- Open Access
Analysing horizontal equity in enrolment in Disease Management Programmes for coronary heart disease in Germany 2008–2010
© Bozorgmehr et al.; licensee BioMed Central. 2015
Received: 11 July 2014
Accepted: 24 February 2015
Published: 10 March 2015
Disease Management Programmes (DMPs) have been introduced in Germany ten years ago with the aim to improve effectiveness and equity of care, but little is known about the degree to which enrolment in the programme meets the principles of equity in health care. We aimed to analyse horizontal equity in DMP enrolment among patients with coronary heart disease (CHD).
Cross-sectional analysis of horizontal inequities in physician-reported enrolment in the DMP for CHD in a large population-based cohort-study in Germany (2008–2010). We calculated horizontal inequity indices (HII) and their 95% confidence intervals [95%CI] for predicted need-standardised DMP enrolment across two measures of socio-economic status (SES) (educational attainment, regional deprivation) stratified by sex. Need-standardised DMP enrolment was predicted in multi-level logistic regression models.
Among N = 1,280 individuals aged 55–84 years and diagnosed with CHD, DMP enrolment rates were 22.2% (women) and 35.0% (men). Education-related inequities in need-standardised DMP enrolment favoured groups with lower education, but HII estimates were not significant. Deprivation-related inequities among women significantly favoured groups with higher SES (HII = 0.086 [0.007 ; 0.165]. No such deprivation-related inequities were seen among men (HII = 0.014 [−0.048 ; 0.077]). Deprivation-related inequities across the whole population favoured groups with higher SES (HII estimates not significant).
Need-standardised DMP enrolment was fairly equitable across educational levels. Deprivation-related inequities in DMP enrolment favoured women living in less deprived areas relative to those living in areas with higher deprivation. Further research is needed to gain a better understanding of the mechanisms that contribute to deprivation-related horizontal inequities in DMP enrolment among women.
Coronary heart disease (CHD) is the leading cause of death and an important cause of morbidity world-wide [1,2] and in Germany . As part of a nation-wide attempt to restructure health care for CHD and other chronic diseases, Disease Management Programmes (DMPs) were introduced between 2002 and 2005 into the German statutory health insurance system (SHI). Their overall goals are to reduce morbidity, improve survival and quality of life of enrolled patients . To this end, DMPs aim to improve quality of care for the chronically ill by aligning service delivery with evidence-based guidelines and with a ‘managed care approach’ that has been shown to increase effectiveness of care . Another major rationale of the DMPs was to provide financial incentives for purchasers (i.e. sickness funds) and health service providers to care for the chronically ill. This would (so the assumption) reduce “cream-skimming”, and increase equity in the system  by allocating resources to sickness funds and providers who care for patients with higher needs.
In contrast to managed care approaches in other countries , German DMPs are tightly linked to sickness funds , and their core content (guideline-oriented care, quality indicators, coordinated transfer between different levels of care, recall for patients) is defined by a national expert group whose recommendations are binding for providers, while allowing for smaller differences among different contracting partners with respect to implementation . The first DMPs for CHD (DMP-CHD) were introduced in 2003. Ten years later, more than 1.7 million patients across Germany were enrolled in about 1,700 accredited DMPs for CHD . Participation in DMPs is voluntary for patients, but physicians have the mandate to enrol only “active patients” with respect to their therapy who can potentially benefit from the programme . Evidence suggests that patients enrolled in the DMP-CHD do not benefit in terms of lower mortality , but receive a better quality of care .
Previous research in the context of DMPs has mainly assessed programme effectiveness [9,11,12]. Despite concerns about selective enrolment favouring patients with higher socio-economic status (SES) , and despite considerably higher resources allocated to patients enrolled in the DMPs , no studies have yet analysed equity in DMP enrolment. The concern with selective enrolment is not only important for conclusions on programme effectiveness [11,13]. Inequitable enrolment in DMPs, i.e. selective enrolment based on socio-economic differences, could (theoretically) lead to intervention-generated inequalities , provided that the programme (or elements thereof) are effective in achieving their goals. We therefore aimed to assess if the principles of horizontal equity (equal treatment for equal need regardless of socio-economic factors ) are met in the context of the DMP-CHD.
