Connect
MJA
MJA

Doctors’ health-seeking behaviour: a questionnaire survey

Sandra K Davidson and Peter L Schattner
Med J Aust 2003; 179 (6): 302-305.
Published online: 15 September 2003

Abstract

Objectives: To explore doctors’ perceptions of the acceptable limits to self-treatment and to identify barriers to doctors seeking appropriate healthcare.

Design: Self-completion, postal survey using three hypothetical case vignettes.

Setting and participants: 896 Australian doctors randomly selected from the Health Insurance Commission database and stratified by sex, discipline (general practitioner or specialist) and location (urban or rural). Data were collected between May and July 2001.

Main outcome measures: Doctors’ self-reported attitudes on illness behaviour and choice of medical care in response to case vignettes.

Results: 358 (40%) doctors returned questionnaires. More participants believed it was acceptable to self-treat acute conditions (315/351; 90%) than to self-treat chronic conditions (88/350; 25%). Nine per cent (30/351) of participants believed it was acceptable to self-prescribe psychotropic medication. A greater proportion of GPs (206/230; 90%) than specialists (101/121; 83%) believed doctors are reluctant to attend another doctor, especially if the problem is psychological. Women and GPs were significantly less likely to report that it was easy to find a satisfactory treating doctor (women, 58/140 [41%]; men, 128/211 [61%]; GPs, 106/231 [46%]; specialists, 80/120 [67%]). Being a specialist was predictive of seeking appropriate healthcare for all three vignettes.

Conclusion: Doctors have varying opinions regarding the acceptability of self-treating chronic conditions, and perceive considerable barriers to seeking appropriate medical care. Strategies are needed to challenge the culture of self-reliance.

Many studies show that, when doctors experience ill health, they disregard the advice they give their patients.1,2 The medical community has developed a culture in which working through illness and self-treating is the norm.3,4

Although doctors have lower overall rates of mortality than the general population, they are at higher risk of certain physical and psychological problems.5,6 The impact of these illnesses may be exacerbated by a culture that discourages admission of health vulnerabilities.7 Personal testimonies of doctors who have experienced major illness suggest that the “pull yourself together and just get on with it” mentality is a significant barrier to seeking appropriate treatment.8 A New South Wales study found that 26% of doctors experienced a condition warranting medical consultation, but felt inhibited about consulting a doctor.9 When doctors do seek external care, evidence suggests they receive a lesser quality of care than lay patients.7,8

Self-treatment for doctors includes diagnosing and treating one’s own illness and prescribing for oneself. It also includes undertaking informal, or “corridor”, consultations and self-referring to a specialist. Self-treatment is inappropriate because of its lack of objectivity.10

We aimed to describe doctors’ perceptions of the acceptable limits to self-treatment, and to determine whether selected attitudinal or demographic variables predict doctors’ decisions to self-treat in three hypothetical situations.

Methods

A stratified, random sample of doctors was drawn from the Health Insurance Commission (HIC) database using their modified synchronised sampling system. General practitioners who claimed more than 375 consultations and specialists who claimed more than 200 consultations in the March quarter of 2001 were eligible for selection. From a population of 26 765, a sample of 896 doctors was randomly selected. The sample was stratified by sex, urban or rural practice, and GP or specialist. Data were collected from May to July 2001.

We developed a 31-item questionnaire. Twelve items asked participants about their attitudes to self-treatment, and four items asked about their own experience of formal healthcare. Responses were provided on a five-point Likert scale. There were 12 demographic questions, including whether respondents had their own GP. The questionnaire ended with three hypothetical case vignettes describing a scenario in which participants experience anxiety, hypertension, or severe asthma. The dependent variable for each vignette was the doctor’s decision to self-treat or to seek appropriate care.

Four elements identified in a preceding qualitative study and a literature review as being potential barriers to seeking formal healthcare were embedded as independent variables in each vignette. The variables were concerns over confidentiality, heavy workload, embarrassment, and confidence in treating the condition. Each variable was included in the vignette either as a concern for the participant, or as not a concern (Box 1). To evaluate the effect of these variables on the decision to seek appropriate care, each variable was fully crossed with the other three variables at both levels, producing 48 possible combinations. A fractional factorial design11 was used to reduce this to 18 combinations: six each for anxiety, hypertension and asthma, in an even distribution. Six versions of the questionnaire were produced, each consisting of one version of the anxiety, hypertension and asthma vignettes.

Versions 1 and 2 of the questionnaire were each sent to 150 participants and versions 3 to 6 were each sent to 149 participants. Questionnaires were anonymous but coded so nonresponders could be followed up. Administrative staff recorded receipt of questionnaires and then gave them to the researchers. A second mail-out was conducted after three weeks. Two weeks later, a random sample of 200 nonresponders was contacted by phone and encouraged to complete and return the questionnaire.

