Connect
MJA
MJA

A comparison of general practice encounters with patients from English-speaking and non-English-speaking backgrounds

Stephanie A Knox and Helena Britt
Med J Aust 2002; 177 (2): 98-101. || doi: 10.5694/j.1326-5377.2002.tb04681.x
Published online: 15 July 2002

Abstract

Objective: To determine whether doctor–patient encounters in general practice with patients from a non-English-speaking background (NESB) differ from encounters with patients of English-speaking background (ESB) in terms of the type of practice where the encounters occur and the type of problems managed.

Design and setting: A national cross-sectional survey of GP–patient encounters from a sample of all active registered GPs in Australia.

Participants: A random sample of 1047 GPs recruited in the 12 months from April 1999 to March 2000, each providing details of 100 consecutive patient encounters.

Main outcome measures: GP demographics, practice characteristics, patient demographics (including whether the patient mainly spoke a language other than English at home), and problems managed at the encounter.

Results: After adjusting for significant predictors, encounters with NESB patients were significantly more likely to occur at solo practices than practices of five or more GPs (odds ratio [OR], 2.15; 95% CI, 1.49–3.09), in metropolitan practices (OR, 6.34; 95% CI, 4.04–9.96), and with GPs who mostly consulted in a language other than English (OR, 5.44; 95% CI, 3.78–7.83). NESB encounters were relatively more likely to involve a respiratory problem (OR, 1.14; 95% CI, 1.04–1.26), endocrine/metabolic problem (OR, 1.41; 95% CI, 1.22–1.63) or digestive problem (OR, 1.14; 95% CI, 1.02–1.27), and relatively less likely to involve a psychological problem (OR, 0.73; 95% CI, 0.61–0.88) or social problem (OR, 0.67; 95% CI, 0.49–0.92).

Conclusion: Differences in morbidity management rates between encounters with NESB patients and ESB patients may reflect both differences in underlying prevalences of some disorders in the population of general practice patients, as well as different reasons among the two groups for attending general practice.

People from culturally and linguistically diverse backgrounds form an integral part of Australia's social fabric. Data from the 1996 census showed that 27% of Australian residents were born overseas and 15% were born in a non-English-speaking country. Among adults of working age, the proportion of those born in non-English-speaking countries was 22%.1

Understanding the utilisation of primary care, in particular general practice, is considered important in evaluating further demands for health services.2 There is ample evidence from Australian and overseas studies of health differences between ethnic groups, although the picture is complex.3-5 Issues of culture and language may constitute barriers to obtaining adequate healthcare,6 and issues of ethnicity and health are nearly always confounded by socioeconomic factors.3,7 Given these complexities, it cannot be assumed that people from culturally and linguistically diverse backgrounds consult general practitioners for the same purposes as the Australian community as a whole.

Most studies of ethnic differences in primary care have focused on one particular health aspect4,8 — very few have looked at group differences in overall use of healthcare services. There is some limited evidence from the United Kingdom of ethnic differences in consultation rates and the nature of problems managed in general practice.9 However, there have been no recent large-scale studies investigating the overall morbidity profile of culturally and linguistically diverse patients consulting GPs in Australia.

We wished to determine whether encounters in general practice with patients from a non-English-speaking background (NESB) differ from encounters with patients from an English-speaking background (ESB) in terms of the characteristics of the patients, the practices where the encounters took place and the problems managed at the encounters.

Methods
Data collection

Our study was based on data from the Bettering the Evaluation and Care of Health (BEACH) program, a national study of Australian general practice. The method used in the BEACH study has been described in detail elsewhere.10 In brief, BEACH is a continuous cross-sectional survey of general practice activity in Australia that commenced in April 1998. A random sample of about 1000 Australian GPs is recruited each year. The sampling is done in a "rolling" manner, designed so that a GP has one chance in three years of being recruited into the study. The unit of measure is the patient encounter with the GP. The sample of encounters is a cluster sample with the GP as the primary sampling unit, each GP providing records of 100 consecutive patient encounters. Since a patient may have more than one encounter with a GP, encounter rates are an indication of both the frequency that patients consult a GP for a problem and the prevalence of the problem in the general practice patient population.

The data presented here are for the 12-month period from April 1999 to March 2000.

Outcome measures

Data were collected on GP and practice characteristics, including GP age and sex, whether the GP conducted most consultations in a language other than English, and the number of GPs in the practice. The location of the practice was classified by postcode as either rural/remote or urban/metropolitan.11 Patient demographic factors recorded in the study included patient age and sex, whether the patient held a healthcare concession card and whether the patient was new to the practice. Problems managed at the encounter were classified according to the International classification of primary care (ICPC-2).12 Morbidity was analysed at both the specific problem level and the broader ICPC-2 chapter-based body-system level.

