OUP user menu

Individual health services and the denial of health services in German medical practices: prevalence, regional differences and socio-demographic determinants

Susanne Richter, Heinke Rehder, Heiner Raspe
DOI: http://dx.doi.org/10.1093/eurpub/ckp145 491-498 First published online: 12 October 2009


Background: Internationally, in many healthcare systems financial pressure has led to the implementation of co-payments, private medical (extra) services and rationing. In Germany, members of statutory health insurances (SHIs) increasingly report the denial of medical services and the offer/demand of privately financed supplementary health services individual health services, (IHSs) in medical practices. The public discussion on both denial and IHSs is chequered, mainly critical, partly polemic. The present study aims to operationalize IHSs and denial and investigates their occurrence, socio-demographic determinants within two regional populations. Methods: Two postal surveys were conducted in 4898 German inhabitants of Lübeck (Northern Germany) and Freiburg (Southern Germany), aged 20–79 years. The survey focused on experiences with IHSs and denial of health services in medical practices among members of SHIs. Results: In all members of SHIs that had consulted a physician during the past 12 months (n =1899), the one-year-prevalence of IHSs and denial of medical services were 41.7 and 20.5%. About 40% were offered a denied medical service as an IHS later. Conclusion: The study presents population-based, quantitative data on IHSs and denial of medical services in German practices. The results partly confirm former findings on the occurrence of IHSs. Contrary to other studies, socio-demographics seemed to play a minor role in the offer/demand of IHSs.

  • denial of medical services
  • Germany
  • individual health services
  • population-based survey
  • statutory health insurance.


Financing healthcare systems is an internationally pressing problem. Instable numbers of premium payers in Bismarck systems, cuts in public sector spending in Beveridge systems and increasing costs have led to rising out-of-pocket costs for patients, e.g. co-payments, cost-sharing, private medical services and top-up treatments. These market developments seem to be linked to implicit and explicit strategies of rationing with regard to medical services, which has also been stressed across systems repeatedly.1,2 Discussions on healthcare markets recurrently stress ethical implications, values, implicit and explicit rationing, access to healthcare and equity issues.3–5 In the British system, a current discussion deals with ethical dilemmas at the microlevel if patients wish to top-up National Health Services (NHSs) with treatments that are not supported by either doctors or the National Institute for Health and Clinical Excellence.6,7 Similar discussions and developments have been reported for Canada, Australia, New Zealand, Italy, Spain, Norway and Sweden.8 In the context of international ongoing trends to extend patients’ personal responsibilities and out-of-pocket medical services, the German experience with potential implications and hazards of an unregulated second healthcare market should provide valuable information for other countries.

Within the German system, healthcare is funded by a statutory contribution system. Statutory and private health insurances co-exist. About 90% of the population is covered by statutory health insurance (SHI) schemes, 10% is completely covered by private health insurances. In Germany, compulsory, voluntary and family SHIs have to be distinguished. In general, all workers are compulsorily SHI-insured if their gross pay does not exceed a defined upper limit (in 2009: €48 600 before tax). Insured persons are free to choose an SHI fund. Many people carry voluntary SHI (e.g. self-employed, civil servants, gross pay exceeding upper limit). Provided certain conditions are met, students, spouses, partners and children of the SHI-insured person are covered as family members who do not have to pay contributions. The SHI payment depends on the income and is unevenly shared between employee and employer. From the year 2009 the SHI contribution is to be fixed by federal state law (at present at 15.5% of the income). The package of services covered by the SHIs is comprehensive and defined in Social Code Book V. A ‘Federal Joint Committee’ decides on the inclusion/exclusion of medical services in/from the SHIs benefits catalogue. Excluded services are partly covered by additional private health insurances (i.e. extended coverage) via reimbursement, or provided as privately financed services. Formal waiting lists apply to transplantations only. Outpatient healthcare services are mainly supplied to the public by independent physicians, who work on a freelance basis under contract of the social health insurances. By law, doctors caring for members of the SHIs have to be registered by the Regional Association of SHI-Accredited Physicians.9,10