Data analysis

Frequency analysis was conducted on the attitudinal data, and independent samples t tests were used to identify differences between GPs and specialists and between men and women.

Logistic regression analysis was performed on the vignette data to determine whether any of the four independent variables or selected demographic variables predicted respondents’ decisions to seek appropriate care. The dependent variable for the logistic regression analysis was the treatment choice (ie, appropriate or inappropriate choice of treatment). Inappropriate treatment was defined as self-treatment, corridor consultation, direct referral to a specialist, or no treatment. Appropriate treatment was defined as consulting a GP. The following 11 variables were entered into the equation: confidentiality, workload, embarrassment, confidence in treating the condition, having one’s own GP, receiving specialist treatment, having attended a GP in the past two years, GP or specialist registration (discipline), sex, age, and practice location.

Ethical approval and consent

This study was approved by the Monash University Standing Committee on Ethics in Research Involving Humans.

Results
Demographic profile

Of the 896 questionnaires, 358 were returned (response rate, 40%). Five questionnaires had a substantial amount of missing data and analysis was generally conducted on between 351 and 353 respondents. Urban male GPs were slightly underrepresented and rural female GPs were overrepresented (Box 2). Fifty-five per cent of respondents were aged 40–55 years. Forty-four per cent of respondents indicated they had long-term health problems, which is much lower than the 87% prevalence of long-term conditions reported in the community.12

Attitudes

Box 3 shows attitudes regarding self-treatment, barriers to seeking formal care, and respondents’ experiences of medical care. Most considered self-treatment of acute minor illnesses acceptable, but there was more disagreement as to the acceptability of self-treating chronic conditions and ordering blood tests on oneself. Most respondents (especially GPs) felt that doctors often feel embarrassed to attend another doctor, especially for psychological problems.

Predictors of appropriate treatment

Being a specialist was a highly significant predictor of choosing appropriate treatment (Box 4). Having one’s own GP was a significant predictor of appropriate treatment choice in the hypertension and asthma vignettes only. Concern over confidentiality was a significant predictor of not choosing appropriate treatment in the anxiety vignette alone. Workload, embarrassment and confidence in treating the condition were not predictors of treatment choice in any of the vignettes.

Discussion

The response rate for our study is rather low, so caution is needed in generalising the results. However, it is comparable to other Australian surveys investigating doctors’ own medical care.9,13 It has been suggested that the personal nature of the topic contributes to low response rates.9

Our study described doctors’ opinions and attitudes, rather than their actual behaviour in managing their medical conditions. Given that recommendations against self-treatment are well known in the profession, responses may have been influenced by social desirability factors, and it is possible that more respondents would actually support self-treatment, including the self-prescribing of psychotropic medicines, in real situations.

Previous studies have found differences between GPs and specialists in their patterns of health-seeking behaviour,1,9 but this is the first study to show that GPs are significantly less likely than specialists to seek appropriate treatment across several situations. The attitudinal data suggest that GPs’ preferences for self-treatment may be associated with their perception that doctors are embarrassed to attend a colleague and that it is difficult to find a good treating doctor.

Our results support previous assertions that the culture of medicine is a barrier to doctors seeking healthcare.1,3,7 The perception that doctors are more likely than other professionals to work through illness may be symptomatic of a culture in which an image of invincibility is encouraged and vulnerability is denied.3 The direct result of this phenomenon is likely to be delayed presentations even for serious conditions.

Fifty-five per cent of respondents had their own GP. This is higher than the 42% reported by Pullen,9 but is considerably less than the New South Wales Doctors Mental Health Policy recommendation that all doctors have their own GP.14 Having one’s own GP was a significant predictor of seeking appropriate care in the hypertension and asthma vignettes, but not in the anxiety vignette. The reason for this discrepancy is unclear, but the emergence of confidentiality as a significant predictor in the anxiety vignette suggests that issues of trust are involved in the more sensitive psychological conditions. This is consistent with a previous study,3 in which concerns about confidentiality were a barrier to seeking psychiatric services. Similarly, confidentiality breaches, combined with professional attitudes discouraging admission of vulnerabilities, have been reported to influence doctors’ reluctance to seek mental healthcare.7

