people get a worse service.’ (CE106) ‘Consultant policy ... influences practice ... thresholds for intervention ... some cardiologists won't take high risk patients; some cardiologists are aggressive, investigate lots; some less interested in older people. I personally check the diary to see which consultant is on duty and decide whether to refer on that basis ... depending on the patient. If I feel I want aggressive treatment for a patient then I refer to a cardiologist, or if I want advice but nothing aggressive then I refer to other types of specialist (e.g. in hypertension).’ (CE121) Characteristics of consultants referred to ‘I chose the consultant by personality, to match the personality of the patient ... (and) the patient's needs’. (GP111) ‘ ... (chose consultant) by personal acquintance ... attendance at post-graduate meetings’. (GP150) Doctors' characteristics and interests ‘I'm sure that (ethnicity) has a huge influence. My interests are in Asian coronary artery disease being an Asian doctor. I probably try to compensate for other doctors who don't take the pain as seriously. (CA117) Patient characteristics ‘Age is a double edged sword because the older you are, the higher the risk of an angiogram, so once you start hitting 75, 80, 85 mark you then start getting put-off because you worry about complications. So those are the clinical risk factors which would sway me.’ (CA106) ‘Quality of life is important not age. If you are in your 80s and your chest pain stops you going to the post-office then it is just as disabling in terms of what your life is all about as in a young person's case.’ (GP125).
|
Access to equipment for interventions. Just over half the doctors who were interviewed, 53% (40), said their access to equipment was poor and thus their threshold for referral for investigations and interventions was higher. Those who did not mention this commented that they had good access to equipment, and that they therefore had a low referral threshold.
Problems raised related to lack of equipment or its limited availability, which led to longer waiting times, which in turn was said to lead to implicit rationing. Of the 40 doctors who said access to equipment was poor, 40% (16) had reported that they did not have direct or open access to cardiological investigations or facilities and had to refer the patient on for these (i.e. to a hospital specialist in the case of GPs), or to another specialist, or to another hospital site. Ten of these were GPs, and they reported that they had no direct or open access to investigative services in cardiology, and instead had to refer patients to a cardiologist for all tests. These GPs complained about their lack of open access to investigative equipment. For example, they admitted that not having an ECG machine in their own practices limited the number of ECG requests they made. A care-of-the-elderly specialist stated bluntly what the consequences of lack of easy access for their patients were (see quote 1, Box 1). This was said to lead to patients' treatment varying by hospital site. The inequities were also said to be between specialist and district general hospitals. The study specialists who worked in both teaching (main post) and non-teaching hospital sites were able to distinguish clearly between the type of care offered between sites owing to the problem of lack of equipment in general hospitals, and which also had implications for waiting lists (see later) with cross-site (hospital to hospital) referrals (see quotes 2, 9 and 10, Box 1). Both hospital doctors and GPs said that restricted access to equipment raised their referral thresholds, thereby having the effect of delaying or rationing patient care.
Time, staffing and budgetary constraints. Forty-three per cent (33) of both hospital doctors and GPs mentioned that lack of time adversely affected their decision making, also commenting that the effect was to make more referrals to specialists (thereby off-loading the time problem), and even lowering the threshold for ordering diagnostic tests owing to insufficient time for an examination and history taking. As one doctor simply stated ‘Lack of time leads to cut corners.’ (GP103)
In addition, 41% (31) commented that inadequate staffing levels, which also led to longer waiting lists, affected their decision making. This, once again, was said to lead to poorer patient care, prioritization and rationing decisions among hospital doctors. Budgetary restraints on both medical and surgical practice were mentioned by almost one-fifth of doctors: 18% (14) said that available budgets directly affected their prescribing behaviour and 17% (13) mentioned other restrictive effects of budgets. The GPs were most likely to mention budgetary restrictions on certain prescriptions, particularly statins (see quotes 6 and 7, Box 1).
Waiting lists. While limited access to equipment and staffing problems were said by doctors to be inevitably linked to longer waiting lists, waiting lists were also raised in their own right as a restraint on clinical practice (see last three quotes, Box 1). Twenty-eight per cent of doctors (n = 21:11 cardiologists, 4 care-of-the-elderly specialists and 6 GPs) said that the length of the waiting list influenced their decision making, again leading to higher referral thresholds and to different standards of care. However, there was no association between mentioning the length of the waiting list as a barrier to practice and the actual waiting list times (range 5–36 weeks in study areas). However, the lack of associations should be interpreted with caution as open-ended methods of questioning elicit responses that are most likely to be the most pertinent, which is not the same as prevalence.
