Skip to main content
  • Research article
  • Open access
  • Published:

Selective enrollment in Disease Management Programs for coronary heart disease in Germany – An analysis based on cross-sectional survey and administrative claims data



In 2002, Disease Management Programs (DMPs) were introduced within the German healthcare system with the aim to increase the quality of chronic disease care. Due to the enrollment procedures, it can be assumed a) that only certain patients actively decide to enroll in a DMP and/or b) that only certain patients get the recommendation for DMP enrollment from their physician. How strong this assumed effect of self- and/or professional selection is, is still unclear.


We used data from a cross-sectional postal-survey linked on individual level with administrative claims data from a German sickness fund. The sample consisted of individuals suffering from coronary heart disease (CHD) who i) were either enrolled in the respective DMP or ii) fulfilled the disease related criteria for enrollment but were not enrolled. We applied multivariate logistic regression analyses to assess factors on patient level associated with DMP enrollment.


We included 7070 individuals in our analyses. Male sex, higher age and receiving old age pension, a higher Charlson Score and a diagnosis of type 2 diabetes increased the odds for DMP-CHD enrollment significantly. Individuals with a diagnosed myocardial infarction (MI) were also more likely to be enrolled in the DMP-CHD. We found a significant interaction effect for MI and sex, indicating that the association between MI and DMP enrollment is stronger for women than for men.


DMP-enrollees and non-enrollees differ in various factors. Studies analyzing the effectiveness of DMP-CHD should carefully take into account these group differences. Furthermore, the results suggest that the DMP-CHD assessed reaches men better than women.

Peer Review reports


In the year 2002, Disease Management Programs (DMPs) were introduced within the German healthcare system, initially for diabetes and breast cancer. The aim was to increase the quality of chronic disease care, based on the provision of coordinated and structured care, following clinical guidelines and by strengthening the patients’ self-care competencies [1].

The first DMPs for coronary heart disease (CHD) were introduced in 2003. By 2014, more than 1.7 million individuals were enrolled in a DMP-CHD [2]. DMPs for CHD aim specifically at: the reduction of mortality; the reduction of cardiovascular morbidity (a reduction of myocardial infarctions (MI) and a reduction in the development of heart failure); and an increase in quality of life [3]. Enrollment in the DMP-CHD is voluntary and without any costs for statutorily health insured. Patients can enroll in the DMP-CHD if they have a documented diagnosis of CHD and are willing to actively participate in the DMP [2].

Due to the enrollment procedures, it can be assumed a) that only certain patients actively decide to enroll in a DMP and/or b) that only certain patients get the recommendation for DMP enrollment from their physician. How strong the effect of self- and/or professional selection is, is still unclear. Comprehensive information on selective enrollment in DMPs is essential for research as well as policy making. First of all, it is important to know if certain population subgroups are not reached although they might benefit from DMPs. Secondly, selective enrollment might influence the results of DMP evaluation [4] and studies trying to evaluate DMPs face the difficulty of selecting suitable control groups [5]. This becomes more important when considering that 14 years after the introduction of DMPs, methodologically solid evidence on their effectiveness is still limited [6]. The mandatory program evaluation according to clause 137f, paragraph 4 in Book Five of Germany’s Social Security Code (SGB V), includes DMP enrollees only, thus a control group is missing and inferences on the effectiveness of DMPs cannot be made.

To overcome the assumed selection bias, many studies analyzing the effectiveness of DMPs apply propensity score matching [5, 7,8,9,10]. Yet, those studies often use routinely collected administrative claims data only, and thus, some factors potentially influencing DMP enrollment cannot be included. For instance, it is still rather unclear, how socio-economic factors influence DMP enrollment [11, 12]. On the other hand, analyses relying on survey data only, face the problem of incorrect or incomplete reporting [13].