Several public endeavours have been made in Germany to control the healthcare expenses, e.g. by health acts in 1993, 2004, 2006 and 2007. In this context, for instance, the benefits catalogue of the SHIs has been slightly reduced, and cost-sharing of patients has successively risen. In 1998 the concept of individual health services (IHSs) was introduced to offer medical services that are (for several reasons) beyond the SHIs benefits catalogue.11 IHSs have not been consistently defined, yet. As a central characteristic, they are not paid or reimbursed by the SHIs, but are to be paid for privately by the patient alone. IHSs catalogues include highly beneficial and evidence-based methods (like travel medicine, malaria prophylaxis), questionable/probably beneficial services (e.g. screening for lung cancer by sputum cytology), dubious/untested methods from alternative/complementary medicine (e.g. ayurveda) and services to enhance beauty, fitness or wellness (e.g. anti-ageing). Comprising an increasing number of currently >300 services, and with regard to the stagnating incomes of SHI-accredited physicians, IHSs have established a second healthcare market.12–14

The public discussion of IHSs and the denial of medical services are chequered, rather critical, and partly polemic. In 2006, the German Medical Association has developed a 10-point paper, which aims at the establishment of a code for SHI-accredited physicians on how to deal with IHSs.15 Current data demonstrate an increasing experience with IHSs and denial among members of the SHIs.16,17 According to a telephone survey conducted among SHI-insured persons in 2008 (n = 3000), 27% were offered or charged an IHS in a medical practice during the past 12 months, with an estimated turnover of one billion Euros.18 At the same time, IHSs gain economic importance for the medical profession.19,20 Simultaneously, the experience with the denial of expected or demanded medical services shows an increasing trend.21 Up to 16% of the SHI-insured population report the denial of a medical service at the most recent consultation, and 58% fear further reductions. Persons suffering from disabilities or chronic diseases seem especially concerned.22 In a regional population survey on low back pain (Lübeck, HL, Germany, 2003), among respondents reporting back pain related consultations within the past 3 months, 19% (n = 58) were offered an IHS, mainly initiated by orthopaedists, two-thirds of the patients accepted the offer.23

How often patients are offered a denied service as an IHS is uncertain. Merely one survey revealed, that one-third of the SHI-patients was subsequently offered a service which was denied as an IHS.24 Thus, the aim of the study was to present independent, population-based data from statutory health insured persons on:

  1. the lifetime and one-year prevalence, aims, explanations and regional differences of offered and demanded IHSs in medical practices;

  2. the lifetime and one-year prevalence, explanations and regional differences of denied medical services in medical practices;

  3. the combined occurrence of denial and IHSs;

  4. socio-demographic determinants on the occurrence of IHSs/denials of medical services.

The study has been approved by the ethics committee of the University at Lübeck (reference number 06-141).



A systematic sample of German inhabitants aged 20–79 years was drawn from the population registries of the cities of Lübeck (HL; n = 2748) in the North and Freiburg (FR; n = 2750) in the South of Germany. Apart from the address, data on age and gender were transferred by the registry offices. In Germany, southern areas are generally more affluent than northern regions. The cities of HL and FR were chosen to investigate a potential north–south divide with regard to the prevalences. Both cities are university towns of similar size (213 000 vs. 220 000 inhabitants), surrounded by rural districts of Western Germany.

Development of the survey instrument

The development of the questionnaire included an in-depth enquiry of existing German surveys on denial and IHSs. No standardized questionnaires are available covering IHSs and the denial of medical services. Therefore, cognitive interviews and postal pre-tests were used to develop a valid instrument. Both techniques shed light on different problems and are intended to complement each other.25