Hospital and consultant policies. A smaller proportion of doctors, 13% (10) said that the hospital, practice or consultant firm's clinical policy/guidelines influenced their decision making. This related to access to facilities (10), adherence to criteria (e.g. for referral to cardiology) (10), age-related policies (6) and prescribing (5). Again, the policies were said to have restrictive effects, in some cases disadvantaging elderly patients (see quotes 1 and 2, Box 2). GPs also commented that hospital policy also affected their practice because it influenced specialists' prescribing criteria and therefore the type of medication GPs were then requested to continue patients on; whether GPs were granted open access to hospital diagnostic facilities and clinics; and the volume of hospital referrals of patients back to the GP for management (to save the hospital's budget).
Characteristics of consultants referred to. Just over one-fifth of specialists and GPs (22%, 17) said that they deliberately chose the consultant they wished to refer (on) to by their personal and professional characteristics. Some mentioned they chose them by their personality (e.g. their approachability), others chose them by their clinical policies, their clinical reputation, efficiency or clinical interests; others (GPs) chose those they had met (e.g. at post-graduate education meetings) (see quotes 3 and 4, Box 2). And as one doctor indicated, the patient's likely pattern of investigation and treatment is likely to be influenced by the GP's decision on the type of specialist referral to make, and also by the particular specialist.
Doctors' characteristics and interests. Forty-six per cent (35) of the doctors said their own characteristics, family medical history, and special interests, influenced their decision making. Sixteen doctors aged said their own age influenced them and made them more sensitive to likelihood of the presence of coronary disease in older patients (although only three over these were aged over 40), five mentioned that their own (Asian) ethnicity was influential (see quote 5 in Box 2), four mentioned their family medical history of heart disease, two said their interests in gender influenced them (one was male, one female), and eight mentioned combinations of these.
Patients' characteristics. Finally, 18% (14) of doctors freely said the patients' own characteristics influenced their clinical decisions. For example, eight doctors said that the patient's age influenced their decision making (six of these said they were less likely to refer older people), three mentioned the patient's ethnicity (Asian men were said to be at higher risk of cardiovascular disease, and thus they were more likely to refer them), two mentioned the patient's gender (women's lower risk of cardiovascular disease made them either more or less cautious in investigating their symptoms), and one doctor said the patient's level of mobility and quality of life affected their clinical decisions (see below). Their comments about age and quality of life varied (see quotes 6 and 7 in Box 2).
The characteristics of the doctors, including grade of specialist (consultant, specialist registrar) and number of years within the specialty, were analysed (controlling for specialty), and there were no significant differences with perceived influence on practice. There were also no significant differences between doctors' referral rates for the interventions and mentioning barriers to practice or type of barrier.
|
Discussion |
|---|
|
TopAbstractBackgroundAimMethodsResultsDiscussionDeclarationReferences |
|---|
The main aims of this study were to investigate referral rates for cardiac interventions by clinical specialty, to document doctors' reasons for referrals, and to explore doctors' perceptions of the factors that influenced their clinical decisions. The clinical judgement study revealed substantial variations between doctors in most of their clinical decision making, consistent with studies of cardiac patients' case notes.15,16 The interview study provided insight into doctors' perceptions of their reasons for referring, and of the main influences on (‘barriers to’) decision making. These provided insight into why failures in intervention occur, and need testing objectively. The findings may not be unexpected, but research data on this topic are rare. While the study was based on a sizeable sample of clinicians, it was not based on a random sample, and the results are therefore of unknown generalizability. However, the intention was to provide insightful information on the process of decision making.
The main barriers to referral identified related to the ways in which services were resourced, staffed and organized locally, including the length of waiting lists. These were often said to influence whether patients were referred for investigation and treatment. In particular, doctors reported that ease of access to equipment was either a key barrier to, or a lever for, appropriate referrals. The ease of access was said to influence their thresholds for intervention, so largely determining whether appropriate referrals were made, and whether patient access to care was equitable. These findings were consistent with those from another, smaller, qualitative study of 30 GPs on barriers to accurate diagnosis and management of heart failure.8
The low referral rates for cholesterol testing, angiography and revascularization among care-of-the-elderly specialists are likely to reflect their lack of specialist role in cardiac medicine. It might also reflect their greater familiarity with very frail older patients for whom invasive procedures might not always be appropriate, thus militating against referrals. However, their referral rates for a cardiology opinion were relatively low and suggest that older patients with indications for intervention are also not being considered for a specialist opinion. This area requires further investigation.
The interview study showed that almost a fifth of doctors said they were influenced by the patient's socio-demographic characteristics, which is consistent with the results of the wider judgement analysis (Harries C, Forrest D, Harvey N, Bowling A, manuscript submitted). A recent study of cardiology patients' records, which showed lower use of treatment and investigation among older patients, concluded that this might be explained by the lack of an evidence base as most clinical trials have excluded elderly patients, and called for investigations and treatment trials to be conducted specifically with patients aged >70 years to establish evidence base against which standards can be set and services can be audited.36
The importance of these results is that the doctors themselves reported variations in practice, inequity of care and differences in treatment between sites. These data contribute to the limited body of knowledge on barriers to appropriate decision making in clinical practice, and have implications for health policy. The current emphasis on achieving targets in the British health service has led to a focus on easily measurable and distortable process targets such as waiting lists. But system failures both create waiting lists and restrict their use.