Design and study population
We conducted a cross-sectional analysis using data of the “Epidemiological Study for the Prevention, Early Diagnosis and Optimal Treatment of Chronic Diseases in an Elderly Population” (ESTHER). This prospective (ongoing) cohort study includes non-institutionalized people from the general population living in the German federal state of Saarland, who were recruited by their general practitioners (GPs) during a general health check-up between 2000 and 2002 (t0). Baseline recruitment (t0) occurred before DMPs were introduced .
A total of 9,949 out of 12,000 invited individuals, aged 50 to 75 years, agreed to participate in the cohort. This sample is representative for the population of Saarland in the respective age range . Numerous GP-reported and patient-reported measures were captured in the GP-practice by questionnaires at baseline, and by postal questionnaires after two (t1), five (t2), eight (t3) and 11 (t4) years of follow-up [17-19]. For the purpose of this study, we focused on the population at t3 (2008–2010) since information on DMP enrolment was not captured in previous follow-ups. At t3, the cohort consisted of N = 7,012 survivors that were still physically and mentally able to participate (response rate 80.9%).
Inclusion and exclusion criteria
Measurement of horizontal (in-)equity in DMP enrolment and statistical analysis
We measured horizontal (in-)equity by estimating the concentration index and 95% confidence intervals [95%CI] for need-standardised enrolment, which has been referred to as the horizontal inequity index (HII) ranging from −1 to +1 . The HII, a measure of (in-)equity in health care use, can be obtained by a direct or an indirect standardisation method [20,21] in order to account for (legitimate) socio-economic differences in enrolment based on differences in need. As pointed out by O’Donnell et al., this procedure rests on the assumption that once observable need indicators have been controlled for, any residual variation in utilization is attributable to non-need factors .
A positive HII indicates a higher share of health care use by groups with higher SES than their share of need, i.e. horizontal inequities favouring the better-off. A negative value of HII represents horizontal inequities favouring groups with lower SES given their share of need. The HII is zero if health care use is equitably distributed across the socio-economic groups , i.e. if there is no inequality in the share of health care use given the share of need between respective socio-economic groups.
Enrolment in the DMP-CHD is voluntary. Patients who are diagnosed with CHD are eligible, provided that their GP offers DMP-CHD and considers the patient to be “active” and “likely to benefit from the programme”. The DMP-CHD consists of regular follow-up visits, need-based pro-active therapy, life-style counselling, psychosocial counselling, and standardised reference pathways to specialists and other levels of care. The programme goals are to reduce mortality and cardiovascular morbidity, prevent recurrent cardiovascular events and heart failure, and increase quality of life . In practice, patients with severe co-morbidities, or suffering from life-limiting conditions, or of very high age might be judged to be “not likely” to benefit from the programme in terms of the goals, as might those with healthy lifestyles and well-controlled risk factors. Therefore, our need variables should ideally account for potential differences in enrolment that are related to differences in co-morbidity and CHD risk factors.
To this end, we used age (three categories), sex (male/female), objective (Cumulative Illness Rating Scale for Geriatrics Severity Index, CIRS-G ) and subjective (self-rated health as dummy: “excellent, very good” vs. “fair, bad, very bad”) measures of morbidity simultaneously as need variables.
The CIRS-G is a comprehensive (physician-rated) assessment of 14 organ systems, reflecting not only the presence, but also the severity of any physical (co-)morbidity. Self-rated health (SRH) in turn is a global measure with high prognostic validity for morbidity and mortality, and is used in this study to reflect the psychosocial dimension of need for continuous and pro-active care as provided by the DMP.
Although further variables could theoretically be considered as reflecting “need to enrol” (e.g. smoking, body mass index etc.), we limited the analysis to the above-mentioned variables mainly because previous analyses in the cohort (unpublished) showed that there are no systematic differences in DMP-CHD enrolment based on life-style factors.
Following the indirect standardisation method with non-linear models [20,21], we obtained need-expected enrolment in the DMP-CHD (dummy) in multi-level logistic regression models (cross-classified models), which contained all need variables (age, sex, CIRS-G, SRH) and accounted for the simultaneous clustering of patients in both municipalities and in GP-practices. Details on the model specification are provided in the supplementary file along with regression coefficients, standard errors, and variance parameters obtained from the multi-level regression models (Additional file 1).