Statistical analysis

Unadjusted differences between NESB and ESB patients were analysed using cross-tabulations. Multiple logistic regression was used to test for differences, after adjusting for age and sex. Stata 7.0 software13 was used to correct for the design effect of the cluster sample. For cross-tabulations the P values for the Pearson χ2 statistic, corrected for the design effect, are reported. For logistic regression, 95% CIs are reported, based on standard errors calculated using the robust variance estimator method.13

Multiple logistic regression was used to explore the significant differences in problem management rates for NESB versus ESB encounters, after adjusting for all other significant explanatory variables. The model was reduced using backward elimination, with variables entered in related groups ("families"): GP/practice characteristics, patient characteristics, and morbidity. Variables were reduced in family order, starting with morbidity, and individual variables were retained or removed based on the P value of the Wald statistic, adjusted for the cluster sample (α, 0.05).

Results

There were 104 700 GP–patient encounter records from 1047 GPs, of which 7372 encounters (7.0%) were with NESB patients.

Morbidity (prior to multiple logistic regression)

After adjusting for age and sex, we found that encounters with NESB patients more frequently involved respiratory, metabolic/endocrine, skin, digestive or general/unspecified disorders, and less frequently involved psychological problems, than encounters with ESB patients (Box 2).

Pregnancy, and problems involving the neurological, eye, urinary, blood and male genital systems were each managed at less than 5% of encounters, with no significant differences between language background groups.

The specific problems most frequently managed in general practice were similar for both NESB and ESB encounters (Box 3). However, hypertension, acute upper respiratory tract infections, diabetes and lipid disorder were all managed significantly more frequently at NESB encounters, after adjusting for age and sex. Depression was significantly less likely to be managed at NESB encounters.

Multiple logistic regression analysis

Adjusted odds ratios for patient, practice and morbidity variables after multiple logistic regression analysis are summarised in Box 4.

After adjusting for other significant variables, patient age, healthcare card status, practice size, practice location and the GP speaking a language other than English remained as independent predictors of NESB encounters.

After adjusting for significant patient and practice characteristics, NESB encounters were significantly more likely to involve the management of a metabolic/endocrine problem and significantly less likely to involve a psychological or social problem.

Discussion

In metropolitan practices, NESB patients consulted GPs in a diverse range of general practice settings. GPs who operated solo practices and those who consulted in a language other than English were more likely to have encounters with NESB patients. This confirms that bilingual GPs have an important role in providing healthcare to many NESB patients.6

Our results support previous findings of different morbidity patterns managed in general practice for patients from different language backgrounds.9 Because statistics for encounters do not distinguish between the number of visits by separate patients and the number of return visits by an individual patient, morbidity rates can not be related directly to underlying differences in the prevalence of certain medical problems among NESB and ESB groups. It is also possible that apparent differences in morbidity between the two groups are more a reflection of differing beliefs about what type of problems are appropriate to discuss with a GP.

The higher rates of diabetes management in NESB encounters may reflect the relatively higher population prevalence of self-reported diabetes among Australians who do not speak English at home.14 However, differences in community prevalence do not readily explain the higher rates of hypertension problems discussed in NESB encounters, as self-reported cardiovascular conditions such as hypertension are no higher among NESB than ESB people in the general population.15

Whether mental disorders are any more or less prevalent among the NESB population than the rest of the community is uncertain, as studies have produced conflicting results.2,6,16 It is unclear from our study whether lower rates of psychological problems managed at NESB encounters reflect lower prevalence of these disorders in NESB groups, different beliefs about the appropriateness of raising such issues with GPs, GPs' skill in detecting psychological problems, or other factors. Previous research has suggested that GPs find it more difficult to detect psychological problems in their NESB patients8 — whether this is because NESB patients more often express psychological distress in somatic rather than psychological terms is a matter for debate.8,17

Although census data indicate that nearly 15% of Australians were born in a non-English-speaking country,1 only 7% of encounters with GPs in our study were identified as being with NESB patients. This discrepancy may be partly a matter of definition: in our study, we defined NESB in terms of the primary language spoken at home, while in the census it was defined according to country of birth, so the two categories are not exactly equivalent. Nevertheless, the number of encounters with NESB patients may have been under-reported in our study, as GPs may have omitted to label certain patients as "NESB" if they spoke English well. There has been some variability over time in the proportion of encounters with NESB patients in the BEACH study, possibly due to GP recording practices, as well as to some small changes to the recording form each year.18

Our findings relate to the population of general practice patients. Any extrapolation to the broader Australian population of NESB people should be made with caution. Furthermore, the category "NESB" includes a diverse range of cultures and ethnic identities, and there are limitations in applying the findings from this study to any specific cultural, language or ethnic group. Studies of specific NESB communities are required to explore the extent that the broad differences identified in this study apply to individual groups.

  • Stephanie A Knox1
  • Helena Britt2

  • General Practice Statistics and Classification Unit, University of Sydney, Wentworthville, NSW.


Correspondence: sknox@med.usyd.edu.au

Acknowledgements: 

The Commonwealth Department of Health and Aged Care, the Commonwealth Department of Veterans' Affairs, the National Occupational Health and Safety Commission, AstraZeneca Pty Ltd (Australia), Aventis Pharma Pty Ltd and Roche Products Pty Ltd contributed financially to the conduct of the BEACH study. We would like to acknowledge the contribution of the GP participants and all BEACH team members, past and present. The General Practice Statistics and Classification Unit is a collaborating unit of the Australian Institute of Health and Welfare.

Competing interests:

None identified.

  • 1. Commonwealth Department of Health and Aged Care. Health Wiz: national social health statistical database. Canberra: Prometheus, 2000.
  • 2. Stuart GW, Klimidis S, Minas IH. The treated prevalence of mental disorder amongst immigrants and the Australian-born: community and primary-care rates. Int J Soc Psychiatry 1998; 44: 22-34.
  • 3. Ahmad WI, Kernohan EE, Baker MR. Influence of ethnicity and unemployment on the perceived health of a sample of general practice attenders. Community Med 1989; 11: 148-156.
  • 4. Marks GB, Bai J, Simpson SE, et al. The incidence of tuberculosis in a cohort of south-east Asian refugees arriving in Australia 1984–94. Respirology 2001; 6: 71-74.
  • 5. de Looper M, Bhatia K. Australian Health Trends 2000. Canberra: Australian Institute of Health and Welfare, 2001. (AIHW Catalogue No. PHE 24.)
  • 6. Stuart GW, Minas IH, Klimidis S, O'Connell S. English language ability and mental health service utilisation: a census. Aust N Z J Psychiatry 1996; 30: 270-277.
  • 7. Chandola T. Ethnic and class differences in health in relation to British south Asians: using the new National Statistics Socio-Economic Classification. Soc Sci Med 2001; 52: 1285-1296.
  • 8. Comino EJ, Silove D, Manicavasagar V, et al. Agreement in symptoms of anxiety and depression between patients and GPs: the influence of ethnicity. Fam Pract 2001; 18: 71-77.
  • 9. Gillam SJ, Jarman B, White P, Law R. Ethnic differences in consultation rates in urban general practice. BMJ 1989; 299: 953-957.
  • 10. Britt H, Miller GC, Charles J, et al. General practice activity in Australia 1999–2000. Canberra: Australian Institute of Health and Welfare, 2000. General Practice Series. (AIHW Catalogue No. GEP 5.)
  • 11. Rural, Remote and Metropolitan Area (RRMA) classification. Canberra: Department of Primary Industries and Energy and Department of Human Services and Health, 1994.
  • 12. Classification Committee of the World Organisation of Family Doctors (WONCA). International classification of primary care. 2nd edition. (ICPC-2). Oxford: Oxford University Press, 1998.
  • 13. Stata statistical software. Release 7. College Station, TX: Stata Corporation, 2001.
  • 14. National health survey 1995: diabetes. Canberra: Australian Bureau of Statistics, 1997. (ABS Catalogue No. 4371.0.)
  • 15. National health survey 1995: cardiovascular and related conditions. Canberra, Australian Bureau of Statistics, 1997. (ABS Catalogue No. 4372.0.)
  • 16. Mental Health and Well-being: profile of adults, Australia, 1997. Canberra: Australian Bureau of Statistics, 1998. (ABS Catalogue No. 4326.0.)
  • 17. Gureje O, Simon GE, Ustun TB, Goldberg DP. Somatization in cross-cultural perspective: a World Health Organization study in primary care. Am J Psychiatry 1997; 154: 989-995.
  • 18. Britt H, Sayer GP, Miller GC, et al. General practice activity in Australia 1989-99. Canberra: Australian Institute of Health and Welfare, 1999. General Practice Series. (AIHW Catalogue No. GEP 2.)

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

Online responses are no longer available. Please refer to our instructions for authors page for more information.