Most studies on DMPs focused on the DMP for type 2 diabetes (DM2) [4,15,16,, 6, 9, 1417]. Little is known about the effects of selective enrollment within the DMP-CHD. To our knowledge, only three publications analyzed selective enrollment in the DMP-CHD. The study by Gapp et al. [18] used data from a MI register, which includes individuals with an acute MI living in the region of Augsburg, southern Germany. Eligible individuals were surveyed and patient characteristics were, if possible, cross-checked with registry data. The study indicated that the odds for DMP enrollment decreased with age and with the time since the last MI. Patients suffering additionally from DM2 were more likely enrolled than those with a cardiac event only [18]. The two publications by Bozorgmehr et al. [11, 12] used data from a cohort study conducted in the German federal state of Saarland. Study participants had either a diagnosis of MI and/or angina pectoris. The focus in both analyses was on inequities in DMP enrollment by education and regional deprivation. In Bozorgmehr et al. [11] the authors reported no horizontal inequities regarding educational attainment, but found that women living in less-deprived areas were significantly more likely to enroll in the DMP-CHD than women in deprived areas. In Bozorgmehr et al. [12] they reported that women generally are less likely to enroll in the DMP-CHD than men but found no effect for education.

The aim of our analyses is to add further knowledge on selective enrollment in the DMP-CHD by identifying factors on patient level that are associated with DMP enrollment. We try to overcome the discussed shortcomings by using a database of survey data linked with individual level claims data by a German sickness fund and by including chronically ill from all federal states of Germany.



Within the project RAC (Responsiveness in ambulatory care [19]), a German-wide postal survey with chronically ill individuals was conducted in October 2013. The sample consisted of persons having CHD who i) were either enrolled in the respective DMP or ii) fulfilled the disease related criteria for enrollment but were not enrolled. The sample was identified based on claims data by a German sickness fund (Techniker Krankenkasse (TK)); from both groups a random sample of 12,999 (not enrolled) and 13,000 (enrolled) individuals was drawn based on defined in- and exclusion criteria (see Table 1). All identified individuals were contacted via post and received the questionnaire, a letter of informed consent regarding the use of administrative claims data and additional informational material (e.g., on data protection issues). Data were only linked and included in the following analyses if both the questionnaire and the signed informed consent were returned. The final dataset comprises i) data from the 2013 postal survey that is linked on individual level to ii) claims data from the TK covering 18 months (from April 2012 to September 2013). Due to the fact that the sample is from one German sickness fund only, it has to be assumed that the results are not representative for the overall German population suffering from CHD. The sample selection, the linkage of survey and claims data as well as the operationalization of variables (including the used ICD-codes) are discussed in more detail in Röttger et al. [19].

Table 1 Inclusion and exclusion criteria


We used binary logistic regression analysis to assess factors associated with DMP enrollment. In the multivariate analysis we specified five models that varied by the included explanatory variables.

Variable description and rationale for variable selection

DMP enrollment: The status of DMP enrollment was generated from administrative claims data. Enrollment was defined as an individual being continuously enrolled in the DMP-CHD since at least 1.1.2012. Individuals within the group “not enrolled” were not enrolled in the DMP-CHD nor in any other DMP (e.g., for DM2) since at least 1.1.2012.

Socio-demographic and socio-economic factors: The following variables were included in all models: Sex, age, living area (urban vs. rural), net equivalent income, employment status and single household. In addition to these “objective” factors, we also included a variable indicating the subjective socio-economic status (subSES). The variables net equivalent income, employment status, single household and subSES were based on survey data. We calculated the net equivalent income by dividing the monthly net income by the number of household members, weighing the first member with 1 and all other family members with 0.5 [19].Footnote 1

Enrolled in integrated care program: Enrollment in an integrated care program (e.g., in a family physician care model) was assessed from claims data. In contrast to DMPs that primarily focus on the ambulatory care level, integrated care programs mainly aim at cross-sector patient care [20]. We assumed that participants of integrated care programs are more likely to be enrolled in a DMP for two reasons: a) they might be more positive about participating in structured care programs in general; and/or b) may have a GP who is more positive about special care programs in general and is thus more likely to promote DMP enrollment.

