|Home | About | Journals | Submit | Contact Us | Français|
The need to improve doctors' access to health care by reducing the barriers they experience has been regularly described in the literature, yet the barriers experienced are not well defined, despite the volume of expert opinion in this area.
To define what is known about doctors' access to health care from the data within the current literature.
A systematic review of studies of doctors' health access.
A systematic search of MEDLINE® and CINAHL, supplemented by citation searches and searches of the grey literature, identified both quantitative and qualitative studies. Two reviewers used specific criteria for inclusion of studies and quality assessment. The data were tabulated and analysed.
Twenty-six articles met the inclusion criteria. The paucity of data and the overall poor quality of those data are highlighted. Despite this, many doctors appear to have a GP, but this does not ensure adequate health access. Systemic barriers to healthcare access (long hours and cultural issues) are more significant than individual barriers.
Expert opinion in this field is supported by poor-quality data. The current knowledge reveals important similarities between doctors and the general population in their healthcare access, especially with mental health issues. Understanding this may help the medical profession to respond to these issues of ‘doctors’ health' more effectively.
Doctors are often said to be healthier than the general population because their standard mortality rate is lower.1 However, doctors have similar rates of chronic illness and have the same preventive health needs as the general community.2 The literature on doctors' health refers to the need to improve doctors' access to health care by reducing the barriers they experience.3–5 This can only be achieved if current understanding of doctors' health-access behaviours and the barriers they experience in accessing this care is adequately informed. As a first step, this article presents a systematic review of the evidence associated with doctors' health-access behaviours and the barriers they experience.
A systematic search of the literature used the electronic databases MEDLINE® (1966–2007) and CINAHL (1982–2007). Keywords were identified through an iterative search strategy, including further search terms as relevant articles were found. Articles were also identified by hand searching citation lists and grey literature. Searches were restricted to articles whose full texts were available in English.
This review presents the current data on doctors' healthcare access. Recognising gaps in knowledge will help focus future research projects. Comparison of the health-access behaviours of doctors with those of the general community is encouraged.
The final search strategy for MEDLINE® was: ((Physicians/ OR Physicians, Family/ OR Physician Impairment/ OR Internal Medicine/ OR Family Practice/) AND (Self Care/ OR Self Medication/)) OR ((Physicians/ OR Physicians, Family/ OR Physician Impairment/ OR Internal Medicine/ OR Family Practice/) AND (Health Behavior/)).
A similar strategy was used for CINAHL.
An assessment tool was used to ensure studies were assessed in terms of their internal validity (identification of bias and confounding factors),6 reporting quality, and reliability (external validity).7 Criteria for questionnaire-based studies included having a response rate?≥40%. Although this is relatively low, this response rate was selected for this systematic review because medical practitioners are known to be difficult to survey.8–10
Assessments for criteria and quality were made by two authors. Disagreements on inclusion were resolved by consensus.
Quantitative results were tabulated. Dominant themes were determined from the qualitative comments, and these were categorised and tabulated. Statistical meta-analysis was inappropriate because of the heterogeneity of study populations, methods, and outcome measures.
Searches of the electronic databases and citations lists led to the inclusion of 26 articles (Figure 1) for this systematic review. Twenty-three surveys of medical practitioners were found (six of which also reported some responders' comments), and three qualitative studies. The details of these are set out in Appendices 1 and and2.2. A list of excluded papers and reasons for exclusion is available.
The data were gathered from many countries including seven from the UK, six from the US, five from Australia, two from Ireland, and one each from Canada, New Zealand, Israel, Finland, Norway, and Switzerland. Most studies used registration lists from medical boards or colleges to randomly select participants. Nine studies focused on GPs, and five focused on junior doctors. Most studies were postal surveys, although the qualitative studies relied on focus groups and interviews (Appendices 1 and and22).