Future health service strategy should include a broader quality assurance element, and even an 'equity audit,37 and build in investigation of system failures leading to barriers to appropriate practice. The greater use of existing, large general practice datasets could also be used to provide basic data to support such exercises. Research has revealed a failure in health policy, and in economic evaluations of health care, to strike an acceptable balance between equity and efficiency goals when these conflict.38,39 However, it can be questioned whether an equitable service is a realistic policy aim, given clinical variations between doctors (e.g. in referral rates),40 and the impact of clinical uncertainty in medicine.41
|
Declaration |
|---|
|
TopAbstractBackgroundAimMethodsResultsDiscussionDeclarationReferences |
|---|
Source of funding: The study was financed by ESRC award number ROO238247. The researchers were independent from the funders.
Ethical approval: not required at the time of the study.
Conflict of interest: none.
Role of authors: DF administered the electronic judgement exercises and conducted the interviews; the audio tape recordings were transcribed by one of three transcribers. CH designed the judgement analysis programme with NH, and analysed the judgement data using Minitab. AB read the transcripts of each interview several times, identified themes, developed a coding frame and programme for categorization of themes using SPSS, conducted the coding, and merged Minitab and SPSS data files. All authors contributed to the design and structure of the study, the electronic programme and interview-coding frame, agreed on the final themes, and inspected the categorized responses. CH and AB analysed the data presented here. AB wrote this paper with CH; all authors contributed. Name of the guarantor of the data contained in this paper is Professor Ann Bowling.
|
Acknowledgments |
|---|
We would like to thank the doctors who gave up their time to participate in the study, Dr Nigel Dudley and Dr Steve Iliffe for their helpful clinical advice, Professor Paul Wallace for his comments on an earlier draft of this paper; and the transcribers of the audio-taped interviews.
|
Notes |
|---|
Bowling A, Harries C, Forrest D and Harvey N. Variations in cardiac interventions: doctors' practices and views. Family Practice 2006; 23: 427–436
|
References |
|---|
|
TopAbstractBackgroundAimMethodsResultsDiscussionDeclarationReferences |
|---|
1 Organisation for Economic Cooperation and Development (OECD). Health data 9. Paris, CREDES-OECD.
2 BHF statistics database British Heart Foundation. (2003) Coronary heart disease statistics. (British Heart Foundation, London).
3 Department of Health National Service Framework on coronary heart disease. (2000) (Department of Health, London).
4 Audit Commission. Waiting for elective admission: acute hospital portfolio. Review of national findings. London: Audit Commission. Available at www.audit-commission.gov.uk/reports/.
5 Devlin N, Harrison A, Derrett S. (2002) Waiting in the NHS: Part 1—a diagnosis. J R Soc of Med 95:223–226.
6 Beecham L. (2003) BMA annual representative meeting. NHS targets are distorting clinical need. Br Med J 327:72.
7 Derrett S, Devlin N, Harrison A. (2002) Waiting in the NHS, Part 2: a change of prescription. J R S Med 95:280–283.
8 Fuat A, Hungin APS, Murphy JJ. (2003) Barriers to accurate diagnosis and effective management of heart failure in primary care: qualitative study. Br Med J 326:196.
9 Dovey SM, Meyers DS, Phillips RL, et al. (2002) A preliminary taxonomy of medical errors in family practice. Qual Saf Health care 11:233–238.[CrossRef][Medline]
10 Reid FDA, Cook DG, Majeed A. (1999) Explaining variation in hospital admission rates between general practices: cross sectional study. Br Med J 319:98–103.
11 Kravitz RL. (1999) Ethnic differences in use of cardiovascular procedures: new insights and new challenges. Ann Intern Med 130:231–233.
12 Rathore SS, Berger AK, Weinfurt KP, et al. (2000) Race, sex, poverty and the medical treatment of acute myocardial infarction in the elderly. Circulation 102:642–648.
13 Gregory PM, Malka ES, Kostis JB, Wilson AC, Arora JK, Rhoads GG. (2000) Impact of geographic proximity to cardiac rehabilitation services on service utilisation. Med Care 38:45–57.[CrossRef][ISI][Medline]
14 Pettigrew M, McKee M, Jones J. (1993) Coronary artery surgery: are women discriminated against? Br Med J 306:1164–1166.[ISI][Medline]
15 Bowling A, Bond M, McKee D, et al. (2001) Equity in access to exercise tolerance testing, coronary angiography, and coronary artery bypass grafting by age, sex and clinical indications. Heart 85:680–686.
