- Open Access
The effectiveness of German disease management programs (DMPs) in patients with type 2 diabetes mellitus and coronary heart disease: results from an observational longitudinal study
Diabetology & Metabolic Syndromevolume 7, Article number: 77 (2015)
Although the population-based German disease management programs (DMPs) for diabetes mellitus (DM) and coronary heart disease (CHD) are among the biggest worldwide, evidence on the effectiveness of these programs is still inconclusive or missing, particularly for high risk patients with comorbidities. The objective of this study was therefore to analyze the impact of DMPs on process and outcome parameters in patients with both, type 2 DM and CHD.
Analyses are based on two postal surveys of patients from the KORA myocardial infarction registry (southern Germany) with type 2 DM and on two postal validation studies with patients’ general physicians (2006, n = 312 and 2011, n = 212). The association between DMP enrollment (being enrolled in either DMP-DM or DMP-CHD) and guideline care (defined by several process indicators) at baseline (2006) and its development until follow-up (2011) was analyzed using logistic regression models accounting for the repeated measurements structure. The impact of DMP enrollment/guideline care on cumulated (quality-adjusted) life years ((QA)LYs) over a 4-year time horizon (2006–2010) was assessed using multiple linear regression methods. Logistic regression models were applied to analyze the association between DMP status and patient self-management at follow-up.
Being enrolled in a DMP was associated with better guideline care at baseline [OR = 2.3 (95 % CI 1.27–4.03)], but not at follow-up [OR = 0.80 (95 % CI 0.40–1.58); p value for time-interaction <0.01]. DMP enrollment was not significantly [+0.15 LYs (95 % CI –0.07, 0.37); +0.06 QALYs (95 % CI –0.15, 0.26)], but treatment according to guideline care significantly [+0.40 LYs (95 % CI 0.21–0.60); +0.28 QALYs (95 % CI 0.10–0.45)] associated with higher (quality-adjusted) survival over the 4-year follow-up period. DMP enrollees further reported a somewhat better self-management than patients not being enrolled into a DMP.
The results of this study concerning the effectiveness of DMPs in patients with DM and CHD are mixed, but are weakly in favor of DMPs. However, we found a clear positive impact of guideline care on quality adjusted survival in this patient group. The development of the association between DMP enrollment and guideline care over the follow-up time indicates some external effects, which should be the subject of further investigations.
Metabolic and cardiovascular diseases (CVD) are interrelated public health problems causing substantial economic pressure on health care systems [1, 2]. As the prevalence of diabetes mellitus (DM) is currently rising in Germany and in most other countries and as diabetes is a major risk factor for CVD, the burden of these diseases is expected to increase further [3–5]. Specifically, high risk individuals with co-existing cardiovascular and metabolic diseases need intensive treatment to prevent macro-vascular complications. Owing to their similar pathogenesis , treatment guidelines in both diabetes and coronary heart disease (CHD) focus on the management of CVD risk factors through modification of lifestyle habits and medication therapy as these strategies have been shown to improve patient outcomes [7–13]. In order to reinforce guideline care and to improve quality of care for chronically ill patients, in the years 2002/2003, disease management programs (DMPs) for DM and CHD were rolled out nationwide in Germany within the statutory health insurance (SHI) system, a system of non-profit health insurance companies insuring ~88 % of the population . In comparison to disease management programs in other countries which often focus on high risk patients and are heterogeneous in their structure, German DMPs are population-based approaches and characterized by a high degree of homogeneity. In Germany, the “Federal Joint Committee” (Gemeinsamer Bundesausschuss) which is responsible for decision-making and quality assurance in the inpatient and outpatient health care sector passes directives and content requirements for structured DMPs . All DMPs must obtain approval from the German “Federal Insurance Agency” (Bundesversicherungsamt), which controls the quality standards passed by the Federal Joint Committee [14, 16]. Within these programs, general practitioners (GPs) act as gatekeepers responsible for (voluntary) enrollment of eligible patients and coordination of care. The key contents of the DMP-DM and DMP-CHD are enforcement of medication therapy, enhanced patient activation and patient self-management education (e.g., diet, physical activity, smoking cessation), continuity of care according to current guidelines and the use of information technology systems for routine documentation/benchmarking . Details including the roles of GPs, patients and health insurances as well as reimbursement schemes have been described elsewhere [17, 18]. Although around 4.0 million patients with type 2 DM and 1.8 million patients with CHD are enrolled in the respective German DMPs, a scientifically sound evaluation of the programs has appeared to be difficult . Important clinical information is only routinely collected for DMP participants, but not for non-DMP patients, and running randomized trials is hardly possible, as by law patients have the right to be enrolled into the appropriate programs [19, 20]. Therefore, the evidence on the effectiveness of DMPs mainly relies on non-randomized group comparisons based on a limited subset of information from routine claims [17, 18, 21, 22] or population-based survey data [23–28]. These studies show mixed results on survival and better quality of care for patients enrolled in the respective DMPs. However, due to the population-based focus of German DMPs, no study has examined the effectiveness of DMPs in terms of process and outcome quality in multi-morbid high risk patients. In addition, although patient education is one of the key DMP contents [16, 29, 30], very little is known about the impact of these education programs on patient self-management.