A first version of the questionnaire was compiled and tested within a cognitive survey phase. The cognitive interviews were conducted in a population sample (n = 11) of HL and six patients of the orthopaedic/gynaecological ward of a single clinic in the HL region in August and September 2006. All interviews were performed semi-standardized by SR and HRe, using ‘concurrent think-aloud’ and ‘probing’ techniques, and gathering general information on the handling of the questionnaire.25 Especially statutory health insured persons, who had never been offered any IHSs, appeared to be overstrained by the provided scientific explanations. Furthermore, some of these definitions seemed to be incomplete.14 Therefore, within the postal pre-tests, IHSs were described as supplementary medical services that (i) have to be privately financed by the patient, and (ii) are not reimbursed by any statutory or additional private health insurance. Denial was defined as medical services that had not been prescribed although patients would subjectively have needed them (e.g. medicaments, examinations). The cognitive survey phase was followed by three postal pre-tests conducted in systematically drawn sub-samples from the study sample (October 2006 to January 2007).

Screening survey

Postal survey stage 1: screening

Following the postal pre-tests, the remaining sub-samples had 2448 persons from HL and 2650 from FR. To allow for equal sample sizes, 2450 addresses were randomly selected from the FR sample. We used a two-stage postal survey. A 4-page screening questionnaire gathered data on the occurrence of IHSs and denial, subjective health status and socio-demographic variables. The subjects were sent the screening questionnaire in February 2007 (cf. table 1; two postal reminders each 2 weeks apart). All addressed subjects were informed about the study and the subsequent (detailed) postal survey by a cover letter and a leaflet on frequently asked questions.

View this table:
Table 1

Characteristics and screening results of participating SHI members (n = 2120)

ParameterValid N FR/HLResponse CategoryTotalFRHLDifference FR/HL
Age (years)896/122420–4069332.736340.533327.0Chi2 = 48.3 P < 0.001
Gender896/1224Females123358.254360.669056.4Chi2 = 3.8 P = 0.051
Status of statutory health insurance896/1224Compulsory158874.964972.493976.7Chi2 = 7.0 P = 0.030
Additional private health insurance813/1109Yes36118.817321.318817.0Chi2 = 5.8 P = 0.016
School educationa (years)882/1198≤974635.922125.152543.8Chi2 = 172.2 P < 0.001
Current/last occupational status847/1138Blue-collar worker40020.211213.228825.3Chi2 = 47.8 P < 0.001
White-collar worker121861.456366.565557.6
Offer/demand of IHSs—ever863/1187Yes107252.345252.462052.2Chi2 = 1.8 P = 0.407
Denial of medical services—ever888/1214Yes55026.222725.632326.6Chi2 = 2.4 P = 0.301
Consultations of a physician—12 mb892/1221Yes189989.979989.6110090.1Chi2 = 0.2 P = 0.698
If consultations: denial—12 m788/1085Yes38420.515820.122620.8Chi2 = 0.2 P = 0.913
If consultations: offer/demand of IHS 12 m787/1079Yes77841.733041.944841.5Chi2 = 1.8 P = 0.406
If consultations and offer/demand of IHSs 12 m: utilization of IHSs328/446Yes56372.724273.832272.0Chi2 = 0.3 P = 0.577
ScaleValid N FR/HLTotal M (SD)FR M (SD)HL M (SD)Difference FR/HL
General health statusc888/12182.6 (0.95)2.4 (0.94)2.7 (0.95)T = −5.8 P < 0.001
  • FR = Freiburg; HL = Lübeck; IHS = individual health service; M = mean; SD = standard deviation; SHI = statutory health service

  • a: Higher = university or advanced technical college entrance qualification/secondary school education, lower = elementary/other/no school education

  • b: 12 m = during the past 12 months

  • c" Scale 1 = very good, 5 = bad

To estimate a potential non-response bias, two strategies were pursued. First, telephone interviews were conducted focusing on six central screening questions (socio-demographics, occurrence of IHSs and denial). Second, an additional questionnaire on reasons for non-participation was sent out with the second postal reminder.

Postal survey stage 2: details on experience with IHSs and denial

All (postal) screening respondents, who had reported the experience with IHSs and/or denial during the past 12 months, were sent a second questionnaire from April to May 2007 collecting detailed information on IHSs (e.g. aims, explanations) and/or denial (e.g. explanations, combination with IHSs; cf. table 2; two postal reminders each 2 weeks apart).