One of the most disturbing findings was that, in spite of professional recommendations to the contrary,14,15 a quarter of respondents thought it was acceptable to self-treat chronic conditions and an even higher proportion thought it was acceptable to order blood tests on oneself to monitor a chronic condition. Other studies have examined self-treatment in relation to specific medications or symptoms, and have not grouped self-treatment into acute conditions, chronic conditions, diagnostic procedures and monitoring categories as we did. Therefore, although it is not possible to directly compare across studies, our results are consistent with a previous finding that doctors would self-treat for seven out of 10 sets of acute and chronic symptoms for which they would recommend other doctors seek care.1 The finding of little consensus regarding acceptability of self-treatment is also consistent with the earlier study.1 Although most respondents were satisfied with the care they had received from other doctors, many thought it was difficult to find a doctor with whom they were comfortable. Reasons for this may include embarrassment or concern about confidentiality for sensitive health issues. This highlights the need for doctors to be appropriately trained for consultations in which the patient is also a doctor.15

Our results indicate that there are significant barriers to doctors seeking formal medical care, and that this may be particularly true for GPs. It is important that this be addressed in medical school, as such attitudes are already in place among many junior doctors.13 Strategies aimed at doctors’ self-treatment should focus on challenging the culture of doctors’ self-reliance.

1: Examples of vignettes used in the questionnaire

Hypertension

You realise that your BP has been elevated on three separate occasions recently. Your overall health is good. Your BMI is 26, you’re a non-smoker, do minimal exercise and have one or two glasses of wine a day. Your total cholesterol is slightly elevated. Although you may be confident managing this condition in your patients you are hesitant to manage it yourself. There is a competent doctor you know and trust to maintain your confidentiality but you are worried about what the doctor you could consult would think of you. Your workload is very manageable and you can take time off if you need to.

In this vignette, embarrassment and confidence in treating the condition are concerns and confidentiality and workload are not.

Anxiety

Over the last 6 months you’ve noticed that you have become increasingly anxious. You find yourself constantly worrying about everyday events that never used to bother you. You are having difficulty sleeping and often wake up during the night. Having a few drinks seems to help you relax. Although your general health is quite good you are experiencing a lot of muscle tension and headaches. You’re also confident in managing this condition with your patients and feel that you could manage it for yourself. You know a competent doctor you could see whom you’re comfortable would be understanding of you, but you are unsure whether you can trust them to maintain your confidentiality. Your practice is so busy it would be hard to find the time.

In this vignette, workload and confidentiality are concerns, while embarrassment and confidence in treating the condition in oneself are not.

2: Representativeness of 353 respondents compared with Health Insurance Commission (HIC) sample of 896 doctors

Urban


Rural


HIC

This study

HIC

This study


General practitioners

Women

136 (15%)

55 (15%)

64 (7%)

49 (14%)

Men

280 (31%)

77 (22%)

104 (12%)

50 (14%)

Specialist

Women

48 (5%)

24 (7%)

16 (2%)

12 (3%)

Men

184 (21%)

59 (16%)

64 (7%)

27 (8%)

Total

648 (72%)

215 (61%)

248 (28%)

138 (39%)

3: Participants who “agree” or “strongly agree” to attitudes on seeking healthcare: comparisons between disciplines and sexes

Total (n = 350–352)*

GPs (n = 229–231)*

Specialists (n = 120–121)*

P

Men (n = 210–212)*

Women (n = 139–140)*

P


It is acceptable to:

Self-treat acute, minor illness

315 (90%)

205 (89%)

110 (91%)

0.175

193 (91%)

122 (87%)

0.517

Self-treat chronic conditions

88 (25%)

63 (28%)

28 (21%)

0.139

57 (27%)

31 (22%)

0.770

Order blood test on self for diagnostic purpose

179 (51%)

112 (49%)

67 (55%)

0.192

112 (53%)

67 (48%)

0.201

Order blood test on self for monitoring purpose

162 (46%)

112 (49%)

50 (42%)

0.177

102 (49%)

60 (43%)

0.166

Self-prescribe psychotropic medication

30 (9%)

22 (10%)

8 (7%)

0.136

18 (9%)

12 (9%)

0.995

Opinions on doctors’ behaviour in seeking healthcare

Embarrassed to attend another doctor

250 (71%)

174 (76%)

76 (63%)

0.003

146 (69%)

104 (74%)

0.074

Reluctant to consult for psychiatric problems

307 (87%)

206 (90%)

101 (83%)

0.027

177 (84%)

130 (93%)

0.015

Doctors make few errors when treating self

24 (7%)

19 (8%)

5 (4%)

0.051

15 (7%)

9 (6%)

0.181

Doctors’ symptoms are under-investigated

88 (25%)

54 (23%)

34 (28%)

0.864

59 (28%)

29 (21%)

0.224

Doctors down-play their symptoms

251 (72%)

166 (72%)

85 (70%)

0.370

149 (71%)

102 (73%)

