Differences in antibiotic treatment and utilization of diagnostic tests in Dutch primary care between natives and non-western immigrants

Abstract

Background: Different expectations and demands of non-western immigrants may prompt physicians to deviate from their routine clinical management strategies resulting in more diagnostic tests and higher prescription rates. Antibiotic prescription is of specific importance, since overuse of antibiotics is a major public health concern. No data are available about possible differences in antibiotic prescription rates between natives and non-western immigrants in Europe. Objectives: To determine whether ethnic origin is an independent determinant for the frequency of antibiotic treatment and additional diagnostic tests in primary care patients with infectious diseases. Methods: In this cross-sectional study, 1015 non-western immigrants were compared to 995 native Dutch, all selected from the Utrecht Health Project database. Data were analysed using multiple logistic regression analyses. Antibiotics most frequently used in Dutch primary care were included in the analyses. For the analyses of additional diagnostic tests, only respiratory tract infections were included, since these are the most common infectious diseases for which general practitioners are contacted. Results: First generation non-western immigrants were more likely to undergo diagnostic tests and received more antibiotics than native Dutch (respectively OR: 2.08; 95% CI: 1.64–2.63; and OR: 1.31; 95% CI: 1.04–1.65). No differences were found for the second generation immigrants. The results found were independent of potential confounders. Conclusions: The highest rates of antibiotic prescription and diagnostic tests concern first generation non-western immigrants. In order to promote a rational and restricted use of antibiotics in outpatients first generation immigrants should be a prioritized target group.

Key words::

• Antibiotics
• diagnostic tests
• infectious diseases
• immigrant patients
• general practice

View correction statement:

CORRIGENDUM

Introduction

In general, non-western immigrants use medical care more often than natives (1,2), not only because they actually have more health problems, but also because their self assessed health is relatively low (1). As was shown in Dutch primary care, different expectations and demands of immigrants may prompt physicians to deviate from their routine clinical management strategies resulting in more diagnostic tests and higher prescription rates (2). Antibiotic prescription is of specific importance, since antibiotic resistance, mainly driven by overuse and misuse of antibiotics, is a major public health threat. Since the number of non-western immigrants in many western countries has risen rapidly over the last decades, it is of growing importance to study this question.

A study in the United States (3) found that even though immigrants expected antibiotics more often, there was no significant difference in antibiotic prescription rates. However, it is uncertain whether these results also apply to Western European countries since antibiotic prescription rates overall are high in the United States and the ethnic origins of immigrants dissimilar. Moreover, the study of Mangione-Smith (3) focused on paediatric population only. No data are available about possible differences in antibiotic prescription rates of non-western immigrants, both children and adults, in Europe. The aim of this study was to determine whether ethnic origin is an independent determinant of antibiotic prescription and use of additional diagnostic tests in primary care patients with an infectious disease.

Methods

Design

A cross-sectional study was conducted in four healthcare centres in a new residential area near the city of Utrecht in The Netherlands. This study was part of the Utrecht Health Project (UHP), a dynamic population study in which all new patients that register at one of the four health-care centres in the area are invited to participate (4). Participants fill out an individual health profile that includes questionnaires on socio-economic status, mental health, cardiovascular risk, dietary intake, lifestyle factors (e.g. smoking and alcohol) and housing. In addition, a standard set of biometrics (including electrocardiogram and lung function test) are done. Disease history and follow-up are registered, using the International Classification of Primary Care (ICPC) (5), in the patient's electronic medical file by general practitioners. More than 50% of the inhabitants in the area are currently participating. All data are gathered continuously in an anonymous central database. The study was approved by the Ethical Committee of the University Medical Center Utrecht.

Data collection

Data were obtained between December, 2000 and January, 2009. The study population comprised all 1015 non-western immigrants, of which 555 first and 460 second generation immigrants and 995 randomly selected native Dutch. Since most immigrants in the Netherlands are of Turkish, Moroccan, Surinamese and Antillean/Aruban origin (1), immigrants of other countries were excluded from the analyses.

Variables

We applied the definition of the Statistics Netherlands, a governmental institution that publishes official Dutch statistics, to determine ethnicity (6). The definition considers a person to be non-Dutch if he/she or at least one parent was born in Turkey, Morocco, Suriname or the Antilles. Immigrants were classified as first-generation immigrants if they were born in one of the countries stated and second-generation if they were born in The Netherlands, but at least one parent was born in one of the selected countries.

Medication prescription was based on the Anatomical Therapeutic Chemical (ATC) classification system (7). Antibiotics most frequently used in Dutch primary care were included in the analyses: amoxicillin (J01CA04); amoxicillin and clavulanate potassium (J01CR02); flucloxacillin (J01CF05); sulfamethoxazole and trimethoprim (J01EE01); erythromycin (J01FA01); clarithromycin (J01FA09); azithromycin (J01FA10); ciprofloxacin (J01MA02); and doxycycline (J01AA02).

For the analyses of additional diagnostic tests, only respiratory tract infections were included, since these are the most common infectious diseases for which a general practitioner is consulted (8). Respiratory tract infections included any of the following ICPC-codes: upper respiratory tract infections (R74), sinusitis (R75), acute tonsillitis (R76), acute laryngitis/tracheitis (R77), acute bronchitis/bronchiolitis (R78), influenza (R80), pneumonia (R81), pleurisy/pleural effusion (R82) and other respiratory infections (R83). Additional diagnostic tests included blood tests (C-reactive protein, CRP; Erythrocyte Sedimentation Rate, ESR; or leukocytes) and chest X-rays.

Covariates included in the analyses were age, sex, education, presence of chronic diseases and smoking. Education was defined as the highest completed level of education and was divided into three groups: low education (incomplete or completed primary education); average education (completed high school or secondary vocational education); and high education (higher vocational education or university). For children under the age of 19, the highest level of education completed by the parents was used. Chronic diseases included those that are associated with higher antibiotic use: diabetes, asthma, COPD, tuberculosis, cystic fibrosis, HIV, or lung cancer. Smoking was defined as current smoking. For children we included also passive smoking (at least one of the parents smoked inside the house).

Statistical analysis

Logistic regression analyses were used to study the independent relationship between the ethnic origin and antibiotic use as well as use of additional diagnostic tests. All analyses were adjusted for age, sex, education, presence of chronic diseases and smoking. We also repeated the analyses including only subjects with low and average education only. We adjusted the analyses for age, sex, presence of chronic diseases and smoking. Data were analysed by using SPSS (version 15) for Windows (SPSS Inc, Chicago, IL, USA).

Results

Population characteristics

As expected, the mean age of second generation immigrants was lower than of first generation immigrants or native Dutch patients (Table I). The highest prevalence of chronic diseases was in first generation non-western immigrants (17.8%). In this group asthma (9.9%), diabetes (7.6%) and COPD (1.6%) were the most common diseases. Asthma and diabetes were also the most common chronic diseases in second generation non-western immigrants and native Dutch patients.

Table I. Characteristics of study population.

Patient characteristics^a	Native Dutch n = 995	First generation non-western immigrants^a n = 555	Second generation non-western immigrants* n = 460
Male	44.1	43.6	49.6
Age (years), mean ± SD	39.4 ± 17.5	41.9 ± 11.1	18.5 ± 11.1
Chronic diseases^b	11.9	17.8	12.4
Current smoking^c	24.6	28.1	27.0
Education^d
Low education	6.7	20.6	13.0
Average education	51.4	57.8	65.4
High education	41.8	21.6	21.6

All data in percentages, or as indicated.

^aIncluding Turks, Moroccans, Surinamese and Antilleans/Arubans.

^bIncluding any of the following diseases: asthma, COPD, diabetes, tuberculosis, cystic fibrosis, HIV, an infection of the circulatory system or lung cancer.

^cIn children: also passive smoking.

^dHighest completed education; in children: highest completed education of the parents. Low education was defined as incomplete or completed primary education, average education as completed high school or secondary vocational education and high education as completed high vocational education or university.

Antibiotics and diagnostic tests

Antibiotics were prescribed in 36.4% of the native Dutch population, in 46.7% of the first and 32.6% of the second generation non-western immigrants. First generation non-western immigrants were more likely to receive antibiotics than native Dutch after adjustment for age, sex, education, presence of chronic diseases and smoking (Table II). No differences were found for the second generation immigrants.

Table II. The effect of ethnic origin on antibiotic use and additional diagnostic tests (adjusted odds ratios; total study population, n = 1839^c).

	Adjusted odds ratio (95% CI)
Determinants	Antibiotic use	P value	Use of additional diagnostic tests	P value
Ethnic origin		0.017		<0.001
Dutch	1 (reference)		1 (reference)
First generation immigrants	1.309 (1.039–1.650)	0.022	2.079 (1.642–2.634)	<0.001
Second generation immigrants	0.857 (0.631–1.165)	0.324	1.385 (1.003–1.911)	0.048
Age (years)	1.006 (0.999–1.013)	0.080	1.035 (1.027–1.043)	<0.001
Sex
Male	1 (reference)
Female	1.590 (1.304–1.938)	<0.001	2.231 (1.816–2.741)	<0.001
Chronic disease^a
No chronic disease	1 (reference)		1 (reference)
Any chronic disease	2.579 (1.943–3.422)	<0.001	1.525 (1.131–2.055)	0.006
Education^b		<0.001		0.001
Low education	1 (reference)		1 (reference)
Average education	0.829 (0.609–1.129)	0.24	0.831 (0.601–1.148)	0.262
High education	0.543 (0.384–0.767)	0.001	0.567 (0.397–0.809)	0.002

Smoking was not significant.

P values for ethnic origin and education show the overall significance of these variables in the model.

^aIncluding any of the following diseases: asthma, COPD, diabetes, tuberculosis, cystic fibrosis, HIV, an infection of the circulatory system or lung cancer.

^bHighest completed education; in children: highest completed education of the parents. Low education was defined as incomplete or completed primary education, average education as completed high school or secondary vocational education and high education as completed high vocational education or university.

^cExcluding 171 subjects having missing values.

We compared antibiotic use between the immigrants and native Dutch patients and found significant differences for amoxicillin and azithromycin. Immigrants were prescribed amoxicillin and azithromycin more often than native Dutch patients (Fisher's test, P <0.05).

In 37.3% of the native Dutch population, in 57.3% of the first and 28.5% of the second generation non-western immigrants additional diagnostic tests were performed. The use of diagnostic tests was more frequent in first generation non-western immigrants than in native Dutch patients also after adjustment for age, sex, education, presence of chronic diseases and smoking in the logistic regression analysis (Table II). As with antibiotic prescription, no differences were found for second generation immigrants.

Educational level

We compared the educational level of both native Dutch and non-western immigrants in our study with the national data for both groups obtained from the Statistics Netherlands. In both groups, our study population was better educated. However, when we repeated the analyses including only low and average educated subjects the results were similar: most antibiotics were prescribed in first generation immigrants (OR: 1.34; 95% CI: 1.03–1.74) and most diagnostic tests were performed in this group as well (OR: 1.88; 95% CI: 1.43–2.48). In second generation non-western immigrants, no differences were found for antibiotic use (OR: 0.84; 95% CI: 0.59–1.22) or diagnostic tests (OR: 1.16; 95% CI: 0.79–1.69).

Discussion

Several studies in primary care have investigated determinants of antibiotic use such as age (9), sex (10) and clinical presentation (9). Our aim was to examine the effect of ethnicity on antibiotic prescription. We found relevant differences in antibiotic prescription rates and the use of additional diagnostic tests between first generation immigrants and native Dutch. No differences were found for the second generation immigrants. This may be explained by an increased integration of the second generation immigrants to Dutch society and better language competence (11).

Our study is in line with a qualitative study conducted in Germany showing that GPs are under pressure to prescribe antibiotics particularly in consultations with Turkish immigrants (12). Carrasco-Garrido et al. (13) also found a trend towards higher antibiotic use among non-western immigrants in Spain. The differences in antibiotic use between immigrants and native Dutch patients were mainly caused by differences in the use of amoxicillin and azithromycin, which together with doxycycline are the most frequently used antibiotics in Dutch primary care.

Strengths and limitations

A limitation of our study was that the questionnaire on the individual health profile was available only in Dutch, thus potentially limiting the number of respondents in lesser educated immigrants. This does not undermine our results but could have lead to an underestimation of the differences we found between first generation immigrants and native Dutch patients. When we compared both groups (native Dutch and immigrants) to the national data we found that education in both groups was higher than their respective national averages. However, when we limited the analyses to low and average educated participants, results were similar. Another potential limitation was that only education was used to correct for socioeconomic status (SES). Additional indicators like employment, income and family composition were not included, because of the high number of missing values. However, in epidemiological studies education is the most frequently used measure of SES (14). Lastly our findings could have been confounded by real differences in health between first generation immigrants and native Dutch persons. However, we adjusted the analyses for the most important and prevalent indicators for health like co-morbidity and age. Therefore, we think that the differences we saw are not due to rational somatic indications but are driven more by cultural and social reasons.

Implications

The highest rates of antibiotic prescription and diagnostic tests concern first generation non-western immigrants. In order to promote a rational and restricted use of antibiotics in outpatients first generation immigrants should be a prioritized target group. More research is needed in order to understand better the reasons for high use of antibiotics and additional diagnostic tests in these ethnic groups to facilitate the development of culturally specific interventions aimed at rationalizing antibiotic use.

Acknowledgments

This work was supported by the University Medical Center Utrecht. The Utrecht Health Project LRGP received grants from the Ministry of Health, Welfare, and Sports (VWS), the University of Utrecht, the Province of Utrecht, the Dutch Organisation of Care Research (ZON), the University Medical Center Utrecht (UMC Utrecht) and the Dutch College of Healthcare Insurance Companies (CVZ). The authors acknowledge the participating inhabitants of Leidsche Rijn, Utrecht, The Netherlands, and the general practitioners working in this area for providing research data from routine care.

Ethical approval

None.

Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.