View this table:
Table 2

Experiences of SHI members with individual health services (n = 601) and denial (n = 238) during the past 12 months

Aspects of IHS
    Aims of IHSsa,bEarly diagnosis of diseases30650.9
Treatment of present diseases15926.5
Travel preparation396.5
Blood/laboratory results335.5
Strengthening the immune system325.5
Usage of CAM325.3
Maintenance of mental health284.7
Increase of physical attractiveness162.7
Capability tests132.2
Special services91.5
Prevention/treatment of environmental diseases81.3
Innovative diagnostics/treatments40.7
    Explanation for IHSsa,bSHI pays no longer34557.4
SHI pays not yet16727.8
SHI pays only in case of suspicious facts9916.5
IHS is more gentle/secure589.7
IHS is more beneficial477.8
Budget ceiling of physician was reached325.3
Do not know/remember305.0
No explanation284.7
Aspects of denial
    Implicit vs. Explicit denialcExplicit: physician mentioned, s/he could not or no longer prescribe or perform the treatment16268.1
Implicit: patient and physician did not talk about the denial, but the patient did not or no longer receive an expected treatment6326.5
    Denied services formerly utilized as SHI servicecYes17678.6
Physician's and patient's explanations for denialb,cSHI pays no longer12853.84217.6
Budget ceiling of physician was reached6527.36226.1
Medically not necessary2811.872.9
SHI pays not yet2711.393.8
SHI pays only in case of suspicious facts2510.593.8
Efficacy not evident198.0187.6
Referral to another physician125.083.4
No explanation104.272.9
Medical service not known to physician41.741.7
Side-effects or risks were too high31.341.7
Former experience of side-effects20.8
Understanding of physician's explanationdYes4724.7
Combined occurrence of denial and IHSs
    Senied service offered as IHSscYes, at once7735.5
Yes, by another physician94.1
Yes, at another consultation83.7
No, no offering at all5224.0
No, other prescription2612.0
No, received as SHI service at another consultation198.8
No, referral to another physician52.3
Unsure, do not know94.1
    If offered as IHSs: utilizationbYes6463.4
    If offered as IHSs: reasons for non-utilizationcToo expensive5565.5
Still irresolute1011.9
Not enough information89.5
Efficacy not evident56.0
Too risky22.4
  • CAM = complementary alternative medicine; EEG = electroencephogram; ECG = echocardiogram; IHS = individual health service; SHI = statutory health service

  • a: Proportion of all statutory health insured that had reported the offer/demand of at least one IHS during the past 12 months (n = 601)

  • b: Multiple answers were possible

  • c Proportion of all statutory health insured that had reported the denial of at least one medical service during the past 12 months (n = 238)

  • d: Proportion of all statutory health insured that were given an explanation by their physician (n = 190)

Prospective sample size calculation

On the basis of the screening survey, within a subgroup of 275 persons (n = 1100 respondents of one city who are members of a SHI, one gender, one additional differentiation) the expected prevalence of IHSs during the past 12 months (∼23%)26 can be estimated with a precision of ±5%. With 80% power, a type I error of 5%, a subgroup of 1100 (or n = 550) persons and an IHS prevalence of about 20%, a two-tailed test can detect statistically significant group differences (e.g. HL vs. FR) of 5% points (or 7%).

Data analysis, statistics

For data entry a PC-based scan programme was used (TeleForm 10.0, ElectricPaper GmbH, Lüneburg/Germany). Data analysis was mainly descriptive and was conducted with SPSS Version 15.0. In the case of nominal and ordinal data, contingency tables were generated and Pearson's chi-square was computed. For continuous variables, Student's independent t-test was used for pair-wise comparisons. Binary logistic regression (method ENTER) was applied to model socio-demographic influences on the prevalence of IHSs and denial. To account for differences of the two regional samples, raw prevalences were adjusted for age and gender by direct standardization to the merged subgroups of SHI-insured respondents with medical consultations during the past 12 months (n = 1899).