This study aims to evaluate the effectiveness of German DMPs in patients with coexistence of diabetes and CHD using longitudinal follow-up data from a registry-based population (myocardial infraction registry). In detail, we analyzed the impact of DMP enrollment (either DMP-DM or DMP-CHD) on guideline care, the impact of DMP enrollment/guideline care on 4-year (quality-adjusted) survival, and on patient self-management.
Data for this study originated from the core documentation of the KORA (Cooperative Health Research in the Region of Augsburg) myocardial infarction (MI) registry and two postal questionnaire waves (2006 and 2011) in MI registry patients and their respective GPs. The KORA MI registry collects information on all cases of coronary deaths and non-fatal MIs of inhabitants aged 25–74 years in the city of Augsburg and two surrounding counties . Of the 2950 participants who were alive in 2006 (n = 4394, response ~67 %) and answered an initial postal survey (patient survey 1) assessing basic information on socioeconomic status, lifestyle habits, risk factors, medication, and quality of life (QoL), 2563 were insured in the SHI system at the date of MI. As DMP names vary considerably between statutory health insurance companies and may not reveal the disease management aspect, the information about the DMP status was assessed from the patients’ GPs. For this, all GPs of statutorily insured participants who reported DMP-CHD enrollment (n = 665) and a subsample of the GPs of statutorily insured patients who denied DMP-CHD enrollment (n = 583) were asked to provide information on the DMP status of their patients (validation survey 1). The GPs of 1128 patients answered this questionnaire. For several reasons, 153 patients were excluded from further study (details in Fig. 1). Of the remaining 975 patients, 312 were identified as having type 2 diabetes. Information on the diabetes status and the type of diabetes was extracted from the GP questionnaires. If the physicians’ information on diabetes status was missing, the information provided by the patient was used (sensitivity 80 %; specificity 99 %). Up to the end of 2010, 141 patients died (62 of those with type 2 diabetes) and 681 of the patients still alive answered a second postal questionnaire in 2011 (patient survey 2), assessing the same variables as in 2006, plus information on patient self-management behavior. As in 2006, GPs were contacted again in 2011, and information on diabetes and DMP status was retrieved for 524 patients through a postal questionnaire (validation survey 2). Of those 524 patients, 210 were identified as having type 2 diabetes; 144 already had a diagnosis in 2006 and 66 were incident cases between 2006 and 2011. Only patients with physician-validated information on DMP status (n = 312 in 2006 and n = 210 in 2011) were included in the analysis, as previous analyses have shown that the validity of patient self-reports on their DMP status is low . The detailed study design and patient selection are illustrated in Fig. 1. Data collection and analyses related to the registry have been approved by the Bavarian Ethics Committee.
Information about the patients’ DMP enrollment status was extracted from the GP questionnaire in 2006 and 2011. All patients in the analyzed sample were eligible for enrollment in both DMP-CHD and DMP-DM. Therefore, a patient was classified as being a DMP enrollee if enrollment in one or in both of these programs was reported.
Treatment of patients in the DMP-CHD and DMP-DM should be based on national treatment guidelines , which include among other criteria, counseling for the topics physical activity, diet, and smoking cessation and treatment with statins, platelet aggregation inhibitors (PAI), and antihypertensive drugs [either renin–angiotensin inhibitors (RAI) or beta blocker] [33, 34]. As in a previous study, we defined guideline care as being present if: (a) a patient had received medical advice over the last 12 months on at least two out of the three lifestyle topics (smoking, diet, exercise) and (b) reported the intake of a PAI and a statin and either RAI or a beta blocker over the last 7 days .
Quality of life, survival, and quality-adjusted survival
Information on the vitality status of patients was obtained by address search and by contacting the regional registration authorities. Survival times started at the fill-in date of the patient’s questionnaire in 2006 and ended at date of censoring or death. Quality of life was assessed in both patient questionnaires (2006 and 2011) using the EQ-5D instrument and utility values were calculated using a scoring algorithm derived from the general German population . Combining survival times and QoL, accumulated (quality-adjusted) life years ((QA)LYs), a measure that accounts for both the length and the quality of life over a certain observation period, were assessed individually for each person according to the methodology described by Laxy et al. .