0.074

Doctors present late

270 (77%)

180 (78%)

90 (74%)

0.568

157 (74%)

113 (81%)

0.067

Doctors are more likely to work when sick

332 (95%)

219 (95%)

113 (93%)

0.259

197 (93%)

135 (96%)

0.006

Attitudes regarding own healthcare

Embarrassed if problem turned out to be minor

95 (27%)

72 (31%)

23 (19%)

0.003

48 (23%)

47 (34%)

0.016

Easy to find treating doctor

186 (53%)

106 (46%)

80 (67%)

< 0.001

128 (61%)

58 (41%)

0.000

Satisfied with own healthcare

297 (85%)

191 (83%)

106 (88%)

0.417

185 (88%)

112 (81%)

0.048

Concerned about confidentiality

68 (19%)

45 (20%)

27 (21%)

0.716

34 (16%)

34 (24%)

0.054


* Denominators vary because of missing data.

4: Significant factors predicting appropriate treatment choice in case vignettes (logistic regression analysis)

Odds ratio

95% CI


Anxiety

Is a GP

0.38

0.22–0.68

Attended GP in last 2 years

2.56

1.31–5.00

Concerned about confidentiality

1.87

1.01–3.45

Hypertension

Is a GP

0.38

0.19–0.74

Has own GP

3.57

1.76–7.23

Asthma

Is a GP

0.42

0.25–0.72

Has own GP

2.84

1.55–5.21

Male

2.33

1.38–3.93

  • Sandra K Davidson1
  • Peter L Schattner2

  • Department of General Practice, Monash University, East Bentleigh, VIC.

Correspondence: 

Acknowledgements: 

We thank Dr Tony Rogers who had the original idea for the study and Dr Remi Guibert who provided advice on the methods section of the article.

Competing interests:

None identified.

  • 1. Chambers RM. What should doctors do if they become sick? Fam Pract 1993; 10: 416-423.
  • 2. Rennert M, Hagoel L, Epstein L, Shifroni G. The care of family physicians and their families: a study of health and help-seeking behaviour. Fam Pract 1990; 7: 96-99.
  • 3. Thompson WT, Cupples ME, Sibbett CH, et al. Challenge of culture, conscience, and contract to general practitioners’ care of their own health: qualitative study. BMJ 2001; 323: 728-731.
  • 4. McKevitt C, Morgan M, Dundas R, Holland WW. Sickness absence and ‘working through’ illness: a comparison of two professional groups. J Public Health Med 1997; 19: 295-300.
  • 5. Schlicht SM, Gordon IR, Ball JRB, Christie DGS. Suicide and related deaths in Victorian doctors. Med J Aust 1990; 153: 518-521.
  • 6. Serry N, Bloch S, Ball R, Anderson K. Drug and alcohol abuse by doctors. Med J Aust 1994; 160: 402-407.
  • 7. Center C, Davis M, Detre T, et al. Confronting depression and suicide in physicians. JAMA 2003; 289: 3161-3166.
  • 8. Armstrong E. Rehabilitating troubled doctors. BMJ 1997; 314: 2-3.
  • 9. Pullen D, Lonie CE, Lyle DM, et al. Medical care of doctors. Med J Aust 1995; 162: 481-484.
  • 10. Richards JG. The health and health practices of doctors and their families. N Z Med J 1999; 112: 96-99.
  • 11. Kirk RE. Experimental design: procedures for the behavioral sciences. 3rd ed. Pacific Grove, CA: Brooks/Cole, 1995.
  • 12. Australian Bureau of Statistics. National health survey: summary of results 2001. Canberra: ABS, 2001. (Cat. No. 4364.0.)
  • 13. Shadbolt N. Attitudes to healthcare and self-care among junior medical officers: a preliminary report. Med J Aust 2002; 177: S19-S20. <MJA full text>
  • 14. NSW Doctors’ Mental Health Implementation Committee. Doctors’ mental health policy. Sydney: NSW Health Department and NSW Branch, Australian Medical Association, 1997.
  • 15. Buchanan J. Towards better healthcare for doctors. Options and strategies to improve policy development for doctors’ healthcare. AMA (Federal) and AMA Queensland, 2001. Available at: www.ama.com.au/web.nsf/doc/WEEN-5NE9JX/$file/DrHealthOptions.pdf (accessed Jul 2003).

Author

remove_circle_outline Delete Author
add_circle_outline Add Author

Comment
Do you have any competing interests to declare? *

I/we agree to assign copyright to the Medical Journal of Australia and agree to the Conditions of publication *
I/we agree to the Terms of use of the Medical Journal of Australia *
Email me when people comment on this article

Responses are now closed for this article.