Co-Morbidity – generic: As previous studies had discussed that DMP-participants are healthier than non-participants [15], we included various measures of (co-) morbidity: the level of long term care entitlement, the Charlson Score and self-rated health on a scale from 0 (worst imaginable state) to 100 (perfect health). We calculated the ambulatory version of the Charlson Score [21] based on administrative claims data.Footnote 2 All diagnoses, on which variables are based, were identified through ICD-10-Codes from ambulatory and hospital care.

(Co-) Morbidity – disease specific: We included four disease specific morbidity measures: i) a diagnosis of DM2, ii) the level of disease severity, iii) a MI and iv) a congestive heart failure (CHF) diagnosis. DM2 was selected for two reasons: a) it is a frequent comorbidity of CHD and b) a DMP for DM2 exists and thus, persons suffering from CHD and DM2 have a higher chance of being informed about DMPs. The level of disease severity comprised three categories (lowest severity: CHD, other ischemic heart diseases; medium severity: angina pectoris, after MI and without a previous MI; highest severity: acute MI, unstable angina pectoris, other acute ischemic heart diseases). The classification was conducted based on the morbidity-categories and ICD diagnoses used by the German Federal (Social) Insurance Authority [22]. Additionally we included a variable indicating a MI (acute and history of MI) and a variable indicating CHF. Previous studies reported an association between MI and DMP enrollment [11]. Moreover, as the DMP for CHD is offered with an additional module for CHF [2], we assumed that DMP enrollment might be especially interesting for persons suffering from CHF.

A gender bias in the treatment of CHD is frequently discussed, with women and men being treated differently and a greater focus of attention on the treatment of men [23, 24]. We thus included interaction terms between sex and all CHD specific variables (MI, CHF, disease severity).

Incomplete data, missing values and plausibility checks

Cases with incomplete (only questionnaire or only administrative claims data) or implausible data (e.g., different information on age and sex in questionnaire and administrative data) were excluded from all analyses.

For enrollment in the DMP-CHD a diagnosis of CHD is required. Therefore we only included cases which had according to our data at least one diagnosis related to CHD. Cases without a diagnosis of CHD may have been in the initially selected sample for two reasons: a) for the sample selection a broader time frame was used and b) a rather sensitive approach was used to identify CHD patients (increasing the chance of “false-positive” cases).

We handled missing values in survey data by using multiple imputation. We imputed 15 datasets by fully conditional specification with multinomial logit models and predictive mean matching [25, 26]. All analyses were conducted with the original and the imputed dataset. Except for the descriptive statistics, we only report the results from the imputed datasets. For the logistic regression we report pooled estimates; for all analyses we report a significance level of p < 0.05. The multiple imputation was conducted in R (package MICE), the regression analysis in SPSS 22. All study participants gave written informed consent. The study was approved by the responsible ethics committee.


From 25,999 contacted persons, 8476 persons returned a filled in questionnaire, the signed consent form for data linkage and the linked cases passed the plausibility checks. These cases were included in a (non-) responder analysis, which revealed that DMP-participants, men and older study participants returned questionnaire and consent form more frequently (see Table 2). The response rates across sex and age groups were similar within the groups enrolled in DMP and not enrolled. However, the rise in response rates with age was steeper in the enrolled in DMP group compared to those not enrolled.

Table 2 Analysis (non-) responder

Of those 8476 that returned the questionnaire and the consent form, 1407 individuals did not fulfil the official requirement for DMP participation, as they had no relevant diagnosis related to CHD within the six quarters studied. Thus, 7070 cases were included in the following analyses. From those cases that had to be excluded due to not meeting the required diagnosis, 1377 were not enrolled in a DMP and only 30 were enrolled.

In the final sample of 7070 cases, 1092 had one or more missing values that were imputed and subsequently included in all further analysis. Eighty two percent of the imputed sample are male and the mean age is 71.4 (min: 25, max: 90, SD: 9.1) years. The majority of participants lives in a two person household, receives old-age pension, has a net equivalent income between 980€ and 1633€ (i.e., 60–100% of German median) and rates themselves as having a medium socio-economic status (see Table 3).