The three qualitative studies were rated as high-quality studies meeting at least six of the seven quality criteria.6 The quality of the surveys, however, was affected by several factors: in the majority of cases insufficient information was available to assess the questionnaire form (given to the physicians), instructions, administration, and previous piloting of the measures used.7 Low response rates, retrospective self-reporting (without objective verification of data), and incomplete data available on non-responders also affected quality. Although some studies collected demographic data of the non-responders, it was not possible to obtain information regarding their healthcare access.8
Table 1 shows that between one-fifth and nearly all responders were registered with a GP. This disparity reflects the different national health systems associated with each study. Some systems require compulsory patient registration with a GP, for example the NHS of the UK. Therefore, studies from the UK would be expected to have a greater than 90% registration (Table 1). The study populations also varied with regard to the specialties, age, and sex of the medical practitioners surveyed, and these factors may account for the differing responses reported. Choosing a GP is a complex process, often fraught with difficulty for doctors.11–14 Female doctors were more likely to describe difficulty with this,10 but were more likely to find an independent doctor.15 Many doctors do not choose a GP who is independent (that is someone who is not a close friend or relative, including spouse, and who is not a practice partner) (Table 1). Regardless of whether or not they have a GP, most doctors are satisfied with their health care.10
Some studies described how often doctors consulted other doctors, either formally or informally. This review used Pullen's definitions of formal and informal consultations.9 That is ‘in a formal consultation, the doctor sees his or her colleague as a usual patient would. An informal consultation may occur in passing; for example, as a “corridor consultation”’.9 Between 24% and 87% of responders stated they had consulted a doctor within the last year (Table 1). The higher rates probably included informal consultations with colleagues, although this was not clearly stated in some studies. Informal or ‘corridor’ consultations appear to be more common than formal care.16,17 Over one-third of doctors used corridor consultations.17,18 Even when the consultation was formal, the doctor may not have consulted their usual doctor.
Two studies found that doctors with a GP were more likely to have received preventive health care;8,19 however, having a GP does not always ensure better health access for all medical conditions. Hypothetical vignettes showed that having a GP predicted an appropriate treatment choice in relation to physical health, but not for mental health issues.10,17 Doctors with a chronic illness were less likely to access formal health care,18 even though they may be more likely to have a GP,17,19 although the data are conflicting.18
A sex difference in relation to consultation with a doctor was also evident. Female doctors consult other physicians more frequently than male doctors,9,12,20,21 although this may reflect a tendency to access more informal care, especially if they had a medical spouse.12
Between one-quarter and almost all doctors reported ‘self-treatment’ (Table 1). Most doctors felt that it was acceptable to self-treat, especially for minor illnesses.10,22 The variation in these reported percentages of self-treatment may lie in the definition of what constitutes ‘self-treatment’. This was not clearly defined in the majority of the studies. Self-treatment was more common for GPs than specialists.9,23 Rosvold and Bjertness reported that even if a doctor had consulted another doctor within the past 3 years (68%), 76% still self-treated.12
Table 2 lists the barriers reported by doctors. To facilitate interpretation, these barriers are grouped into categories and subcategories as described in Table 2. The ‘Patient’ category reflects the barriers specifically related to the doctor–patient seeking health care. The ‘Provider’ category details the barriers predominantly under the control of the provider of the medical care. The ‘System’ category reflects barriers within the medical system itself. Although these system barriers may be expressed in the individual behaviours of a patient or provider, they are barriers that are difficult for any individual to alter without the system itself changing.
Embarrassment was a recurrent theme identified in a number of articles. Davidson and Schattner found that 71% of doctors described themselves as embarrassed when seeing another doctor.10 However, the term ‘embarrassment’ is used by doctors to reflect many different experiences. McKevitt and Morgan clarified this, explaining that some doctors used it to refer to their general discomfort with the patient role, while others were expressing more specific concerns, for example, that their treating doctors might think they were over-reacting to a trivial illness.11 The ‘sense of duty’ felt by the doctors was considered an internalised expression of this embarrassment.11 Many doctors described concern (embarrassment) that they should not impose upon another doctor's time, especially if the illness was a trivial one.10,13 Shadbolt reported that 70% could only justify seeing a doctor if they were really ill.24 Up to one-third of doctors stated they had a condition that they felt they should consult a doctor for, but had not done so because they were a doctor themselves.9,13,24 Other doctors worried that their own diagnosis or treatment would be found to be wrong if they sought medical help.13 Embarrassment was reported to be more prominent for mental health problems.12,13,25 Doctors were also more reluctant to seek help for ‘less-defined’ illnesses, such as stress, sexual difficultly, and alcohol dependence.9,10,12,13,17 Some of the barriers identified were derived from doctors responding to hypothetical vignettes rather than actual experiences.9,10,16,17,25
Lack of time was commonly reported as a significant barrier (Table 2). Cost was a more complex barrier, involving both direct and indirect costs, including fees charged, time off, and access to disability and business insurance cover. Cost remained an important barrier to healthcare access, even though physicians may be in the upper socioeconomic group.