Non-need variables included in the analysis were the number of social contacts, i.e. family members/friends whom participants can count upon or discuss problems with (included as a proxy of loneliness in the elderly), and an immigration background (dummy), defined as having (i) a foreign nationality or (ii) a German nationality and a place of birth outside of Germany. The number of social contacts was initially operationalised as a variable with three categories (“0-1”, “2-4”, “5 and more”). It was collapsed to a binary variable (“0-1” vs. “2 and more”) for inclusion in the regression model to avoid collinearity. Non-need variables were set to their means in obtaining the predictions .
Inequality in need-standardised DMP enrolment
The predicted values obtained from the multi-level logistic regression models can be interpreted as need-expected enrolment, adjusted for non-need variables. Random effects (at the level of the cross-classification between municipalities and GP-practices) were set to zero in predicting need-expected enrolment. As explained in details elsewhere [20,21], need-standardised enrolment was obtained by subtracting need-expected enrolment from the observed actual enrolment.
In a next step, we calculated the concentration index for need-standardised DMP-CHD enrolment (HII) to assess inequalities across two measures of SES stratified by sex. First, the highest educational attainment at baseline, defined as an ordinal, individual-level SES variable with three levels contrasting low (level I - no formal degree or at least 9 years of schooling, Hauptschule), medium (level II – at least 10 years of schooling qualifying for professional training, Realschule/Mittlere Reife) and high (level III – at least 12 or 13 years of schooling qualifying for university entrance, Fachhochschulreife/Abitur).
Second, small-area regional deprivation of patients’ residential area using the German Index of Multiple Deprivation (GIMD) . The GIMD includes seven domains of area deprivation: income, employment, education, municipal revenue, social capital, environment, and security . Patients were assigned a distinct value of regional deprivation at municipality level by linking their ZIP codes (at t3) to one of 52 municipalities in Saarland. Municipalities were grouped into quintiles (Q1: lowest SES/highest deprivation – Q5: highest SES/lowest deprivation) in relation to all 9,620 municipalities in Germany.
When predicting the need-standardised enrolment for analysis across regional deprivation as SES indicator, we included educational attainment as non-need variable in the multi-level logistic regression model (Additional file 1). The education variable with three categories was collapsed to a binary variable (“level I” vs. “level II + II”) to avoid collinearity. The rationale for including education as non-need variable was to assess the inequity across regional deprivation while controlling for the effects of individual-level education and to reduce the possibility of omitted variable bias.
We plotted concentration curves for the HIIs across each SES indicator to illustrate the distribution of DMP-CHD enrolment across social groups. For descriptive purposes, DMP enrolment rates stratified by sex and socio-economic group were calculated in addition to sex-stratified absolute frequencies/proportions for all variables used in this study.
All analyses were performed using Stata® 12.1. The Distributive Analysis Stata Package (DASP)  with the commands “igini” and “clorenz” was used to calculate the concentration indices of need-standardised enrolment (HII) including standard errors and 95% CIs, and to plot the concentration curves respectively. Details on the calculation of standard errors by the DASP command “igini” (following personal communication with the authors of the package) are provided in Additional file 2. To assess the robustness of inferences we additionally estimated CIs at the 90% level for all HIIs.
Patients without ZIP codes (n = 14) or without GP-identifier (n = 25) were not included in the regression analysis. Missing data in covariables were treated as missing at random and a complete case analysis was performed.
Descriptive characteristics of included participants of the ESTHER study with coronary heart disease at the 8-year follow-up (2008–2010) (N = 1,280)
Freq. (col %)
Socio-economic status (SES)
Highest educational attainment*
Level I (lowest)
Level III (highest)
Regional deprivation of patients’ residential areas (GIMD)
Q1 (lowest SES/highest deprivation)
Q5 (highest SES/lowest deprivation)
Lifetime prevalence of physician-reported CHD defining morbidities/index diseases
Cumulative Illness Rating Scale for Geriatrics (CIRS-G) severity index**
DMP utilization (physician-reported)
Enrolled in DMP
Duration of enrolment in DMP
Social contacts: family members/friends whom participants can count upon/discuss problems with
5-10 and more
The proportion of women in lower age-groups (55–64) was 10 percentage-points lower compared to men, while the proportion of women was higher (7 percentage-points) in the highest age-group (75–84). The severity of (co-)morbidities was fairly balanced between both sexes, while a higher share of women (8 percentage-points more than men) rated their health status as “fair, bad, very bad” (Table 1).
Horizontal inequity indices
Horizontal inequity indices and 95% confidence intervals for enrolment in the Disease Management Program for coronary heart disease across highest educational attainment and regional deprivation by sex
Horizontal inequity index - HII (SE) [95% CI]
Highest educational attainment*
Regional deprivation of patients’ residential areas (GIMD)**
Regional deprivation-related inequities
Compared to the education-related inequities, horizontal inequities across regional deprivation showed a positive sign and thus favoured groups with higher SES. Deprivation-related inequities among women were significantly different from zero (HII = 0.086 [0.007; 0.165]), whereas no such differences were seen among men (HII = 0.014 [−0.048; 0.077]). Deprivation-related inequities were not significant across the whole population at the 95% level (HII = 0.036 [−0.012; 0.085]), but were marginally significant at the 90% level (Table 2).
This is the first study analysing horizontal equity in DMP enrolment among elderly people with CHD in Germany, taking into account individual need by means of both objective and subjective measurements while controlling for both non-need variables and the (simultaneous) clustering of patients in municipalities and GP-practices.
We found evidence for horizontal inequities favouring women living in less deprived municipalities relative to those living in municipalities with higher deprivation. The enrolment of women in the DMP-CHD did not correspond with their level of need, adjusted for differences in educational attainment and other non-need variables (social contacts, immigration background).
Our findings provide evidence that the principle of equal access or treatment for equal need is not fulfilled with respect to the systematic differences depending on small-area regional deprivation. Women in socially deprived areas are more disadvantaged than male patients with the same need. The degree to which “potential access” to the DMP-CHD (guaranteed by the SHI system) is converted into “realised access”  appears to be affected by regional deprivation and the (gender-specific) mechanisms underlying this association deserve further investigation.
A very positive finding is that DMP enrolment widely meets the principles of equity in health care as far as individual educational attainment (interpreted as an indicator for SES) is concerned. In this respect, primary care services in the federal state of Saarland produced an equitable distribution of health care services and resources.
This finding is consistent with two studies that assessed the possibility of selection effects in enrolment in DMP-CHD using self-reported  and physician-reported  enrolment status in an other German federal state. These studies [9,10] tested the hypothesis that there are no differences between enrolled and non-enrolled patients using a wide range of factors, including individual-level educational attainment. However, the comparability is limited because these studies [9,10] (and other studies focusing on the DMP Diabetes [12,13,28]) primarily assessed average “differences” across educational groups, and did not explicitly focus on equity or inequity in enrolment. This means that differences in need or morbidity were not taken into account when analysing potential education-related differences in enrolment.
Strengths and limitations
The major strength of our study is that we assessed equity in GP-reported DMP-CHD enrolment using individual-level data from a population-based cohort study. This allowed the use of appropriate measures of need, while controlling for relevant non-need factors. It might be argued that no need-standardisation would be necessary if the aim of DMPs is to include everyone diagnosed with a CHD. Reaching this ‘ideal’ might have been an intrinsic motivation to introduce DMPs, but insights from qualitative studies show that implementation of DMPs at GP-practice level is challenged by a lack of time and human resources, so that the programme is in reality not offered to everyone with a relevant condition . Consideration of some aspect of need thus remains crucial when trying to assess equity in enrolment, given that implicit judgements on eligibility and need are made by physicians in light of limited resources.
We used the CIRS-G severity index and SRH to operationalise need among CHD patients for a structured and intensified treatment of their condition in the scope of the DMP-CHD. We believe the variables are a valid reflection of need for enrolment in this context, but it is important to note that the legal regulation  does not explicitly formulate specific need criteria. It only combines formal eligibility criteria (CHD diagnosis) with vaguely defined criteria amenable to subjective judgements by physicians. Studies that operationalise need for treatment in the DMP-CHD differently might thus yield different results.
Previous studies on DMPs in Germany were not able to assess the possibility of socio-economic selection processes between participants and non-participants due to the limited availability of specific information on SES and/or lack of control groups [11-13,30]. It is also not possible to evaluate equity in enrolment by means of routine or claims data because of the lack of control groups (i.e. data on non-participants) and rudimentary information on individual covariables. The availability of individual-level educational attainment for both enrolled and non-enrolled patients, and the possibility of linking patients to area-based measures of SES (GIMD) allowed a comprehensive assessment of equity aspects in our study. This study thus adds to previous studies on selection effects in DMP enrolment [9,10,12,13,28] in using a comprehensive measure of SES such as the GIMD in addition to educational attainment. If previous studies used SES indicators to assess selective enrolment, then they relied on educational attainment alone, which is an arguably imperfect indicator of SES in elderly populations . The socio-economic variables were sometimes even treated as covariables to be adjusted for in the analysis [10,12]. In contrast to the above studies, this analysis focused explicitly on the socio-economic distribution of need-standardised propensity of DMP enrolment.
Another strength is the use of concentration indices (HII) to estimate (in-)equity in enrolment, which – to the best of our knowledge – has not been done in the context of health care use in Germany previously. This measure takes into account inequalities across the whole distribution of socio-economic groups, in contrast to conventional epidemiological methods (e.g. measures of association such as frequency ratios or odds ratios) which establish the presence of inequalities by comparing two or more socioeconomic groups without taking into account how individuals are distributed in each SES category . The use of the concentration index is hence recommended for comparisons between populations (e.g. from other federal states of Germany with different sizes of respective socio-economic groups), and over time  where differences in sizes of socio-economic groups might affect estimates derived from conventional epidemiological methods. Our estimates of (in-)equity in enrolment of the DMP-CHD can thus be used as baseline for monitoring change in (in-)equity over time e.g. in the ongoing ESTHER study or in other cohort studies with comparable data basis [9,10]. As such, our study contributes to attempts for improving the evaluation of German Disease Management Programmes which has been referred to as insufficient elsewhere .
Several methodological issues [34-36] should be noted in this context. First, our study analyses inequity in a binary outcome. As O’Donnell et al. point out, the concentration index is an appropriate measure of socio-economic health (care) inequality when health care is measured on a ratio scale with nonnegative values . In line with common approaches [34,38,39], we have therefore transformed the binary outcome to a ratio scale in using predicted probabilities as outcome.
The fact that predicted probabilities are bounded (i.e. have a range between zero and 100%) creates two other challenges: a) The range of the concentration index is linked to the mean of the variable [34,36], in our case the mean prevalence of DMP enrolment. b) The “mirror principle”  is not fulfilled, which means that inequalities may vary depending on whether we measure inequality in attainment (e.g. DMP enrolment) or in shortfall (e.g. non-enrolment in DMPs) . In other words: inequalities may vary depending on which attribute is coded 0 and 1 .
To circumvent challenge (a), a normalisation to the mean of the outcome has been proposed . The underlying value judgements of the normalised concentration index, however, lead to the fact that different pro-rich distributions yield the same value of inequality . In line with suggested options to deal with the “dilemma”  faced by the analyst, we decided to stick with the (unnormalized) concentration index but “be aware of the limitations of the analysis” (which may not arise unless one attempts to measure inequities in enrolment between federal states or over time where the average rate of enrolment varies strongly).
As for challenge (b), several options have been suggested to achieve the desired mirror principle . Others argue that the issue is of less concern if there is a convention on what to code “1” . Although there is no explicit convention in our case, all studies on DMPs in Germany have yet analysed “enrolment”. We therefore think that the mirror issue is less of a concern if future analyses on equity in DMP enrolment stick to the analysis of “enrolment” as outcome (instead of “non-enrolment”) when using concentration indices.
Another methodological issue worth mentioning is that we used the GIMD as ordinal variable in quintiles as is common practice in analyses using area-based deprivation measures . Although the literature on income-related health inequality usually uses the socio-economic ranking variable on a continuous scale, it is important to note that an ordinal scale is sufficient [38,43-45]. There are also examples where continuous measures of SES are transformed to an ordinal scale for use as ranking variable in order to calculate the concentration index based on quintiles , as done in our analysis with the GIMD.
Our findings regarding equity or inequity in DMP-CHD enrolment can be generalised to Saarland but not necessarily to Germany as a whole. This is mainly due to (i) the intra- and inter-regional heterogeneity to which DMPs are implemented across each federal state, and (ii) the non-significance or marginal statistical significance of some of the HII estimates in our study (Table 2). It should also be noted that our results provide insights into equity distributions in the respective time period, but that the situation might have changed since data for this study was collected. This underlines the importance of monitoring of equity aspects in health services utilisation on a routine basis.
Our findings are limited by the cross-sectional nature of the study and the possibility of attrition bias, because the DMP enrolment status of those who deceased or dropped out before t3 was not known. Furthermore, we had information only on current enrolment, which means that individuals who might have been enrolled before t3 but eventually opted out of the DMP until t3, were categorised as “non enrolled”. In performing a complete case analysis, we excluded 560 individuals from the calculation of HIIs, which might have affected our results.
Need-standardised DMP enrolment in the federal state of Saarland was fairly equitable across educational levels. The principles of horizontal equity, however, were not fulfilled for CHD patients - particularly women - since systematic differences existed between those living in deprived and less deprived municipalities. Deprivation-related inequities in DMP enrolment significantly favoured women living in less deprived areas relative to those in areas with higher deprivation. Further research (in form of decomposition analyses or qualitative approaches) is needed to gain a better understanding of the mechanisms that contribute to larger horizontal inequities in DMP-CHD enrolment among women. Our study, conducted as proof of concept, underlines the importance of analysing equity in DMP enrolment in Germany as primary research focus using specific measures of SES beyond the routine national programme evaluation.
The ESTHER study was approved by the ethics committees of the medical faculty of the University of Heidelberg and of the medical board of the state of Saarland.
The authors would like to thank the ESTHER study participants, the study physicians and the ESTHER study office in the state of Saarland, as well as the general practitioners of the state of Saarland for their very good and longstanding cooperation.
- Murray CJ, Vos T, Lozano R, Naghavi M, Flaxman AD, Michaud C, et al. Disability-adjusted life years (DALYs) for 291 diseases and injuries in 21 regions, 1990–2010: a systematic analysis for the Global Burden of Disease Study 2010. Lancet. 2012;380:2197–223.View ArticlePubMedGoogle Scholar
- Lozano R, Naghavi M, Foreman K, Lim S, Shibuya K, Aboyans V, et al. Global and regional mortality from 235 causes of death for 20 age groups in 1990 and 2010: a systematic analysis for the Global Burden of Disease Study 2010. Lancet. 2012;380:2095–128.View ArticlePubMedGoogle Scholar
- Federal Statistical Office (Destatis). [Todesursachen in Deutschland]. www.destatis.de Fachserie 12, Reihe 4. 2014. Wiesbaden, Statistisches Bundesamt. Gesundheit. 4/1/14.
- Joint Federal Commission (G-BA). [Zwanzigste Verordnung zur Änderung der Risikostruktur-Ausgleichsverordnung (20.RSA-ÄndV) Vom 23. Juni 2009]. Bundesgesetzblatt. 2009;I:1551–61.Google Scholar
- Szecsenyi J, Rosemann T, Joos S, Peters-Klimm F, Miksch A. German diabetes disease management programs are appropriate for restructuring care according to the chronic care model: an evaluation with the patient assessment of chronic illness care instrument. Diabetes Care. 2008;31:1150–4.View ArticlePubMedGoogle Scholar
- Busse R. Disease Management Programs In Germanys Statutory Health Insurance System. Health Affair. 2004;23:56–67.View ArticleGoogle Scholar
- Pilnick A, Dingwall R, Starkey K. Disease management: definitions, difficulties and future directions. Bull World Health Organ. 2001;79:755–63.PubMed CentralPubMedGoogle Scholar
- German Federal (Social) Insurance Office (BVA). [Zulassung der Disease Management Programme (DMP) durch das Bundesversicherungsamt (BVA)]. http://www.bundesversicherungsamt.de. 2014. 4/1/14.
- 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.View ArticlePubMedGoogle Scholar
- Stark R, Kirchberger I, Hunger M, Heier M, Leidl R, Scheidt W, et al. Improving care of post-infarct patients: effects of disease management programmes and care according to international guidelines. Clin Res Cardiol. 2014;103:237–45.View ArticlePubMedGoogle Scholar
- 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.View ArticlePubMedGoogle Scholar
- Miksch A, Laux G, Ose D, Joos S, Campbell S, Riens B, et al. Is there a survival benefit within a German primary care-based disease management program? Am J Manag Care. 2010;16:49–54.PubMedGoogle Scholar
- 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.View ArticlePubMed CentralPubMedGoogle Scholar
- Greß S, Focke A, Hessel F, Wasem J. Financial incentives for disease management programmes and integrated care in German social health insurance. Health Policy. 2006;78:295–305.View ArticleGoogle Scholar
- Tugwell P, De SD, Hawker G, Robinson V. Applying clinical epidemiological methods to health equity: the equity effectiveness loop. BMJ. 2006;332:358–61.View ArticlePubMed CentralPubMedGoogle Scholar
- Starfield B. Improving equity in health: a research agenda. Int J Health Serv. 2001;31:545–66.View ArticlePubMedGoogle Scholar
- Raum E, Rothenbacher D, Low M, Stegmaier C, Ziegler H, Brenner H. Changes of cardiovascular risk factors and their implications in subsequent birth cohorts of older adults in Germany: a life course approach. Eur J Cardiovasc Prev Rehabil. 2007;14:809–14.View ArticlePubMedGoogle Scholar
- Schöttker B, Haug U, Schomburg L, Köhrle J, Perna L, Müller H, et al. Strong associations of 25-hydroxyvitamin D concentrations with all-cause, cardiovascular, cancer, and respiratory disease mortality in a large cohort study. Am J Clin Nutr. 2013;97(4):782-93.Google Scholar
- Schöttker B, Herder C, Rothenbacher D, Roden M, Kolb H, Müller H, et al. Proinflammatory cytokines, adiponectin, and increased risk of primary cardiovascular events in diabetic patients with or without renal dysfunction: results from the ESTHER study. Diabetes Care. 2013;36:1703–11.View ArticlePubMed CentralPubMedGoogle Scholar
- Wagstaff A, Doorslaer E. Measuring and testing for inequity in the delivery of health care. J Hum Resour. 2000;35:716–33.View ArticleGoogle Scholar
- O'Donnell O, van Doorslaer E, Wagstaff A, Lindelow M. Measuring and Explaining Inequity in Health Service Delivery. In: Analyzing Health Equity Using Household Survey Data: A Guide to Techniques and their Implementation, vol. 1. Washington D.C.: World Bank; 2008. p. 177–85. http://siteresources.worldbank.org/INTPAH/Resources/Publications/459843-1195594469249/HealthEquityFINAL.pdf (Accessed: 27/02/2015).Google Scholar
- van Doorslaer E, Wagstaff A, van der Burg H, Christiansen T, De Graeve D, Duchesne I, et al. Equity in the delivery of health care in Europe and the US. J Health Econ. 2000;19:553–83.View ArticlePubMedGoogle Scholar
- Miller MD, Paradis CF, Houck PR, Mazumdar S, Stack JA, Rifai AH, et al. Rating chronic medical illness burden in geropsychiatric practice and research: application of the cumulative illness rating scale. Psychiat Res. 1992;41:237–48.View ArticleGoogle Scholar
- Maier W, Holle R, Hunger M, Peters A, Meisinger C, Greiser KH, et al. for the DIAB-CORE Consortium: The impact of regional deprivation and individual socio-economic status on the prevalence of Type 2 diabetes in Germany. A pooled analysis of five population-based studies. Diabet Med. 2013;30:e78–86.View ArticlePubMedGoogle Scholar
- Maier W, Fairburn J, Mielck A. Regional deprivation and mortality in Bavaria. Development of a community-based index of multiple deprivation. Gesundheitswesen. 2012;74:416–25.View ArticlePubMedGoogle Scholar
- Abdelkrim A, Duclos J-Y. DASP: Distributive Analysis Stata Package. 2007. PEP; World Bank; UNDP; Université Laval. http://dasp.ecn.ulaval.ca/index.html (Accessed: 27/02/2015).
- Aday LA, Andersen R. Equity of access to medical care: a conceptual and empirical overview. Med Care. 1981;19:4–27.View ArticleGoogle Scholar
- Elkeles T, Kirschner W, Graf C, Kellermann-Muehlhoff 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;17:205–16.View ArticleGoogle Scholar
- Miksch A, Trieschmann J, Ose D, Rölz A, Heiderhoff M, Szecsenyi J. General practitioners’ opinion and attitude towards DMPs and the change in practice routines to implement the DMP “Diabetes mellitus Type 2”. Z Evid Fortbild Qual Gesundhwes. 2011;105:427–33.View ArticlePubMedGoogle Scholar
- Bestehorn K, Jannowitz C, Karmann B, Pittrow D, Kirch W. Characteristics, management and attainment of lipid target levels in diabetic and cardiac patients enrolled in Disease Management Program versus those in routine care: LUTZ registry. BMC Public Health. 2009;9:280.View ArticlePubMed CentralPubMedGoogle Scholar
- Galobardes B, Shaw M, Lawlor DA, Lynch JW, Davey SG. Indicators of socioeconomic position (part 1). J Epidemiol Community Health. 2006;60:7–12.View ArticlePubMed CentralPubMedGoogle Scholar
- Regidor E. Measures of health inequalities: part 1. J Epidemiol Community Health. 2004;58:858–61.View ArticlePubMed CentralPubMedGoogle Scholar
- Blümel M, Busse R. Disease Management Programs - Time to evaluate. Health Policy Monitor 2009. Available at: http://www.hpm.org/de/Surveys/TU_Berlin_-_D/13/Disease_Management_Programs___8211__Time_to_evaluate.html (Accessed 27/02/2015).
- Wagstaff A. The concentration index of a binary outcome revisited. Health Econ. 2011;20:1155–60.View ArticlePubMedGoogle Scholar
- Erreygers G, Van OT. Putting the cart before the horse. A comment on Wagstaff on inequality measurement in the presence of binary variables. Health Econ. 2011;20:1161–5.View ArticlePubMed CentralPubMedGoogle Scholar
- Wagstaff A. Reply to Guido Erreygers and Tom Van Ourti‘s comment on ’The concentration index of a binary outcome revisited’. Health Econ. 2011;20:1166–8.View ArticlePubMedGoogle Scholar
- O'Donnell O, van Doorslaer E, Wagstaff A, Lindelow M. The Concentration Index. In: Analyzing Health Equity Using Household Survey Data: A Guide to Techniques and their Implementation, vol. 1. Washington D.C.: World Bank; 2008. p. 95–108.Google Scholar
- Erreygers G, Van OT. Measuring socioeconomic inequality in health, health care and health financing by means of rank-dependent indices: a recipe for good practice. J Health Econ. 2011;30:685–94.View ArticlePubMed CentralPubMedGoogle Scholar
- Hosseinpoor AR, van Doorslaer E, Speybroek N, Naghavi M, Mohammad K, Majdzadeh R, et al. Decomposing socio-economic inequality in infant mortality in Iran. Int J Epidemiol. 2006;35:1211–9.View ArticlePubMedGoogle Scholar
- Erreygers G. Correcting the concentration index. J Health Econ. 2009;28:504–15.View ArticlePubMedGoogle Scholar
- Wagstaff A. The bounds of the concentration index when the variable of interest is binary, with an application to immunization inequality. Health Econ. 2005;14:429–32.View ArticlePubMedGoogle Scholar
- Noble M, Wright G, Smith G, Dibben C. Measuring multiple deprivation at the small-area level. Environ Plann A. 2006;38:169–85.View ArticleGoogle Scholar
- O’Donnell O. Measuring health inequalities in Europe: Methodological issues in the analysis of survey data. Eurohealth. 2009;15:10–4.Google Scholar
- O'Donnell O, van Doorslaer E, Wagstaff A, Lindelow M. Data for Health Equity Analysis: Requirements, Sources and Sample Design. In: Analyzing Health Equity Using Household Survey Data: A Guide to Techniques and their Implementation, vol. 1. Washington D.C.: World Bank; 2008. p. 13–27.Google Scholar
- Wagstaff A, Watanabe N. What difference does the choice of SES make in health inequality measurement? Health Econ. 2003;12:885–90.View ArticlePubMedGoogle Scholar
- Monteiro CA, Benicio MH, Conde WL, Konno S, Lovadino AL, Barros AJ, et al. Narrowing socioeconomic inequality in child stunting: the Brazilian experience, 1974–2007. Bull World Health Organ. 2010;88:305–11.View ArticlePubMed CentralPubMedGoogle Scholar
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.