16 Bond M, Bowling A, McKee D, et al. (2003) Does ageism affect the management of ischaemic heart disease. J Health Serv Res Policy 8:40–47.[CrossRef][Medline]
17 Stone PH, Thompson B, Anderson HV, et al. (1996) Influence of race, sex and age on management of unstable angina and non-Q-wave myocardial infarction. The TIMI III registry. JAMA 275:1104–1112.[Abstract]
18 McLaughlin TJ, Soumerai SB, Willison DJ, et al. (1996) Adherence to national guidelines for drug treatment of suspected acute myocardial infarction: evidence for under treatment in women and the elderly. Arch Intern Med 156:799–805.[Abstract]
19 European Secondary Prevention Study Group. (1996) Translation of clinical trials into practice: a European population-based study of the use of thrombolysis for acute myocardial infarction. Lancet 347:1203–1207.[ISI][Medline]
20 Barakat K, Wilkinson P, Deaner A, et al. (1999) How should age affect management of acute myocardial infarction? A prospective cohort study. Lancet 353:955–959.[CrossRef][ISI][Medline]
21 Dudley N., Bowling A., Bond M, et al. (2002) Age- and sex-related bias in the management of heart disease in a district general hospital. Age Ageing 31:37–42.
22 Hannan EL, Kilburn H, O'Donnell JF, Lukacik G, Shields EP. (1991) Interracial access to selected cardiac procedures for patients hospitalised with coronary artery disease in New York State. Med Care 29:430–441.[ISI][Medline]
23 Wenneker MB, Weissman JS, Epstein AM. (1990) The association of payer with utilisation of cardiac procedures in Massachusetts. JAMA 264:1255–1260.[Abstract]
24 Gatsonis CA, Epstein AM, Newhouse JP, Normand SL, McNeil BJ. (1995) Variations in the utilisation of coronary angiography for elderly patients with an acute myocardial infarction. Med Care 33:625–642.[ISI][Medline]
25 Gregory PM, Malka ES, Kostis JB, Wilson AC, Arora JK, Rhoads GG. (2000) Impact of geographical proximity to cardiac revascularization services on service utilization. Med Care 38:45–57.[CrossRef][ISI][Medline]
26 Ayanian JZ and Epstein AM. (1991) Differences in the use of procedures between men and women hospitalised for coronary heart disease. N Engl J Med 325:221–225.[Abstract]
27 Steingart RM, Packer M, Hamm P, et al. (1991) Sex differences in the management of coronary artery disease. Survival and ventricular enlargement investigators. N Engl J Med 325:226–230.[Abstract]
28 Clarke KW, Gray D, Keating NA, Hampton JR. (1994) Do women with acute myocardial infarction receive the same treatment as men? Br Med J 309:563–566.
29 National Audit of Cardiac Surgical Database Report, 1998. (1998) Birmingham: Society of Cardiothoracic Surgeons of Great Britain and Ireland.
30 Woodman R. (1999) Cardiac surgery audit raises concern over equity of access. Br Med J 319:277.
31 Kee F, McDonald P, Kirwan JR, Patterson CC, Love G. (1997) The stated and tacit impact of demographic and lifestyle factors on prioritisation decisions for cardiac surgery. QJM 90:117–123.
32 Harries C, Evans JStBT, Dennis I, Dean J. (1996) A clinical judgment analysis of prescribing decisions in general practice. Le Travail Humain 59:87–111.
33 Chapple A and Gatrell A. (1998) Variations in use of cardiac services in England: perceptions of general practitioners, general physicians and cardiologists. J Health Serv Res Policy 3:153–158.[Medline]
34 Hughes ML, Leslie SL, McInnes GK, McCormac K, Pedan NR. (2003) Can we see more out-patients without more doctors? J R Soc Med 96:333–337.
35 Cooksey RW. (1996) Judgment Analysis. Theory, Methods and Applications (Academic Press, San Diego).
36 Williams RI, Fraser AG, West RR. (2004) Gender differences in management after acute myocardial infarction: not ‘sexism’ but a reflection of age at presentation. J Public Health (Oxf) 26:259–263.[Medline]
37 Whitehead M. (1994) Equity issues in the NHS: who cares about equity in the NHS? Br Med J 308:1284–1287.
38 Sassi F, Le Grand J, Archard L. (2001) Equity versus efficiency: a dilemma for the NHS. Editorial. Br Med J 323:762–763.
39 Sassi F, Archard L, Le Grand J. (2001) Equity and the economic evaluation of health care. Health Technol Assess 5:1–138.[Medline]
40 Bevan G. (1998) Taking equity seriously: a dilemma for government from allocating resources to primary care groups. Br Med J 316:39–42.
41 Djulbegovic B. (2004) Lifting the fog of uncertainty from the practice of medicine. Br Med J 329:1419–1420.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||