Table 3 Sample characteristics, divided by enrollment-status, reported in % (values after multiple imputation)

Persons enrolled in a DMP have more frequently a reported MI, CHF and a higher number of comorbidities according to the Charlson Score. Considering that all participants suffer from a chronic disease, the self-rated health (mean: 64.5; median: 70) is still rather good (the mean value of the German population was 77.4 in a representative survey in 2002/2003 [27]).

Multivariate analyses

In the multivariate analyses five binary logistic regression models with varying explanatory variables were specified. Due to the previous imputation of missing values, all models included the full sample of 7070 cases.

Socio-demographic and socio-economic factors: Male sex, higher age and receiving old age pension have higher odds for DMP enrollment in all five models (Table 4). While we found a slight association in Model 1 and 2 between the net equivalent income and DMP enrollment, the association is insignificant when controlling for co-morbidities.

Table 4 Results from multivariate analysis - dependent variable: DMP enrollment

Integrated care program: The enrollment in an integrated care program significantly increases the odds of being enrolled in a DMP (OR: 1.54–1.57, p < 0.05).

Co-Morbidity – generic: Individuals with higher Charlson Scores, i.e., those with at least two or those with more severe co-morbidities, are more likely to be enrolled. Yet, this effect vanishes for those with even more comorbidities (Charlson Score above 3).

(Co-) Morbidity – disease specific: Individuals with a diagnosed DM2 and individuals with a diagnosed MI have higher odds for DMP enrollment (OR: 2.65, p < 0.05; OR: 2.15, p < 0.05). Yet, a high disease severity according to the classification of the German Federal (Social) Insurance Authority significantly decreases the odds for DMP enrollment (OR: 0.49–0.65, p < 0.05).

Interaction effects: In addition to the single effects, we find a strong significant interaction effect between MI and sex. While a MI increases the odds of being enrolled in a DMP, the effect is stronger for women than for men. This effect is illustrated in Fig. 1, which depicts the conditional probability for DMP enrollment, by sex and MI.

Fig. 1
figure 1

Conditional probability for DMP enrollment by sex and MI

We conducted the same analyses with the dataset without imputed values using complete case analysis. In these analyses 5978 cases were included and the results were very similar to the results with the imputed dataset.


The aim of this analysis was to identify factors on patient level that are associated with DMP enrollment. Based on administrative claims data of a large German sickness fund, the two groups “DMP-enrollee” and “not enrolled, but fulfills criteria for DMP enrollment” were determined. From these two groups, a random sample each was drawn. Thus, assuming the enrollees and non-enrollees do not differ systematically regarding their characteristics, one could expect to find no differences between the two selected groups. At the same time, it can be assumed that differences between the groups indicate that these are either due to “selective enrollment” or that the DMP participation/non-participation led to these differences.

We found significant effects for age and sex, with higher age and a male sex increasing the odds for DMP enrollment. Our findings regarding sex are consistent with the results reported by Bozorgmehr et al. [12]. However, both findings are different to the study results by Gapp et al. [18]. The differences in the effect of age are probably due to the year of the survey/study (2006 vs. 2013). While the DMP-CHD was still rather new during the survey of Gapp et al., individuals suffering from CHD had have up to 9 years to enroll in the DMP at the point of our study. Differences considering the gender distribution (Gapp et al. did not find a significant association between sex and enrollment) are probably resulting from distinct CHD morbidities [18]. While the study by Gapp et al. [18] included exclusively individuals with a history of a MI, we find that differences between men and women are especially strong for those who did not have a diagnosed MI.

As assumed, the association between disease severity and DMP enrollment is affected by the patient’s sex: for women the association between MI and DMP enrollment is a lot stronger than the same association for men. Thus, it seems that women need a worse disease state then men, before they consider or are asked to enroll in the DMP-CHD. These results support existing studies focusing on gender differences in health care settings and particularly in the treatment of CHD [23, 28].