Patient self-management was defined by means of previous established criteria [26, 37–40]. The binary indicators, comprised of regular moderate physical activity (≥5 h per week)  and regular (at least once a week) self-monitoring of blood glucose (SMBG), blood pressure (SMBP), and body weight (SMBW), were extracted from the patients’ self-reports [26, 37, 38]. Medication adherence was assessed according to the German version of the Morisky Medication Adherence Scale (MMAS-4 score <1) . As an indicator of patient knowledge, it was asked if patients knew about the term “HbA1c” as a performance measure for long-term glucose control. Furthermore, it was assessed whether patients participated in an education class for diabetes or blood pressure control.
Descriptive statistics of the samples’ characteristics in 2006 (n = 312) and 2011 (n = 210) and p-values (Chi square tests for categorical variables and t-tests for continuous variables) for the differences between DMP enrollees and non-enrollees and between patients receiving guideline vs. those not receiving guideline care are reported.
In order to test whether DMP enrollment is a predictor for receiving guideline care and to test how this association developed over time, we used a generalized estimating equation (GEE) model with a logit-link, which accounts for the partly repeated measurement structure of the sample and included an interaction term between DMP status (yes vs. no) and time (2006 vs. 2011). Linear regression and Cox proportional hazard regression was applied to test the impact of the DMP status/guideline care on QoL changes and mortality over the 4-year follow-up period (2006–2010). Differences in accumulated (quality-adjusted) life years over the follow-up period for DMP status/guideline care were tested through linear regression in order to obtain an absolute measure for the achieved health gains. Finally, we applied logistic regression to assess the cross-sectional association between DMP/guideline care and the patients’ self-management in 2011. A more sophisticated analysis of the association with self-management was not possible, as self-management was only assessed in the 2011 survey.
As part of our additional exploratory analyses, we further examined separately the impact of DMP-DM and DMP-CHD to capture potential differences in the effectiveness of these programs.
All models were adjusted for the covariates age (continuous), sex, educational status (primary education: ≤9 years of school; secondary and tertiary education: >9 years of school), smoking status (never smoker, ex-smoker, current smoker), weight status (normal weight: Body Mass Index (BMI) < 25; overweight: 25 ≤ BMI < 30; obese: BMI ≥ 30), number of re-infarctions in 2006 (continuous), treatment status (no medication, oral antidiabetic medication only, insulin therapy with or without oral antidiabetic medication) and diabetes duration (<5 years, 5≤ years <10 years, ≥10 years), which were extracted from self-reports of the surveys in 2006 and 2011 and from the core database of the MI registry. Linear regression models analyzing the difference in accumulated QALYs were also adjusted for QoL at baseline (2006). For each association, a separate regression model was built up including only the predictor of interest and the specified set of covariates.
Characteristics of the study population
The characteristics of the 2006 and 2011 sample are shown in Table 1. The mean age of the sample was 68.2 years in 2006 and 72.1 years in 2011. Mean diabetes duration (mean duration since the first registry-registered MI) averaged ~9 (~9) years in 2006 and ~10 (~14) years in 2011. As patients originate from a MI registry, almost 80 % of people were male. In 2011, 77 % (2006, 76 %) were enrolled in one of the two DMPs, and 61 % (2006, 62 %) were treated according to guideline care. DMP enrollees and non-enrollees were quite comparable but patients receiving guideline care were on average significantly younger and reported higher QoL values than those not receiving guideline care.
Association between DMP status and guideline care
Table 2 depicts the association between DMP status and guideline care and how this relationship changed from 2006 to 2011. In 2006, patients enrolled in one of the two DMPs were 2.3 times (95 % CI 1.27–4.03) more likely to receive guideline care than patients not enrolled in a DMP. However, although the portion of DMP enrollees receiving guideline care remained relatively stable until 2011, the portion of non-DMP enrollees increased substantially, resulting in a non-significant association between DMP enrollment and guideline care (OR = 0.80, 95 % CI 0.40–1.58). The interaction term for this effect was highly significant (p = 0.009). A similar pattern was observed for enrollment in DMP-CHD or DMP-DM only.
