How do patients with a Turkish background evaluate their medical care in Germany? An observational study in primary care

Abstract

Background

Patients’ evaluation of medical care is an essential dimension of quality of care and an important aspect of the feedback cycle for health care providers. The aim of this study was to document how patients with a Turkish background evaluate primary care in Germany and determine which aspects of care are associated with language abilities.

Methods

The study was based on an observational design. Patients with a Turkish background from German primary care practices completed the EUROPEP (European Project on Patient Evaluation of General Practice Care) questionnaire consisting of 23 items. Seventeen primary care practices were involved with either German (n=8) or Turkish (n=9) general practitioners (GPs).

Results

A convenience sample of 472 patients with a Turkish background from 17 practices participated in the study (response rate 39.9%). Practices with a German GP had a lower response rate (19.6%) than those with a Turkish GP (57.5%). Items evaluated the highest were “keeping data confidential” (73.4%) and “quick services for urgent health problems” (69.9%). Subgroup analysis showed lower evaluation scores from patients with good or excellent German language abilities. Patients who consulted a Turkish GP had higher evaluation scores.

Conclusion

The evaluation from patients with a Turkish background living in Germany with either Turkish or German GPs showed lower scores than patients in other studies in Europe using EUROPEP. However, our results had higher evaluation scores than those of Turkish patients evaluating GPs in Turkey. Therefore, different explanation models for these findings should be explored in future studies.

Keywords:

• migrant health
• quality of care
• primary health care
• patient satisfaction
• patient evaluations

Introduction

Patients’ evaluation of medical care is an essential dimension of quality of care and consists of a combination of both patient expectations with regard to health care providers as well as actual patient experience.¹ It is also an important aspect of the feedback cycle in quality assurance to identify opportunities for improvement.^2,3 In primary care settings in Germany, general practitioners (GPs) belong to the most frequented group of physicians. Over 90% of the population consults a GP annually. Increasingly, in Germany, the position of GPs as coordinators of care has been emphasized in recent years.⁴ GPs have contact with different groups of patients including those who have an immigrant background in their daily practice.⁵

Almost 16 million people in Germany have a migrant background with 2.5 million Turkish people being the largest group of migrants.⁶ If they are legal immigrants they could have a health insurance. Furthermore, migrants have a lower utilization of health services than the German nationals. The health care of this population group is mostly influenced by socioeconomic factors (eg, lower educational level) and the access to health care could be limited due to language problems but also due to cultural aspects.⁷ However, quality of care for migrant patients has been shown to be underdeveloped combined with an enormous research gap.⁸ It has been shown that quality of primary health care for Turkish immigrants is below that of German nationals.^9,10 There is a research gap on the perspective of Turkish patients regarding their medical encounter with the GPs.

The aim of this exploratory study was to document how patients with a Turkish background evaluate their primary care in Germany and determine which aspects of care are associated with language abilities.

Methods

Design and participants

This study was based on an observational design in which patients with a Turkish background from 17 GP practices completed a questionnaire. Eight GPs were German and nine GPs had a Turkish background (and therefore, could communicate with the patient in Turkish fluently). Participating German GPs represented a convenience sample and were recruited due to personal contact whereas GPs with a Turkish background were identified via the register of the National Association of Statutory Health Insurance Physicians (Kassenärztliche Vereinigung). In all, 143 GPs with a Turkish background were invited to participate in our study by mail. An inclusion criterion for the Turkish GPs was to speak with the patient in Turkish fluently.

Inclusion criteria for the patients were: age over 18 years, sufficient Turkish competence to answer the written questionnaire without support, and formerly known in the practice (no casual patients). Patient participation was voluntary and anonymous. Patients were invited to participate at the GP reception desk to fill in the questionnaire in the waiting room and to put it in a sealed box. The box was subsequently forwarded to the research department. Data collection took place in July and August 2011. The return of the anonymous paper-based questionnaire was classified as informed consent. Because of the recruitment procedure, the participating GP practices were located in four different federal states in Germany (Bavaria, Baden-Wuerttemberg, Lower Saxony, and North Rhine-Westphalia).