Further morbidity indicators are significantly associated with DMP enrollment: a higher level of disease severity decreased the odds for DMP enrollment whereas a diagnosed MI increased the odds for enrollment. Thus, the DMP-CHD seems to be especially interesting for patients with a low severity of CHD and for patients who suffered from a MI. In contrast, suffering from unstable angina pectoris (disease severity high) reduces the odds for DMP enrollment. This effect might be explained by the initial aims of the DMP, which is to help patients managing their disease and to avoid a worsening of the disease. Thus, those who are still in a lower disease severity might have a better chance to profit from the DMP enrollment than those with a high severity (e.g., patients that already suffer from unstable angina pectoris). In addition, the required “active participation” of the patients for DMP enrollment, might be a barrier for very ill patients.

The results indicate a strong association for DM2 and DMP-CHD enrollment. As previously discussed, this is probably due to the fact that a DMP program for DM2 exists, and thus, the chances to get to know about DMPs at all are higher for individuals having both conditions.

Employment status and DMP enrollment also correlate, with employed individuals being less likely to enroll in a DMP. A possible reason is that employed individuals cannot spend or do not want to spend as much time with their disease management as, for instance, individuals receiving old-age pension. This explanation is supported by studies showing that employed individuals also report to forgo care more often than others [29].

In the interpretation of our results several limitations have to be considered. First of all, although the study includes patients from all over Germany, all of them are insured at one sickness fund (TK). Hence, our sample is not representative of all CHD patients in Germany, and the generalizability of our results may be limited [30]. In addition, our results might be affected by differences in the survey-response behavior. It is possible that certain groups of patients did not participate in our study e.g., very ill individuals. However, we were able to analyze the (non-) response rates for age, sex and DMP enrollment. These revealed an overall higher response rate for individuals enrolled in a DMP, but the comparison of the response rates in relation to sex and age split by DMP enrollment status, showed rather similar results (e.g., women had in both groups a lower response rate than men).

We used cross-sectional survey data from October 2013 and administrative claims data from 01 April 2012 to 30 September 2013, while the date of DMP enrollment was any time before 01 January 2012. Thus, for some variables we do not know if they influence the DMP enrollment or if enrollment influences the occurrence of e.g., a certain diseases. Yet, it is rather unlikely that e.g., DMP enrollment leads to more MI in women. The interpretation that women are more frequently enrolled if they had a history of MI seems therefore legitimate. As in all analyses using administrative claims data, it has to be considered that the primary reason for data collection was different to the aim of the conducted analyses. Especially in the comparison of DMP-enrollees and non-enrollees it has to be considered that the medical documentation for patients within a DMP might be slightly better.


This is the first study that comprehensively analyses the association between patient characteristics and DMP-CHD enrollment for patients from all over Germany. Although this study focuses solely on selective enrollment within German DMPs, selective enrollment may also exist in other voluntary care programs (e.g., in the context of integrated care). As care programs address a predefined target population, some intended selection effects are desirable. The problem of adverse selective enrollment arises when the selection effects are not the defined target effects but disadvantage certain population groups.

The results indicate that DMP-enrollees and non-enrollees differ in various factors, e.g., sex, age, employment-status, (co-) morbidities and that selective enrollment may exist. Studies evaluating the DMP-CHD should carefully take into account the differences between both groups. The results indicate that the CHD assessed reaches men better than women. These findings needs further evaluation, as they suggest that the program either suits men better than women or that it is promoted to men more successfully than to women.


  1. We did not differentiate between adults and children under 14 years of age (as it is done in the modified OECD scale [31]), as the survey only included a question regarding the number of children under 18 years of age. But due to the high mean age of the sample, we assume a very small number of under 14 year old children in our sample.

  2. We excluded the diseases MI, CHF and DM2 from the calculation of the Charlson Score.



Coronary heart disease


Type 2 diabetes


Disease management program


Myocardial infarction


Subjective socio-economic status


Techniker Krankenkasse


Congestive heart failure


  1. Busse R. Disease management programs in Germany’s statutory health insurance system. Health Aff. 2004;23(3):56–67.

    Article  Google Scholar 

  2. German Federal (Social) Insurance Authority (2014): Zulassung der strukturierten Behandlungsprogramme (Disease Management Programme – DMP) durch das Bundesversicherungsamt (BVA), [Authorization of Disease Management Programs by the German Federal (Social) Insurance Authority] Updated December 2014. Accessed 4 Jan 2015.