Association between DMP status/guideline care and (quality-adjusted) survival
Table 3 shows the relationship between DMP status/guideline care and QoL changes and mortality respectively. Effect estimates from the linear regression model show that neither DMP enrollment nor guideline care was significantly associated with QoL change over the follow-up period. The results from the Cox proportional hazard regression show that enrollment in a DMP was not significantly associated with mortality (HR = 0.80; 95 % CI 0.45–1.41). Separate analyses for DMP-CHD (HR = 0.74; 95 % CI 0.44–1.25) and DMP-DM (HR = 0.91; 95 % CI 0.54–1.54) also showed non-significant associations.
Treatment according to guideline care was also not associated with QoL change, but was a very strong predictor of mortality (HR = 0.27; 95 % CI 0.15–0.47). More detailed analyses on the single dimensions of guideline care are illustrated in Additional file 1: Appendix 1. All dimensions were associated with a significant or non-significant mortality risk reduction (advice diet, HR = 0.57, p = 0.12; advice exercise, HR = 0.65, p = 0.15; beta-blockers, HR = 0.27, p < 0.001; statins, HR = 0.40, p < 0.01; PAI, HR = 0.47, p = 0.05; RAI = 0.69, p = 0.10).
Table 4 shows the relationship between DMP status/guideline care and (quality-adjusted) survival. Enrollment in a DMP was not significantly associated with accumulated LYs (+0.15, 95 % CI −0.07 to 0.37) and QALYs (+0.06; 95 % CI −0.15 to 0.26). Additional analyses showed that DMP-CHD was (borderline) significantly associated with accumulated LYs (β = 0.21, 95 % CI 0.02–0.40) and QALYs (β = 0.16, 95 % CI 0.00–0.33). The discrepancy between the non-significant HR and the (borderline) significant absolute effect estimates (LYs, QALYs) for DMP-CHD status results mainly from a shorter survival time of people who die, which is not accounted for by the non-parametric Cox model.
Guideline care was strongly associated with longer survival (+0.40 LYs, 95 % CI 0.21–0.60) and quality-adjusted survival (+0.28 QALYs, 95 % CI 0.10–0.45) over a time frame of ~4 years. Detailed analyses showed, that particularly patients taking beta-blockers (+0.29 QALYs, 95 % CI 0.06–0.52), statins (+0.25 QALYs, 95 % CI 0.04–0.46), and PAI (+0.25 QALYs, 95 % CI 0.04–0.46) had a substantially higher quality-adjusted survival over the follow-up period.
Association between DMP status/guideline care and self-management
Figure 2 shows the association between DMP/treatment status and important dimensions of self-management. Overall, patients enrolled in a DMP reported better self-management (almost all OR >1); however, only patient knowledge (“knowing the term HbA1c”) (OR: 3.05, 95 % CI 1.13–8.20) was significantly associated with DMP enrollment. Separate analyses for DMP-DM and DMP-CHD showed a similar pattern. Overall, the uncertainty around effect estimates was large.
This is the first study in Germany and one of the first studies internationally evaluating the effect of structured DMPs in individuals with both DM and CHD, a patient group known to have a high risk of fatal and non-fatal cardiovascular events [29, 41]. In addition, to our knowledge it is the only study in Germany on DMPs which uses data after 2007 [26, 29]. Overall, we found mixed results for the association between DMP enrollment and patient outcomes in high risk patients, which are weakly in favor of DMPs. DMP enrollment was positively associated with guideline care at the beginning of the study (2006), but not any longer the end of the study (2011). Although guideline care at baseline (2006) was identified as a strong predictor for (quality-adjusted) survival, we found no evidence for an impact of DMP enrollment on (quality-adjusted) survival over a time period of 4 years. In contrast, DMPs were weakly associated with better patient self-management.
At baseline (2006) the likelihood of receiving guideline care was more than doubled for those patients enrolled in a DMP for CHD and/or DM. This is congruent with previous studies, which clearly have shown that DMPs are associated with improvements in process parameters such as medical examinations, lifestyle counseling, and medication for secondary prevention of CHD [26–28, 42]. Interestingly, although the prevalence of receiving guideline care in the group of DMP enrollees stayed constant or went down slightly, the prevalence of receiving guideline care in the non-DMP group in this study went up substantially from 2006 to 2011. This significant interaction between the effects of DMP and time could be an indicator of a spill-over effect. It is possible that, with the implementation of DMPs and corresponding physician education requirements, physicians started to apply the adopted standards to all treated patients. However, alternative explanations such as altered patterns of enrollment regarding patient characteristics (education, socioeconomic status) or other external effects might be possible, as well.
The results of our study further show that general DMP enrollment in high risk patients was weakly, but non-significantly associated with (quality-adjusted) survival. Previous studies based on claims data from the BARMER (n ~40,000, mortality: RR = 0.48) and the AOK (n ~4000, HR = 0.68) health insurances companies reported significant mortality risk reductions for general diabetes patients enrolled in a DMP-DM [18, 24]. However, our non-significant results are in line with a study by Stark et al. on all participating CHD patients (with and without diabetes) from the same registry-based data source, which showed a moderate but non-significant mortality risk reduction (n = 975, HR = 0.90) for patients being enrolled in a DMP-CHD . In Stark et al.’s study, guideline care was found to be a strong predictor of mortality in CHD patients (HR = 0.40). Our study shows that this association is even stronger in high risk CHD patients with metabolic comorbidity (HR = 0.28), resulting in a 0.28 QALYs longer quality-adjusted survival over a 4-year time frame. The main driver for this strong effect in our study was medication with beta-blockers, statins, and PAI indicating that secondary prevention according to current German national guidelines might prolong life expectancy at a high level of QoL and should therefore have a high priority in the disease management process. Beyond medication and regular check-ups, involvement and education of patients aiming to enhance self-management behavior are core contents of the DMPs [16, 17]. This study shows that at follow-up patients with diabetes and CHD who were enrolled in a DMP reported slightly better self-management; however, owing to a high magnitude of uncertainty, most of these associations were not significant. These findings contribute to the very limited literature on this topic. The only German study that analyzed the association between DMP enrollment and self-management was based on diabetes patients only and reported that DMP enrollees were more likely to have participated in a diabetes education class, however, behavioral aspects such as self-monitoring of feet, blood glucose, or weight were also not significantly improved . As previous research indicated that self-management is a crucial factor for intermediate and long-term health outcomes, future studies evaluating the effectiveness of DMPs should incorporate dimensions of self-management as separate outcome measures to improve the evidence on this important topic [38, 43].
Our exploratory analyses indicated that both programs, DMP-DM and DMP-CHD, were associated with slightly better outcomes in this high risk population. Although some effects were slightly stronger for DMP-CHD, there is not enough evidence to assume that these differences occurred systematically.
The strong association between guideline care and (quality-adjusted) survival and the diminishing association between DMP enrollment and guideline care over the follow-up time are of great (political) relevance when it comes to the discussion about the meaningfulness of German DMPs: given the assumption that positive effects of DMPs are translated through delivery of guideline care, it is likely that the positive impact of DMPs on survival that had been reported in studies on data before 2007 will have been attenuated in recent years. Contrarily, one can argue that in case of a potential treatment spill-over the implementation of DMPs was an effective strategy for improving health care and health outcomes on the population level. Due to its moderate sample size and potentially limited generalizability this study might not provide enough evidence to derive externally valid conclusions on this question, however, the results of this study can be seen as an important guidepost for future research. To answer the question about the overall effectiveness of German DMPs in normal and high-risk patients, larger studies with up to date data are needed. Linking registry-based or population-based survey data, comprising detailed information on behavioral and clinical parameters, with administrative claims data, comprising detailed information about delivered care, seems to be a promising approach in this context .
The strength of this study is its registry-based longitudinal design and the analysis of different disease management dimensions. The use of patient self-reports and physician-validated information on DMP status at baseline and follow-up made it possible to comprehensively analyze the association between DMP status, guideline care, mortality, and self-management over the follow-up period. In addition, parallel to classical Cox proportional hazard regression, we applied a method that allows the quantification of quality-adjusted survival based on individual-level longitudinal survey data . This approach has two advantages. First, it not only accounts for the length of life, but also for the health state people live in. This is particularly of interest in older multi-morbid populations where health gains (longer survival) are often achieved through intensive treatment regimens, which in turn might negatively affect QoL. Second, absolute measures for health gains are generally needed for the assessment of the cost-effectiveness of interventions. Taking the means and standard errors of published effects (QALYs) and a certain societal willingness to pay threshold allows one to roughly estimate up to which costs DMPs and guideline care would be considered to be cost-effective.
The major limitation of this study is its non-randomized observational design. Therefore, selection bias and, consequently, effects of reverse causation cannot be ruled out. One big source of selection bias might be the inclusion criteria of German DMPs themselves, in which it is stated that patients need to be willing to actively participate in the DMP . Consequently, it could be assumed that better educated and more active patients with fewer comorbidities were more likely to be enrolled in one of the DMPs. Previous studies have reported such selection effects. Although we could not observe such effects in our data, unobserved selection bias could have occurred resulting in overoptimistic effect estimates [22, 28, 45]. Also, the strong association between guideline care and mortality might be partly attributable to selection processes. It is plausible that severely ill patients with low life expectancy or other severe illnesses such as cancer are treated less “aggressively” with medication for the secondary prevention of CHD. In contrast, we also have some indication for dilution of the observed positive effects: We do know that many of the patients initially not enrolled in a DMP were subsequently enrolled after the start of the study . This treatment “crossing over”, for which we could not account in our models, as well as the potential spill-over effects that occurred with the implementation of DMPs are likely to have biased the mortality effects in the direction of the null. Despite these limitations, non-randomized evaluations like this can be considered as the best available evidence, as randomized designs are hardly possible within the given legal framework . Another limitation of the study is its potentially limited power. Although we observed considerable effect sizes for many of the analyzed associations, tests statistics were often not significant.
The German DMPs for DM and CHD are among the biggest programs worldwide, but evidence of effectiveness is weak. This study indicates mixed results, speaking weakly in favor of DMPs. However, treatment according to guideline care seems to be highly effective in this patient group. The development of the DMP-guideline care association over time indicates external effects, which should be subject of further investigations. More sophisticated and larger studies are needed to draw more reliable conclusions about the effectiveness of these programs for patients with diabetes and CHD.
Narayan KM, Gregg EW, Fagot-Campagna A, Engelgau MM, Vinicor F. Diabetes—a common, growing, serious, costly, and potentially preventable public health problem. Diabetes Res Clin Pract. 2000;50(Suppl 2):S77–84.
Leal J, Luengo-Fernandez R, Gray A, Petersen S, Rayner M. Economic burden of cardiovascular diseases in the enlarged European Union. Eur Heart J. 2006;27(13):1610–9.
Wild S, Roglic G, Green A, Sicree R, King H. Global prevalence of diabetes: estimates for the year 2000 and projections for 2030. Diabetes Care. 2004;27(5):1047–53.
Whiting DR, Guariguata L, Weil C, Shaw J. IDF diabetes atlas: global estimates of the prevalence of diabetes for 2011 and 2030. Diabetes Res Clin Pract. 2011;94(3):311–21.
Dokken BB. The pathophysiology of cardiovascular disease and diabetes: beyond blood pressure and lipids. Diabetes Spectr. 2008;21(3):160–5.
Grundy SM, Benjamin IJ, Burke GL, Chait A, Eckel RH, Howard BV, Mitch W, Smith SC, Sowers JR. Diabetes and cardiovascular disease: a statement for healthcare professionals from the American Heart Association. Circulation. 1999;100(10):1134–46.
Knight K, Badamgarav E, Henning JM, Hasselblad V, Gano AD Jr, Ofman JJ, Weingarten SR. A systematic review of diabetes disease management programs. Am J Manag Care. 2005;11(4):242–50.
Norris SL, Nichols PJ, Caspersen CJ, Glasgow RE, Engelgau MM, Jack L Jr, Isham G, Snyder SR, Carande-Kulis VG, Garfield S, et al. The effectiveness of disease and case management for people with diabetes: a systematic review. Am J Prev Med. 2002;22(4, Supplement 1):15–38.
Khunti K, Stone M, Paul S, Baines J, Gisborne L, Farooqi A, Luan X, Squire I. Disease management programme for secondary prevention of coronary heart disease and heart failure in primary care: a cluster randomised controlled trial. Heart. 2007;93(11):1398–405.
Egginton JS, Ridgeway JL, Shah ND, Balasubramaniam S, Emmanuel JR, Prokop LJ, Montori VM, Murad MH. Care management for type 2 diabetes in the United States: a systematic review and meta-analysis. BMC Health Serv Res. 2012;12:72.
Stellefson M, Dipnarine K, Stopka C. The chronic care model and diabetes management in US primary care settings: a systematic review. Prev Chronic Dis. 2013;10:E26.
JBS 2. Joint British Societies’ guidelines on prevention of cardiovascular disease in clinical practice. Heart. 2005;91(Suppl 5):v1–52.
American Diabetes Association. Standards of medical care in diabetes—2013. Diabetes Care. 2013;36(Suppl 1):S11–66.
Busse R. Disease management programs in Germany’s statutory health insurance system. Health Aff (Project Hope). 2004;23(3):56–67.
Ferderal Joint Committtee (Gemeinsamer Bundesausschuss). Directives regarding disease management programmes. www.g-ba.de/downloads/62-492-353/2009-07-01-RSAV20.pdf. Accessed 24 Mar 2015.
Federal Insurance Agency (Bundesversicherungsamt). Approval of DMPs. http://www.bundesversicherungsamt.de/weitere-themen/disease-management-programme/zulassung-disease-management-programme-dmp.html. Accessed 30 Sep 2014.
Drabik A, Büscher G, Sawicki PT, Thomas K, Graf C, Müller D, Stock S. Life prolonging of disease management programs in patients with type 2 diabetes is cost-effective. Diabetes Res Clin Pract. 2012;95(2):194–200.
Stock S, Drabik A, Buscher G, Graf C, Ullrich W, Gerber A, Lauterbach KW, Lungen M. German diabetes management programs improve quality of care and curb costs. Health Aff (Project Hope). 2010;29(12):2197–205.
Drabik A, Graf C, Büscher G, Stock S. Evaluation der Effektivität eines Disease Management Programms Diabetes Mellitus in der GKV - Erste Ergebnisse und methodische Überlegungen. Zeitschrift für Evidenz, Fortbildung und Qualität im Gesundheitswesen. 2012;106(9):649–55.
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 (Bundesverband der Ärzte des Öffentlichen Gesundheitsdienstes (Germany)). 2012;74(8–9):496–501.
Linder R, Ahrens S, Koppel D, Heilmann T, Verheyen F. The benefit and efficiency of the disease management program for type 2 diabetes. Deutsches Ärzteblatt Int. 2011;108(10):155–62.
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. Gesundh ökon Qual manag. 2011;16(04):209–15.
Elkeles T, Heinze S, Eifel R. Health care by a DMP for diabetes mellitus type 2—results of a survey of participating insurance customers of a health insurance company in Germany. J Public Health. 2007;15(6):473–80.
Miksch A, Laux G, Ose D, Joos S, Campbell S, Riens B, Szecsenyi J. Is there a survival benefit within a German primary care-based disease management program? Am J Manag Care. 2010;16(1):49–54.
Schafer I, Kuver C, Gedrose B, Hoffmann F, RuSZ-Thiel B, Brose H-P, van den Bussche H, Kaduszkiewicz H. The disease management program for type 2 diabetes in Germany enhances process quality of diabetes care—a follow-up survey of patient’s experiences. BMC Health Serv Res. 2010;10(1):55.
Stark RG, Schunk MV, Meisinger C, Rathmann W, Leidl R, Holle R. Medical care of type 2 diabetes in German disease management programmes: a population-based evaluation. Diabetes Metab Res Rev. 2011;27(4):383–91.
Stark R, Kirchberger I, Hunger M, Heier M, Leidl R, von Scheidt W, Meisinger C, Holle R. Improving care of post-infarct patients: effects of disease management programmes and care according to international guidelines. Clin Res Cardiol. 2014;103(3):237–45.
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 (Amsterdam, Netherlands). 2008;88(2–3):176–85.
Fuchs S, Henschke C, Blumel M, Busse R. Disease management programs for type 2 diabetes in Germany: a systematic literature review evaluating effectiveness. Deutsches Ärzteblatt Int. 2014;111(26):453–63.
‘Feder Insurance Agency’ (Bundesversicherungsamt). Education programs for patients with diabetes (Schulungsprogramme für Patientinnen und Patienten). http://www.bundesversicherungsamt.de/fileadmin/redaktion/DMP-Veranstaltungen/Leitfaden-Anhang_2a.pdf. Accessed 27 July 2015.
Lowel H, Meisinger C, Heier M, Hormann A. The population-based acute myocardial infarction (AMI) registry of the MONICA/KORA study region of Augsburg. Gesundheitswesen (Bundesverband der Ärzte des Öffentlichen Gesundheitsdienstes (Germany)). 2005;67(Suppl 1):S31–7.
Stark R, Schunk M, Leidl R, Meisinger C, Holle R. Process evaluation of disease management programs for type 2 diabetes based on a representative population in the area of Augsburg (KORA). Betriebswirtschaftliche Forsch Praxis. 2009;3:283–301.
The German Medical Association (Bundesärztekammer); The National Association of Statutory Health Insurance Physicians (Kassenärztliche Bundesvereinigung); The Association of the Scientific Medical Societies in Germany (Arbeitsgemeinschaft der Wissenschaftlichen Medizinischen Fachgesellschaften). National guidelines for the treatment of type-2-diabetes—long version. http://www.leitlinien.de/mdb/downloads/nvl/diabetes-mellitus/dm-therapie-1aufl-vers4-lang.pdf. Accessed 30 July 2015.
The German Medical Association (Bundesärztekammer); The National Association of Statutory Health Insurance Physicians (Kassenärztliche Bundesvereinigung); The Association of the Scientific Medical Societies in Germany (Arbeitsgemeinschaft der Wissenschaftlichen Medizinischen Fachgesellschaften). National guidelines for the treatment of chronic coronary heart disease—long version. http://www.leitlinien.de/mdb/downloads/nvl/khk/khk-3aufl-vers1-lang.pdf. Accessed 30 July 2015.
Greiner W, Claes C, Busschbach JJ, von der Schulenburg JM. Validating the EQ-5D with time trade off for the German population. Eur J Health Econ. 2005;6(2):124–30.
Laxy M, Hunger M, Thorand B, Meisinger C, Kirchberger I, Holle R. The burden of diabetes mellitus in patients with coronary heart disease: a methodological approach to assess quality adjusted life years (QALY) based on individual-level longitudinal survey data. Value Health. 2015;17:7(A494).
Schunk M, Schweikert B, Gapp O, Reitmeir P, Meisinger C, Mielck A, Holle R. Time trends in type 2 diabetes patients’ disease management and outcomes: evidence from two KORA surveys in Germany. Exp Clin Endocrinol Diabetes. 2009;117(2):88–94.
Laxy M, Mielck A, Hunger M, Schunk M, Meisinger C, Ruckert IM, Rathmann W, Holle R. The association between patient-reported self-management behavior, intermediate clinical outcomes, and mortality in patients with type 2 diabetes: results from the KORA-A study. Diabetes Care. 2014;37(6):1604–12.
Morisky DE, Green LW, Levine DM. Concurrent and predictive validity of a self-reported measure of medication adherence. Med Care. 1986;24(1):67–74.
How much physical activity do older adults need? Physical activity is essential to healthy aging. www.cdc.gov/physicalactivity/everyone/guidelines/olderadults.html. Accessed 24 Mar 2014.
Haffner SM, Lehto S, Ronnemaa T, Pyorala K, Laakso M. Mortality from coronary heart disease in subjects with type 2 diabetes and in nondiabetic subjects with and without prior myocardial infarction. N Engl J Med. 1998;339(4):229–34.
Elkeles T, Kirschner W, Graf C, Kellermann-Mühlhoff P. Health care in and outside a DMP for type 2 diabetes mellitus in Germany-results of an insurance customer survey focussing on differences in general education status. J Public Health. 2009;17(3):205–16.
Norris SL, Nichols PJ, Caspersen CJ, Glasgow RE, Engelgau MM, Jack L, Isham G, Snyder SR, Carande-Kulis VG, Garfield S, et al. The effectiveness of disease and case management for people with diabetes. A systematic review. Am J Prev Med. 2002;22(4 Suppl):15–38.
Swart E, Thomas D, March S, Salomon T, Knesebeck OV. Experience with the linkage of primary and secondary claims data in an intervention trial. Gesundheitswesen (Bundesverband der Ärzte des Öffentlichen Gesundheitsdienstes (Germany)). 2011;73(12):e126–32.
Schafer I, Kuver C, Gedrose B, von Leitner EC, Treszl A, Wegscheider K, van den Bussche H, Kaduszkiewicz H. Selection effects may account for better outcomes of the German Disease Management Program for type 2 diabetes. BMC Health Serv Res. 2010;10:351.
ML designed the concept for the paper, performed the statistical analysis, interpreted the data, and drafted the manuscript. RH was involved in the coordination of the study, co-designed the concept of the study, and commented on the manuscript. RS co-designed the concept of the study and commented on the manuscript. CM, IK, MH, and WvS were involved in collecting the data and commented on drafts of the manuscript. All authors read and approved the final manuscript.
The KORA research platform (KORA, Cooperative Research in the Region of Augsburg) was initiated and financed by the Helmholtz Zentrum München-German Research Center for Environmental Health, which is funded by the German Federal Ministry of Education and Research and by the State of Bavaria. Since the year 2000, the collection of myocardial infarction (MI) data has been co-financed by the German Federal Ministry of Health to provide population-based MI morbidity data for the official German Health Report (see www.gbe-bund.de). Steering partners of the MONICA/KORA Infarction Registry, Augsburg, include the KORA research platform, Helmholtz Zentrum München, and the Department of Internal Medicine I, Cardiology, Central Hospital of Augsburg. This work was further supported by the German Center for Diabetes Research, which is funded by the Federal Ministry of Education and Research (BMBF).
Compliance with ethical guidelines
Competing interests The authors declare that they have no competing interests.
Adjusted hazard ratios (HR) and adjusted mean differences on the association between dimensions of guideline care and mortality/QoL-change.
Adjusted odds ratios (OR) on the association between DMP status/guideline care and different dimensions of self-management.