Personality issues were independently related to accessing health care.19,26 Considering three health locus of control domains, a physician who has a strong belief that a physician can help is more likely to have a GP. A physician who has a high chance locus of control is less likely to have a GP, and a physician who has a strong belief that health is under their control is also less likely to have a GP.19 The specialty practice of the doctor may also affect health access. Gross et al found that paediatricians and psychiatrists were more likely to have a GP, compared to surgeons and pathologists.19
Having special knowledge in the health field affects doctors' health access. This knowledge enhances doctors' understanding of the potential implications of symptoms and the ways in which illness might have an impact upon different aspects of practice, including qualifying for insurance and remaining registered.11,13 Doctors are also aware of how an illness and its treatment is likely to have an impact on their work (for both their colleagues and their patients),11,14,23 even before the illness is diagnosed. Medical knowledge enables doctors to justify their symptoms as ‘insignificant’. When doctors' reluctance to seek health care for minor illness is coupled with this tendency to rationalise symptoms as minor (trivial) conditions, the potential seriousness of this barrier is recognised.11–13,22
Medical knowledge also makes doctors more aware of the limitations of medical care. Awareness of the history of medicine, together with the recent emphasis on evidence-based medicine within the medical literature, highlights the lack of evidence for many diagnostic and treatment procedures.27 This encourages scepticism in the doctor–patient, and sometimes this scepticism may be expressed as a lack of compliance,18 just as it is in the general community.28
The provider determines how confidential health information will remain, and this was a key issue. Confidentiality can also be considered an umbrella term reflecting different fears regarding how personal details will be handled in the medical setting. Doctors may be very aware of the limitations to confidentiality within a system where obligatory legal requirements for reporting and the closeness of social networks can make confidentiality difficult to assure.8 Having a professional partner as a GP further complicates this issue.13 Anxiety about confidentiality is greater for those with mental health problems.10 The provider also determines the quality of care provided, and it is concerning to see how frequently the personal experience of poor medical care was reported.11,20,22 This is consistent with the many personal accounts of doctors seeking health care.29,30 A negative experience may reduce the future health access for the doctor–patient.
Specific structural problems within the healthcare system were identified. The long hours of duty and difficulty accessing locums simply made it physically difficult for the doctor to access care.11–13,24 Doctors also lacked training in how to access appropriate self-care and how to treat their peers (in both formal and informal settings). Qualitative data showed that the doctor–patient expected to be treated like a‘normal’ patient, yet the treating doctor often failed to satisfy this expectation.11,22,31,32
Underlying many of the aspects of doctors' health-seeking behaviours identified was the culture of medicine and medical socialisation. For example, doctors commonly faced intense pressure to be healthy or to control their own illness. This pressure came from both medical colleagues and the community (Table 2). Self-treatment was actively encouraged by peers,10,13,20 and it is not culturally appropriate to acknowledge illness in a peer.11,13,14 Rosvold and Bjertness found that one-quarter of doctors hide their illness from colleagues.12 Corridor consultations were accepted practice, and fostered self-treatment.12,24
This systematic review describes the current knowledge of doctors' health-access behaviours and the barriers they experience accessing health care. Firstly, this review demonstrates that very little information is available, despite the importance of this issue for doctors' health. Secondly, collation of these data enables comparison between doctors' health access and that of the general community.
It should be noted that despite the different styles of studies (descriptive and qualitative), the different populations studied, and the different countries where the data were collected, there was a high level of concordance in the health-access experiences described by doctors. The diversity of the study populations (different countries, specialties, age groups, and sex) also improves the generalisability of the data.
A number of the articles reviewed are relatively old. Ten papers are over a decade old. Many of these articles are still regularly cited and continue to inform the current debate on doctors' health. The two most recent studies in this review report similar findings to the earlier studies.21,33 Uallachain surveyed younger doctors, who may have been exposed to more education about self-care in their training, yet the results are similar.21
The quality of the data does limit the conclusions that can be made. Imprecise definitions within the studies affect the interpretation of the results. Selective reporting of comments from the questionnaires may have caused potential reporting bias (Appendix 2). Care should also be taken when accepting hypothetical data, because doctors report how they should access health care differently from how they actually access health care.34
Having a GP enhances the opportunities for both doctors and the general community to access preventive health care.19,35 Professional colleges and registration boards recognise this advantage when they encourage doctors to have their own GP.36–38 However, this review reveals that health access is far more complex than just having a GP. Most doctors who have their own GP still self-treat and access informal health care. This is not surprising when we acknowledge that self-treatment and informal care have always been a normal part of the pathway to formal health care for all. This issue has been well described for the general community within the context of the lay referral pathway.39 Understanding doctors' health-access behaviours in this broader context makes it evident that simply encouraging doctors to have their own GP is not enough to improve healthcare access; it is only the first step.
The data show that the barriers that doctors describe are also similar to those experienced by the general community, especially once other social determinants of health (for example, financial, physical, and education barriers) are considered.40 Most people who do not have a GP simply do not believe they need one,35 and it is possible that since most doctors are satisfied with the care they receive,17 many doctors may not believe they need a GP. Lack of time and cost issues are common concerns for the general community as well as doctors (Table 2).
Embarrassment is a common concern for all patients. For doctors, this may be complicated by concern about imposing upon a colleague and the potential exposure of error in self-diagnoses and treatment. These are complex issues, and the use of umbrella terms such as ‘embarrassment’ in surveys makes interpretation of such data difficult. Being worried that the problem might be trivial is also a concern for patients,41 although the magnitude of this barrier for doctors may be greater. Confidentiality is another important issue for both the community and doctors, especially in relation to mental health issues.42 Most studies show that the barriers that doctors experience are more severe for those with mental health problems, just as they are for the general community.43
The qualitative data emphasise the importance of the system barriers that reduce health access for doctors. System barriers are very difficult for any individual doctor to conquer alone. They need to be addressed by the profession as a whole. Some structural barriers are currently being addressed effectively. For example, the European Working Time Directive is ensuring safe work hours throughout the European Union.44
Considering the system barriers, the acculturation process is especially important in doctors' health access. These cultural barriers are the most difficult for the individual doctor to address. Acculturation begins in medical school.45,46 Barriers are created when role models fail to normalise health access, and systems propagate the stigma of illness for doctors.47 Cultural change is needed. It is only when it becomes acceptable (culturally normal) for a doctor to seek health care for physical and mental health problems that health access will improve.
Educational programmes have been developed to improve doctors' health. These include training in self-efficacy,48 and peer group support.3 Continuing medical education has also been reported to be a protective factor for burnout.49 Many programmes focus on reducing the stress that doctors experience and on the individual doctors' response to their health. They are often regarded positively by the participants,50 although there are limitations.51 Certainly, there is little evidence to suggest that these programmes effectively address health-access barriers, and given the systemic nature of the barriers this is not surprising.
This review highlights a significant gap in current knowledge of an important aspect of doctors' health — the health-access behaviours of doctors. Further research targeting the understanding of doctors' health-access behaviours is encouraged. Presenting future research within a broader context will add an important perspective to the doctors' health literature.