  3. Bundesgesetzblatt Jahrgang 2009 Teil I Nr. 35, Zwanzigste Verordnung zur Änderung der Risikostruktur-Ausgleichsverordnung, [Twentieth regulation on alteration of the risk structure compensation ordinance]; Bonn, (20. RSA-ÄndV) 23. Juni 2009: 1542–69.

  4. Schäfer I, Küver C, Gedrose B, von Leitner EC, Treszl A, Wegscheider K, et al. Selection effects may account for better outcomes of the German Disease Management Program for type 2 diabetes. BMC Health Serv Res. 2010;10:351. doi:10.1186/1472-6963-10-351.

    Article  PubMed  PubMed Central  Google Scholar 

  5. Drabik A, Sawicki PT, Muller D, Passon A, Stock S. The methods within the evaluation of disease management programmes in control-group designs using the example of diabetes mellitus - a systematic literature review. Gesundheitswesen. 2012;74:496–501. doi:10.1055/s-0031-1301273.

    Article  CAS  PubMed  Google Scholar 

  6. Fuchs S, Henschke C, Blümel M, Busse R. Disease management programs for type 2 diabetes in Germany – a systematic literature review evaluating effectiveness. Dtsch Arztebl Int. 2014;111:453–63. doi:10.3238/arztebl.2014.0453.

    PubMed  PubMed Central  Google Scholar 

  7. Drabik A, Büscher G, Thomas K, Graf C, Müller D, Stock S. Patients with type 2 diabetes benefit from primary care-based disease management: a propensity score matched survival time analysis. Popul Health Manag. 2012;15:241–7. doi:10.1089/pop.2011.0063.

    Article  PubMed  Google Scholar 

  8. Linder R, Ahrens S, Köppel D, Heilmann T, Verheyen F. The benefit and efficiency of the disease management program for type 2 diabetes. Dtsch Arztebl Int. 2011;108:155–62. doi:10.3238/arztebl.2011.0155.

    PubMed  PubMed Central  Google Scholar 

  9. Nolting HD, Gottberg A, Schiffhorst G, Buhr S, Engel J. Einfluss der Teilnahme am DMP Diabetes mellitus Typ 2 auf die Entwicklung der Leistungsausgaben‐Ergebnisse einer retrospektiven kontrollierten Studie auf der Basis von GKV-Routinedaten. [Impact of a Disease Management Program for Diabetes on Health Care Costs – Results from a Retrospective Cohort Study with Matched Controls using Claims Data]. Gesundheitsökonomie Qualitätsmanagement. 2011;16:209–15. doi:10.1055/s-0029-1245920.

    Article  Google Scholar 

  10. Schulte T, Mund M, Hofmann L, Pimperl A, Dittmann B, Hildebrandt H. Pilotstudie zur Evaluation des DMP Koronare Herzkrankheit – Entwicklung einer Methodik und erste Ergebnisse [A pilot survey to evaluate the DMP for coronary heart disease – Development of a methodology and first results]. Z Evid Fortbild Qual Gesundheitswes. 2016;110–111:54–9. doi:10.1016/j.zefq.2015.11.003.

    Article  Google Scholar 

  11. Bozorgmehr K, Sebastian MS, Brenner H, Razum O, Maier W, Saum KU, et al. Analysing horizontal equity in enrolment in Disease Management Programmes for coronary heart disease in Germany 2008–2010. Int J Equity Health. 2015;14:28. doi:10.1186/s12939-015-0155-1.

    Article  PubMed  PubMed Central  Google Scholar 

  12. Bozorgmehr K, Maier W, Brenner H, Saum KU, Stock C, Miksch A et al. Social disparities in Disease Management Programmes for coronary heart disease in Germany: a cross-classified multilevel analysis. J Epidemiol Community Health. 2015. Online first. doi: 10.1136/jech-2014-204506.

  13. Hunger M, Schwarzkopf L, Heier M, Peters A, Holle R, KORA Study Group. Official statistics and claims data records indicate non-response and recall bias within survey-based estimates of health care utilization in the older population. BMC Health Serv Res. 2013;13:1. doi:10.1186/1472-6963-13-1.