Postal survey stage 1: screening

Sample characteristics

The screening samples of FR and HL were statistically comparable in terms of gender (total: 52.8% females; FR: 54.1% females; HL: 51.5% females; chi-square = 3.35, P = 0.067). However, the HL sample was significantly older than the FR sample (age in years; total: M = 47.4, SD = 16.5; FR: M = 45.2, SD = 16.5; HL: M = 49.5, SD = 16.3; T = –9.04, P < 0.001).

Non-response analysis

Figure 1 illustrates the study design and participants. Women (n = 1451, 56.1%) were more willing to participate than men (n = 1155, 49.9%; chi-square = 18.8, P < 0.001), and the respondents (M = 49.4, SD = 16.3) were significantly older than the non-respondents (M = 45.0, SD = 16.6; T = 9.4, P < 0.001). The participants in FR were significantly younger and more frequently females than the respondents in HL.

Figure 1

Participant flow. IHS = individual health service; SHI = statutory health service

Screening results

Further analyses concentrated on the 2120 SHI members only, and figure 1 and table 1 contain detailed characteristics and results.

No regional differences in the (raw) one-year and lifetime prevalence of IHSs and denial were found, even after adjustment for age and gender (standardized rates IHSs: one-year prevalence HL 41.3%, FR 42.3%; lifetime prevalence HL 51.7%, FR 53.3%; standardized rates denial: one-year prevalence HL 21.2%, FR 20.1%; lifetime prevalence HL 27.1%, FR 25.1%).

A total of 82 telephone interviews were conducted to estimate a potential non-response bias. Compared with the mailed survey, the interviewees were older and reported a better general health status, possibly in part due to a positive shift of self-assessment in telephone surveys as described in the literature.27 The one-year prevalence of IHSs was lower among the interviewees (30.4%) than in the mail respondents (41.7%), whereas the report of denial was comparable (interview: 19.6%; mail: 20.5%). Of the 210 respondents of the questionnaire on reasons for non-participation, 29.5% withdrew from the study because of missing experience with IHSs and denial.

Postal survey stage 2: details on IHSs and denial

Subsequently, 915 subjects were sent a second questionnaire gathering detailed information on IHSs and denial. The respondents were now younger (n = 690; M = 51.2, SD = 15.2) than the non-respondents (n = 225; M = 45.8, SD = 15.7; T = −4.52, P < 0.001), but they did not differ significantly in terms of gender (P = 0.272), region (P = 0.513) or school education (P = 0.514). Figure 1 shows details of the experiences.

IHSs (n = 601)

IHSs aimed at prevention and treatment of existing conditions and were mainly explained with missing coverage by SHIs (cf. table 2).

Dependent on the specific IHS, 71.7–76.6% would reuse the utilized IHSs, and 49.4–62.3% would recommend their IHSs to friends and relatives (analyses not shown). In our multivariate analysis, gender, school education and subjective health had no significant impact on the offer/demand of IHSs within the past 12 months. Higher occupational status (white-collar workers) and age (respondents aged >40 years) had a significant, but remarkably low influence. Denial of health services was the most relevant predictor for the experience of IHSs (cf. table 3).

View this table:
Table 3

Influence of socio-demographic variables on the occurrence of IHSs and denial of medical services

B (SE)Exp(B) (95% CI)
IHSs past 12 monthsa includedb
 Age (years)
        41–600.45 (0.15)1.56 (1.16–2.10)
        61–790.45 (0.17)1.57 (1.12–2.20)
        Female0.09 (0.20)1.10 (0.74–1.63)
 School education (years)
        10–0.15 (0.25)0.86 (0.53–1.40)
        ≥120.17 (0.23)1.18 (0.75–1.87)
 Occupational status
        Blue collar workerReference
        White-collar worker0.53 (0.17)1.69 (1.20–2.37)
        Self-employed0.28 (0.27)1.32 (0.79–2.22)
 Subjective health
        No chronic conditionReference
        At least one chronic condition0.14 (0.13)1.15 (0.90–1.48)
 Denial of medical services within past 12 months
        Yes1.17 (0.014)3.23 (2.46–4.24)
Denial of medical services past 12 monthsa includedcB (SE)Exp(B) (95%-CI)
 Age (years)
        41–600.11 (0.17)1.11 (0.80–1.55)
        61–79–0.47 (0.20)0.63 (0.42–0.94)
        Females–0.41 (0.32)0.67 (0.36–1.23)
 School education (years)
        10–0.27 (0.28)1.32 (0.76–2.28)
        ≥12–0.14 (0.29)0.87 (0.50–1.52)
 Occupational status
        Blue collar workerReference
        White-collar worker–0.20 (0.20)0.82 (0.55–1.21)
        Self-employed0.24 (0.30)1.27 (0.71–2.26)
 Subjective health
        No chronic conditionReference
        At least one chronic condition–0.05 (0.23)0.96 (0.61–1.49)
 Offer/demand of IHSs within past 12 months
        Yes1.18 (0.14)3.24 (2.47–4.25)
  • IHS = individual health service

  • a: Logistic regression; n = 1381 valid cases out of n = 1899 members of SHIs with medical consultations during the past 12 months.

  • b: Adjusted for interaction gender*school education Nagelkerke's R2 = 0.13.

  • c: Adjusted for interactions gender*health status and gender*school education; Nagelkerke's R2 = 0.12.

  • Relevant influence: printed in bold

Denial of medical services (n = 238)

As the most frequent explanations for denial, physicians mentioned limits of the SHIs benefits catalogue, whereas patients mainly suspected the tight budget of the physician. Only one-fourth of the patients was appreciative to the explanation of the physician. 43.3% were offered the denied service as an IHS, and about two-thirds of the concerned patients accepted the offer (cf. table 2). Socio-demographic variables had again no significant influence. The risk of denial was 3-fold higher among SHI members reporting offers/demand of IHSs within the past 12 months, compared with respondents without an IHS experience (cf. table 3).


IHSs and denial of medical services are frequently reported by SHI-members. No regional differences were found in the lifetime and the one-year prevalences. Contrary to other studies, socio-demographics seemed to play a minor role in the offer/demand of IHSs.

Operationalization of IHSs

The cognitive interviews revealed the difficulties of finding a comprehensive definition of IHSs. Obviously, the understanding of the IHSs concept depends on the former experience with these services. In a population-based survey, an operationalization as privately financed medical services seems to be most appropriate. Former surveys used other IHSs descriptions.26,28

Characteristics of participants

Comparing the regions, the social status of respondents differs clearly. The participants in FR were younger, higher educated, more often voluntarily health insured and more frequently engaged in white-collar jobs than the respondents in HL (cf. table 2).

Non-response bias

About 53% of the sample participated in the screening survey. To estimate a potential non-response bias at this stage, telephone interviews among non-respondents were conducted. The telephone interviews indicate a potential overestimation of the one-year prevalence of IHSs in the postal screening survey. We assume that persons who were experienced with IHSs were more willing to participate in the study. Similar results are known from back pain related epidemiological research.23 Furthermore, the possible overestimation of the IHSs prevalence is emphasised by a second strategy to estimate a potential non-response bias: in additional questionnaires on reasons of non-participation, 29.5% wished to withdraw from the study because of missing experience with IHSs and denial.

Prevalences and regional differences

No significant regional differences in the prevalence of IHSs and denial were found, even not after direct standardization to age and gender. A comparison of the results with other studies needs to consider the mode of data collection, since postal and telephone surveys might produce different results on the same issue.27 The lifetime prevalence of IHSs was about 52%. The one-year prevalence of IHSs was 42% in all statutory health insured, which seems relatively high compared with a telephone-based survey among 3006 (27%) SHI-insured persons in 2008.18 But a comparison of these data needs to consider the answering format: in our study, we first asked the participants to indicate, if they had consulted a physician during the past 12 months. Only in the case of an affirmative answer the questions on denial and IHSs during these consultations had to be filled in. In contrast, the other study did not include information on consultations; therefore, the one-year prevalence of IHSs might be lower due to SHI members who negated the question simply because they had not had any medical consultations.18 Even under the assumption that the proportion of respondents who had consulted a physician during the past 12 months resembled our study (88.9%; i.e. n = 2673 out of n = 3006), and restricting the calculation of the 1-year-prevalence of IHSs on the respondents with consultations, the prevalence within the telephone-based survey did not remarkably change (30%) and still undercuts our own result (42%). The lifetime prevalence of denial (26%) was lower than suggested in studies by a private health insurance (41%).21 Only 25% reported appreciation of the physicians’ reasons for the denial of medical services.