|Author(s) and year||Country||Number studied||Population studied/controls||Study design||Responders, n||Age, years||Response rate, %|
|Allibone et al, 198123||UK||1444||Medical graduates 1935–1959||Postal questionnaire||932||52.2a||65|
|Glanz et al, 198252||US||636||Physicians at Temple University Health Sciences Centre 1980||Distributed questionnaire||296||28a||47|
|Kahn et al, 19888||US||221/416b||Doctors in university faculty/non-MD academics||Postal questionnaire||144/283b||n/a||65/68|
|Richards, 198913||UK||501||Family physicians contracted with Avon Family Practitioner Committee at 1 July 1987||Postal questionnairec||431||≤35: 27.4%; 36–50: 47.8%; ≥51: 23.7%||86|
|Rennert et al, 199018||Israel||366||Members of Israeli Association of Family Physicians||Postal questionnaire||243||47a||66|
|Bortz, 199253||US||152||Members of Palo Alto Medical Clinic||Survey||126||47.6a||83|
|Chambers and Belcher, 199215||UK||275||Family physicians attending a course or postgraduate lecture||Given questionnaire at meeting||247||42.2a||90|
|Pullen et al, 19959||Australia||2564||Registered doctors||Postal questionnaire||1125||n/a||44|
|Baldwin et al, 199716||UK||150||Junior doctors who graduated from a Scottish university||Structured interviec||142||33a||95|
|Clarke et al, 199820||Ireland||85||GPs, GP trainees, hospital consultants||Postal questionnaire||76||n/a||89|
|Wines et al, 199854||Australia||275||Doctors attending of Given urology meeting||Given questionnaire at meeting||205||n/a||75|
|Forsythe et al, 199925||UK||1733||Consultants and principals in general practice structured||Postal semi-structured questionnairec||1151||n/a||66|
|McCall et al, 199926||Australia||544||Family physicians||Postal questionnaire||318||<35: 21%; 35–54: 64%; >55: 15%||58|
|Richards, 199931||New Zealand||500||Registered doctors||Postal questionnaire||311||<30: 13.5%; 30–49: 67.5%; >50: 18.9%||62|
|Gross et al, 200019||US||1337||Medical graduates Postal 1948–1964||Postal questionnaire||915||61a||68|
|Rosen et al, 200032||US||389||Internal medicine residents (four training programmes)||Postal questionnaire||316||29a||83|
|Toyry et al, 200055||Finland||4477||Finnish Medical Association members||Postal questionnaire||3313||42.3a||74|
|Shadbolt, 200224||Australia||300||Junior house officers||Survey (issued) and focus groups||156||n/a||52|
|Rosvold and Bjertness, 200212||Norway||1476||Norwegian Medical Association members||Postal questionnaire||1015||42a||69|
|Campbell and Delva, 200317||Canada||215||Residents in Faculty of Medicine at Queen's University||Mailbox questionnaire||122||n/a||57|
|Davidson and Schattner, 200310||Australia||896||Health Insurance Commission register||Postal questionnairec||358||40–55: 55%||40|
|Uallachain, 200721||Ireland||112||National Association of GP trainees at annual general meeting 2003||Given questionnaire at meeting||100||<25: 8%; 26–30: 81%; 31–35: 10%; >36: 1%||89|
|Schneider et al, 200733||Switzerland||2756||Swiss primary care physician members of Federatio Medicorum Helveticorum||Postal questionnaire||1782||51a||65|
|Author(s) and year||Country||Number studied||Population studied||Study design||Responders, n||Age, years||Responders adding comments, na|
|Allibone et al, 198123||UK||1444||Medical graduates 1935–1959||Postal questionnaire||932||52.2b||561|
|Richards, 198913||UK||501||Family physicians contracted with Avon family Practitioner Committee at 1 July 1987||Postal questionnairec||431||≥35: 27.4%; 36–50: 47.8%; ≥51: 23.7%||n/a|
|Chambers and Belcher, 199215||UK||275||Family physicians attending a course or postgraduate lecture||Given questionnaire at meeting||247||42.2b||98|
|McKevitt and Morgan, 199711||UK||64||Doctors with an illness invited to participate in research||Interviews: face-to-face or telephoneb||64||27–65|
|Richards, 199931||New Zealand||500||Registered doctors||Postal questionnaire||311||<30: 13.5%; 30–49: 67.5%; >50: 18.9%||109|
|Rosen et al, 200032||US||389||Internal medicine residents(four training programmes)||Postal questionnaire||316||29b||59|
|Thompson et al, 200114||UK (Northern Ireland)||172||Family physicians||Focus groups and interviews||27||n/a|
|Shadbolt, 200224||Australia||300||Junior house officers||Survey (issued) and focus groups||156||n/a|
|Fromme et al, 200422||US||38||Physician-patients with cancer||Semi-structured interview||23||28–83||n/a|
Additional information can be found in the online version of this article
University of Queensland Research (Confirmation) Scholarship. Early support for this research paper was received from Primary Health Care Research Education and Development Fund, a programme of the Australian Government Department of Health.
Contribute and read comments about this article on the Discussion Forum: http://www.rcgp.org.uk/bjgp-discuss