    Article  PubMed  PubMed Central  Google Scholar 

  14. Berthold HK, Bestehorn KP, Jannowitz C, Krone W, Gouni-Berthold I. Disease management programs in type 2 diabetes: quality of care. Am J Manag Care. 2011;17:393–403.

    PubMed  Google Scholar 

  15. Dunkelberg S, Zingel D, Noack A, van den Bussche H, Kaduszkiewicz H. Welche Patienten werden (nicht) in das DMP Diabetes eingeschlossen? [Which Patients are (not) Included in the DMP Diabetes Programme?]. Gesundheitswesen. 2006;68:289–93. doi:10.1055/s-2006-926768.

    Article  CAS  PubMed  Google Scholar 

  16. Flamm M, Panisch S, Winkler H, Sönnichsen AC. Impact of a randomized control group on perceived effectiveness of a Disease Management Programme for diabetes type 2. Eur J Pub Health. 2012;22:625–9. doi:10.1093/eurpub/ckr147.

    Article  Google Scholar 

  17. Fullerton B, Erler A, Pohlmann B, Gerlach F. Predictors of dropout in the German disease management program for type 2 diabetes. BMC Health Serv Res. 2012;12:8. doi:10.1186/1472-6963-12-8.

    Article  PubMed  PubMed Central  Google Scholar 

  18. Gapp O, Schweikert B, Meisinger C, Holle R. Disease management programmes for patients with coronary heart disease—An empirical study of German programmes. Health Policy. 2008;88:176–85. doi:10.1016/j.healthpol.2008.03.009.

    Article  PubMed  Google Scholar 

  19. Röttger J, Blümel M, Engel S, Grenz-Farenholtz B, Fuchs S, Linder R, et al. Exploring health system responsiveness in ambulatory care and disease management and its relation to other dimensions of health system performance (RAC) – study design and methodology. Int J Health Policy Manag. 2015;7:431–7. doi:10.15171/ijhpm.2015.97.

    Article  Google Scholar 

  20. Busse R, Blümel M. Germany: health system review. Health Syst Transit. 2014;16(2):1–296.

    PubMed  Google Scholar 

  21. Charlson ME, Charlson RE, Peterson JC, Marinopoulos SS, Briggs WM, Hollenberg JP. The Charlson comorbidity index is adapted to predict costs of chronic disease in primary care patients. J Clin Epidemiol. 2008;61:1234–40. doi:10.1016/j.jclinepi.2008.01.006.

    Article  PubMed  Google Scholar 

  22. German Federal (Social) Insurance Authority (2012): Anlage 1 zu den Festlegungen nach § 31 Abs. 4 RSAV vom 28.09.2012 ICD-spezifische Gültigkeitskriterien zur Gruppierung von ICD-Kodes in Dx- (DXG) und (hierarchisierte) Morbiditätsgruppen ((H)MG) für das Ausgleichsjahr 2013 [Annex 1 to the requirements according to § 31 Abs. 4 RSAV from 28.09.2012 ICD specific criteria for the categorization of ICD-Codes in Dx (DXG) and (hiearchical) morbidity groups ((H)MG) for the year 2013], Accessed 27 Jul 2015.

  23. Kuhlmann E. Gender Mainstreaming in den Disease Management-Programmen – das Beispiel Koronare Herzerkrankung. [Gender mainstreaming in disease management programs – the example of coronary heart disease] Expertise im Auftrag der Bundeskoordination Frauengesundheit/des Arbeitskreises Frauengesundheit, Bremen: Zentrum für Sozialpolitik, Universität Bremen; 2003.

  24. Mittag O. Genderspezifische Aspekte in der Sekundärprävention der koronaren Herzerkrankung. [Gender specific aspects of secondary prevention in coronary heart disease]. Clin Res Cardiol Suppl. 2013;8:52–6. doi:10.1007/s11789-013-0053-2.