Aims and explanations of IHSs

IHSs mainly aimed at prevention and early diagnosis of diseases or treatment of existing conditions. IHSs were most often accounted for with the exclusion from the SHIs benefits catalogue or yet missing coverage by SHIs. About 5% of the respondents reported, IHSs were offered because the budget ceiling of the physician had been reached. It is legally not allowed to offer services as IHSs if they are included in the SHIs benefits catalogue (cf. table 2).

Explanations for denial and combined occurrence with IHSs

Physicians explained denial most often with the SHIs benefits catalogue (patient-reported data). In both the physicians’ and patients’ view, budget ceilings accounted for a large proportion of denial. More than 40% were later offered a denied service as an IHS, whereas a previous finding assumed a proportion of about one-third.24 More than three-fourths accepted the offered IHSs (cf. table 2).

Socio-demographic influences

Previous findings suggest a higher prevalence of IHSs among SHI members with higher income and higher education.18 We did not survey income as it showed to have low acceptance among respondents in our pre-test phase. Socio-demographic variables had a low influence on the offer/demand of IHSs (age ≥ 40 years, white-collar- workers; cf. table 3). However, due to limited variables available the regression model was far from being exhaustive (Nagelkerke's R2 0.12 and 0.13).

Conclusions for health policy

In general, privately paid services like IHSs in Germany offer the opportunity to top-up services by health insurances or NHSs. But the experiences from Germany show that unregulated markets like IHSs have potential hazards, mainly resulting from conflicts of interests of physicians that offer SHIs and IHSs in parallel. First, clearly defined differentiations of SHIs services and IHSs are needed, combined with independent, evidence-based, patient-centred information on their indication, benefits, risks, costs of medical services. Second, mandatory regulations to control the second healthcare market are required, e.g. in terms of quality assurance, documentation and transparency of charging and the regulation of the financial responsibility for consequences of IHSs (like additional diagnostics, false-positive test results, side-effects). Other countries have already implemented clearer regulations, especially on the issue of financial responsibility for side-effects, e.g. Norway.8

Future research

Further explorations should focus on details of IHSs and denial, which can hardly be covered in standardized postal surveys. Therefore, we will conduct focus groups to complement the survey findings. They are intended to enable an in-depth analysis of the patients’ experiences with and their appraisal of IHSs and the denial of medical services, the benefit of IHSs as well as potential ethical implications of denial and IHSs on physician–patient relationship, patients’ role perception, perception of the medical profession, the identification of useful medical methods and services, the understanding of medicine and its goals and the understanding of the German SHIs system.14 The integration of quantitative and qualitative findings will be condensed in recommendations on the handling of IHSs and denial for SHI-accredited physicians and their patients.

Supplementary data

Supplementary data are available at Eurpub online.


This study was supported by the German Medical Association [grant 06-72] and by the Institute for Social Medicine, University at Lübeck.

Conflicts of interest: None declared.

Key points

  • IHSs and the denial of medical services are frequently reported by German members of social health insurances.

  • The data partly confirm former findings on the prevalence of IHSs and add new insights on denial of medical services.

  • Applying multivariate analyses, socio-demographic variables had only a low impact on the occurrence of IHSs and denial of health services.


The authors would like to thank Christina Präcklein for her support in the arrangement of the cognitive interviews, Christel Zeuner for the data management and all respondents for their participation in the survey.


View Abstract