Ethics

The ethics committee of the University Hospital Heidelberg informed us that approval by an ethics committee was not necessary for this observational study. The questionnaires were completed anonymously. No identifying information or data from patients were collected.

Instrument

The validated Turkish version of the EUROPEP (European Project on Patient Evaluation of General Practice Care) questionnaire was used.¹¹ The EUROPEP is an established and internationally validated instrument reflecting a set of indicators for patients to evaluate the quality of primary care.¹² This questionnaire is a multidimensional instrument that comprises 23 items consisting of five dimensions: “relationship and communication”, “medical care”, “information and support”, “continuity and cooperation”, and “availability and accessibility”. Patients were asked to evaluate care that they received in the previous 12 months by means of a 5-point Likert scale ranging from “poor” (1) to “excellent” (5) for each item. Additionally, the questionnaire included 11 questions on demographic characteristics (eg, education level, parental birthplace, German language ability) (Table 1). These are based on the basic set of indicators for mapping migrant status developed by Schenk et al.¹³

Table 1 Characteristics of patients (n=472)*

Characteristics	Number	Percent
Sex
Men	220	46.6
Women	252	53.4
Age, years
18–40	226	47.9
41–59	175	37.1
≥60	47	10.0
Highest level of education
No school attendance	15	3.2
“Grundschule”, primary school	121	25.6
“Hauptschule”, secondary school	93	19.7
High school	195	41.3
More than high school	41	8.7
Nationality
German	105	22.2
Turkish	357	75.6
Self-reported residence status
Limited	45	9.5
Unlimited	406	86.0
Self-reported language abilities
Excellent	127	26.9
Good	98	20.8
Medium	194	41.1
Poor	46	9.7
Very poor	5	1.1
Number of visits to a general practitioner (per year)
1–4	189	40.0
5–8	129	27.3
9–12	78	16.5
>12 times	65	13.8
Start of seeing a general practitioner
Less than 1 year	45	9.5
Between 1 and 4 years	126	26.7
Between 5 and 10 years	142	30.1
More than 10 years	146	30.9

Note:

* n varies due to missing values.

Data analysis

Analyses were performed using SPSS 20.0 (SPSS Inc., IBM Corporation, Armonk, NY, USA). Continuous data were summarized using means and standard deviations. Categorical data were presented as frequency counts and percentages. The proportion of all Turkish patients and Turkish patients with good or excellent German language abilities who answered with a category 5 (excellent) of the items of EUROPEP was reported. The comparison of the nationality of GPs (Turkish or German) regarding the evaluation of care from Turkish patients was analyzed with the nonparametric Mann–Whitney U-test. Furthermore, the mean scores for each of the five domains (relationship and communication, medical care, information and support, continuity and cooperation, and facilities availability and accessibility) were computed. Moreover, a linear regression analysis was performed for the whole study population. The self-reported language ability was the outcome variable and the five domains of EUROPEP, the characteristics of participating practices, GPs of these practices (sex and nationality of GP), and patients (age, sex, number of visits to the GP per year and the point of time patients started seeing a GP) were potential predictors. An α level of P<0.05 was used for tests of statistical significance.

Results

In total, data from 17 practices (85%) were collected. From 1,190 patient questionnaires that were handed out, 472 were returned, which resulted in a response rate of 39.9%. Practices with a German GP (eight practices) showed a considerably lower response rate (19.6%) than practices with a Turkish GP (nine practices) (57.5%). The mean age of the patients was 40.7 years (ranging between 18 and 92 years). The mean age of the GPs was 51.2 years (ranging between 37 and 66 years). The characteristics of the participating patients and GPs are illustrated in Tables 1 and 2.

Table 2 Characteristics of general practitioners (n=17)

Characteristics	Number	Percent
Sex
Men	13	76.5
Women	4	23.5
Age, years
<40	2	11.8
40–49	4	23.5
50–60	6	35.3
>60	5	29.4
Nationality
German	8	47.1
Turkish	9	52.9
Mode of practice
Solo	12	70.6
Group practice	5	29.4
Location of practice
≥15,000 inhabitants	14	82.4
<15,000 inhabitants	3	17.6

Patients’ evaluation of primary care

Table 3 shows the descriptive results of Turkish patients’ evaluation of their health care for each item of the EUROPEP questionnaire, including the proportion of patients who answered a category 5 (excellent). The number of missing values was negligible. Most of the positive ratings were given for the domain “relationship and communication” (“listening to you”, 69.2%; “keeping your records and data confidential”, 73.4%) and for the domain “medical care” (“thoroughness”, 66.4%; “physical examination”, 67.2%). The most negative perceptions were found in the domain “availability and accessibility” (“being able to speak to the GP on the telephone”, 46.8% and “waiting time in the waiting room”, 31.1%).

Table 3 Evaluation of care by Turkish patients (n=357)

EUROPEP questionnaire	Percentage of all patients (n=357)*	Percentage of patients with good or excellent German language abilities (n=153)*
Relationship and communication
1. Making you feel you had time during consultations	49.3	42.5
2. Interest in your personal situation	57.4	53.6
3. Making it easy to tell about your problems	57.4	49.7
4. Involving you in decisions about medical care	51.8	48.4
5. Listening to you	69.2	64.1
6. Keeping your records and data confidential	73.4	70.6
Medical care
7. Quick relief of your symptoms	49.3	43.8
8. Helping to perform your normal daily activities	51.3	44.4
9. Thoroughness	66.4	58.2
10. Physical examination	67.2	60.8
11. Offering you services for prevention	59.1	54.9
Information and support
12. Explaining the purpose of tests and treatments	63.0	58.2
13. Telling about your symptoms and/or illness	64.1	57.5
14. Help in dealing with emotional problems	54.1	49.7
15. Helping understand the importance of following advice	55.2	46.4
Continuity and cooperation
16. Knowing what has been done during previous contacts	54.9	47.7
17. Preparing what to expect from specialists	54.3	45.8
Facilities availability and accessibility
18. The helpfulness of the staff	59.4	50.3
19. Getting an appointment to suit you	62.7	55.6
20. Getting through to the practice on telephone	63.3	58.2
21. Being able to speak to the GP on the telephone	46.8	41.8
22. Waiting time in the waiting room	31.1	28.1
23. Quick services for urgent health problems	69.9	52.9

Note:

* With answer 5 (excellent).

Abbreviations: EUROPEP, European Project on Patient Evaluation of General Practice Care; GP, general practitioner.

The evaluation scores of patients with good or excellent German language abilities were lower than those of other participants. The results of the subgroup analysis for patients attending GPs of either German or Turkish nationality, including mean, standard deviation, and P-values of the EUROPEP, are given in Table 4.

Table 4 Turkish patients’ perspective on care: a comparison of Turkish and German GPs

EUROPEP questionnaire	Nationality of GP				P-value*
	German		Turkish
	n	Mean (SD)	n	Mean (SD)
Relationship and communication
1. Making you feel you had time during consultations	108	4.0 (1.10)	346	4.14 (1.07)	0.18
2. Interest in your personal situation	108	3.96 (1.31)	347	4.31 (1.07)	0.009
3. Making it easy to tell about your problems	109	3.99 (1.36)	342	4.40 (0.90)	0.009
4. Involving you in decisions about medical care	106	4.08 (1.18)	333	4.30 (1.02)	0.06
5. Listening to you	108	4.22 (1.20)	351	4.52 (0.90)	0.02
6. Keeping your records and data confidential	108	4.25 (1.26)	343	4.59 (0.93)	0.007
Medical care
7. Quick relief of your symptoms	107	3.96 (1.31)	337	4.16 (1.17)	0.15
8. Helping to perform your normal daily activities	106	3.93 (1.32)	338	4.20 (1.17)	0.04
9. Thoroughness	107	4.22 (1.17)	348	4.50 (0.86)	0.048
10. Physical examination	109	4.25 (1.13)	346	4.52 (0.89)	0.03
11. Offering you services for prevention	107	4.06 (1.26)	336	4.43 (1.01)	0.002
Information and support
12. Explaining the purpose of tests and treatments	105	4.18 (1.23)	335	4.46 (0.97)	0.02
13. Telling about your symptoms and/or illness	106	4.19 (1.18)	348	4.51 (0.88)	0.006
14. Help in dealing with emotional problems	106	3.82 (1.52)	342	4.30 (1.11)	0.003
15. Helping understand the importance of following advice	107	4.06 (1.31)	337	4.41 (0.88)	0.03
Continuity and cooperation
16. Knowing what has been done during previous contacts	105	3.98 (1.37)	343	4.34 (0.95)	0.06
17. Preparing what to expect from specialists	108	3.90 (1.43)	331	4.31 (1.13)	0.005
Facilities availability and accessibility
18. The helpfulness of the staff	108	3.95 (1.28)	337	4.34 (1.20)	<0.001
19. Getting an appointment to suit you	107	3.92 (1.32)	346	4.33 (1.22)	<0.001
20. Getting through to the practice on telephone	109	3.90 (1.34)	340	4.37 (1.16)	<0.001
21. Being able to speak to the GP on the telephone	107	3.57 (1.57)	319	3.95 (1.49)	0.006
22. Waiting time in the waiting room	107	3.41 (1.39)	338	3.68 (1.38)	0.04
23. Quick services for urgent health problems	109	4.00 (1.46)	340	4.39 (1.13)	0.005

Note:

* P<0.05.

Abbreviations: GPs, general practitioners; SD, standard deviation.

Overall, nearly 75% of the participants with poor German language abilities visited Turkish GPs. Table 4 shows statistically significant differences within almost all items of EUROPEP when comparing patients consulting a Turkish GP with those consulting a German GP.

Table 5 shows the linear regression analysis of the study population characteristics and the five domains of EUROPEP on language abilities of patients. The model explained 35% (R²=0.352) of the variance on the dependent variable “self-reported language ability”. Significant associations were found for two domains of EUROPEP questionnaire: “information and support” and “continuity and cooperation”. Patients with poorer language abilities were older and visited the GP more often and their “information and support” domain was worse, but interestingly “continuity and cooperation” was rated better than that of patients with better language abilities (self-reported).

Table 5 Associations of the study population characteristics and the five domains of EUROPEP on language abilities (results of the linear regression analysis, under specification of standardized β-coefficient, α=5%)

Associated items	β-coefficient	P-value*
Domains of EUROPEP questionnaire
Relationship and communication	0.027	0.728
Medical care	−0.016	0.841
Information and support	−0.175	0.039
Continuity and cooperation	0.193	0.005
Facilities availability and accessibility	0.031	0.585
Practice and GP characteristics
Type of practice	0.060	0.212
Nationality of GP	0.024	0.617
Sex of GP	−0.074	0.132
Age of GP	−0.115	0.018
Patient characteristics
Sex	0.106	0.009
Age	0.496	<0.001
Number of visits to the GP per year	0.177	<0.001
The point of time patients started	0.004	0.926
seeing a GP
R^2	0.352

Note:

* P<0.05.

Abbreviation: GP, general practitioner.

Discussion

The aim of this pilot study was to evaluate the care experiences reported by patients with a Turkish background living in Germany and determine which EUROPEP domains were associated with patients’ language abilities. In general, patients were very positive about their care with the exception of “waiting time in the waiting room”. This finding is comparable with another study using EUROPEP in Germany.¹⁴

Both dental care and primary care patients showed the lowest satisfaction level regarding the item “waiting time in the waiting room”.^14,15 Similar results were shown for patient satisfaction in Slovenia.¹⁶

Moreover, a study about patient satisfaction with family practice in Turkey showed the lowest satisfaction regarding “being able to speak to the GP on the telephone” and “getting through to the practice on the phone”.¹⁷ These results are comparable to our study for patients with poor German language skills.

In subgroup analysis, language abilities showed the greatest impact on the factor of patient satisfaction. Patients with poor German language skills consulted a Turkish GP most frequently (75%). According to our subgroup analysis, there were several interpretation levels. Turkish GP visits were associated with higher satisfaction rates (Table 4). These findings reflect the effect of ethnic doctor–patient concordance on patient satisfaction.^18,19

However, there is another important aspect related to German language ability. In our study, patients with good or excellent German language abilities showed overall less satisfaction in the whole study population. These findings were surprising because the impact of communicational barriers on effective doctor–patient interaction has been widely discussed.^20,21 It was anticipated that patients with good German language abilities would adapt and reflect the German patients in the study population, as it was found for other aspects of medical care, for example, the utilization of health services.²² It could be assumed that patients who are able to communicate on an equal level could be more critical and more discerning about health care providers. Moreover, apart from communication problems, other factors in the doctor–patient encounter could have an impact on treatment as perceived by these patients. Racial or ethnic discrimination in health care has already been discussed and identified as an important but variable factor in medical encounters.²³ Too little is known about patients’ perspectives and experiences with German GPs.

Results from the regression analysis showed that patients with poorer German language abilities evaluated their “information and support” lower but “continuity and cooperation” better than that of patients with better German language abilities. Information about treatment or understanding the importance of following advice is an important component within a doctor–patient interaction. This needs more attention for patients with poorer language abilities. Furthermore, continuity and cooperation are very important not only for patients with long-term or chronic conditions, but also with poorer language abilities.²⁴ More research is needed to evaluate the meaning and consequences of continuity and cooperation for patients with poorer language abilities.

These results could be also associated with a cultural understanding and interpretation of the medical encounter and expectations toward the GP that will differ from other European countries. Previous studies showed that patients’ experiences are related to expectation rather than to GP performance.^25,26 In Turkey, great respect is given to doctors based on a more hierarchical perception of the doctor as a “learned man”.²⁷ It can be assumed that children of second and third generation migrants could also apply this interaction model inherited from their parents. It could also be assumed that high expectations cannot be fulfilled by GPs for multiple reasons (such as lack of time, different understandings of illness and treatment) or even paradigm shifts as the ideal of the omnipotent physician falls apart in modern society, which all lead to lower patient satisfaction rates. Overall, more research is needed to further explore the importance of this aspect.²⁸

Strengths and limitations

A major strength of this study is the utilization of the patients’ native language in the questionnaire. Many surveys exclude foreign language-speaking patients for methodological reasons, meaning little is known about the needs and wishes of these patient groups. The study is limited by the overall low return rate of questionnaires from practices with German GPs. Due to the small number of participating practices, the results of this study cannot be seen as representative for Germany. A selection bias may have been introduced through the mode of recruiting the practices and participating patients. We did not assess the health condition of the participating patients. However, we compared our data with evaluations of European patients with chronic diseases. Since previous studies showed associations with chronic conditions and lower patient satisfaction rates,^29,30 results of direct comparisons need to be interpreted carefully. Nevertheless, we oriented our study design, performance, and data analysis strongly to previous EUROPEP studies.^14,15 Following the analysis of EUROPEP in other studies, only the numbers scoring “5” were reported.^14,15 In general, the EUROPEP shows a ceiling effect not only in our study but also in previous studies.^14,15 There is no cut-off of the EUROPEP tool for clinically significant differences available. Moreover, we believe that the developed hypotheses and theories can implicate important new aspects and views of patients with a Turkish background to their primary care.

Conclusion

Satisfaction with primary care of patients with a Turkish background is, in this study, overall lower than in comparable EUROPEP studies in Europe, though higher than satisfaction rates in Turkey itself. These findings indicate differences in perceived quality of care based on several levels and different aspects of primary care (health care system, doctor/staff–patient relationship, culture, language abilities) that need further research. Particular attention should be paid to the influence of German language skills. Though good or excellent language competencies seem to equalize ethnical doctor–patient discordance, German-speaking patients were overall less satisfied with the received health care. In the context of the ongoing debate about how effective practice in health care for migrant patients should be implemented, these results could raise issues that need attention in the German health care policy. Given the current situation of migration to Germany, it is important to evaluate the actual needs of patients with different cultural backgrounds for ensuring good quality of care and sensitize GPs for the caring of the migrant population.

Acknowledgments

We would like to thank Dr Zekeriya Aktürk from the Department of General Practice of Trakya University in Edirne, Turkey, for providing the Turkish version of the EUROPEP instrument. We also like to thank the participating practices and patients. Finally, we also gratefully thank Sarah Berger, a native speaker, for reviewing this manuscript.

Disclosure

The authors report no conflicts of interest in this work.