    Article  Google Scholar 

  25. Finch WH. Imputation methods for missing categorical questionnaire data: a comparison of approaches. J Data Sci. 2010;8:361–78.

    Google Scholar 

  26. White IR, Royston P, Wood AM. Multiple imputation using chained equations: Issues and guidance for practice. Stat Med. 2011;30:377–99. doi:10.1002/sim.4067.

    Article  PubMed  Google Scholar 

  27. König HH, Bernert S, Angermeyer MC. Gesundheitszustand der deutschen Bevölkerung: Ergebnisse einer repräsentativen Befragung mit dem EuroQol-Instrument. [Health Status of the German Population: Results of a Representative Survey Using the EuroQol Questionnaire]. Gesundheitswesen. 2005;67:173–82. doi:10.1055/s-2005-857991.

    Article  PubMed  Google Scholar 

  28. Babitsch B, Bormann C, Gohl D, Ciupitu-Plath CC. Gender and Utilization of Health Care. In: Janssen C, Swart E, von Lengerke T (eds). Health Care Utilization in Germany - Theory, Methodology, and Results. New York: Springer. 2014.

  29. Levesque JF, Pineault R, Hamel M, Roberge D, Kapetanakis C, Simard B, et al. Emerging organisational models of primary healthcare and unmet needs for care: insights from a population-based survey in Quebec province. BMC Fam Pract. 2012;13:66. doi:10.1186/1471-2296-13-66.

    Article  PubMed  PubMed Central  Google Scholar 

  30. Hoffmann F, Icks A. Unterschiede in der Versichertenstruktur von Krankenkassen und deren Auswirkungen für die Versorgungsforschung: Ergebnisse des Bertelsmann-Gesundheitsmonitors. [Structural Differences between Health Insurance Funds and their Impact on Health Services Research: Results from the Bertelsmann Health-Care Monitor]. Gesundheitswesen. 2011;74:291–7. doi:10.1055/s-0031-1275711.

    Article  PubMed  Google Scholar 

  31. Vos K, Zaidi MA. Equivalence scale sensitivity of poverty statistics for the member states of the European community. Rev Income Wealth. 1997;43:319–33. doi:10.1111/j.1475-4991.1997.tb00222.x.

    Article  Google Scholar 

Download references


The findings and results presented in this article were generated in the project “Exploring health system responsiveness in ambulatory care and disease management and its relation to other dimensions of health system performance (RAC)”. The project was conducted in cooperation with the Scientific Institute of TK for Benefit and Efficiency in Health Care (WINEG). The project was funded as part of the Berlin Centre for Health Economics Research by the German Federal Ministry of Education and Research (grant no. 01EH1202A).


The project was funded as part of the Berlin Centre for Health Economics Research by the German Federal Ministry of Education and Research (grant no. 01EH1202A). The funding organization had no role in study design, data collection and analysis, and writing the manuscript.

Availability of data and materials

Due to data protection regulations of the German sickness funds, the data cannot be made publicly available. Permission to access the sickness fund data was granted by the German Federal (Social) Insurance Office and the respective sickness fund, the Techniker Krankenkasse.

Authors’ contributions

JR wrote a first draft of the manuscript and had the main responsibility in the data analysis and interpretation. MB was involved in drafting the manuscript and in the interpretation of data. JR and MB had the main responsibility in the acquisition of data. RB revised the manuscript critically for important intellectual content. All authors made substantial contributions to conception and design of the study. All authors read and approved the final manuscript.

Competing interests

The authors declare that they have no competing interests.

Consent for publication

Not applicable.

Ethics approval and consent to participate

The study was approved by the ethics committee of the Technical University Berlin. All survey participants gave their written consent for participation and data linkage.

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Author information

Authors and Affiliations


Corresponding author

Correspondence to Julia Röttger.

Rights and permissions

Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (, which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver ( applies to the data made available in this article, unless otherwise stated.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Röttger, J., Blümel, M. & Busse, R. Selective enrollment in Disease Management Programs for coronary heart disease in Germany – An analysis based on cross-sectional survey and administrative claims data. BMC Health Serv Res 17, 246 (2017).

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: