Social disparities in Disease Management Programmes for coronary heart disease in Germany: a cross-classified multilevel analysis

Disaggregating level-specific effects in cross-classified multilevel models

In psychology and other fields, data often have a cross-classified structure, whereby observations are nested within multiple types of non-hierarchical clusters (e.g., repeated measures cross-classified by persons and stimuli). This paper discusses ways that, in cross-classified multilevel models, slopes of lower-level predictors can implicitly reflect an ambiguous blend of multiple effects (for instance, a purely observation-level effect as well as a unique between-cluster effect for each type of cluster). The possibility of conflating multiple effects of lower-level predictors is well recognized for non-cross-classified multilevel models, but has not been fully discussed or clarified for cross-classified contexts. Consequently, in published cross-classified modeling applications, this possibility is almost always ignored, and researchers routinely specify models that conflate multiple effects. In this paper, we show why this common practice can be problematic, and show how to disaggregate level-specific effects in cross-classified models. We provide a novel suite of options that include fully cluster-mean-centered, partially cluster-mean-centered, and contextual effect models, each of which provides a unique interpretation of model parameters. We further clarify how to avoid both fixed and random conflation, the latter of which is widely misunderstood even in non-cross-classified models. We provide simulation results showing the possible deleterious impact of such conflation in cross-classified models, and walk through pedagogical examples to illustrate the disaggregation of level-specific effects. We conclude by considering additional model complexities that can arise with cross-classification, providing guidance for researchers in choosing among model specifications, and describing newly available software to aid researchers who wish to disaggregate effects in practice.

Person-Centered Models for Cardiovascular Care: A Review of the Evidence: A Scientific Statement From the American Heart Association

Cardiovascular disease remains the leading cause of death and disability in the United States and globally. Disease burden continues to escalate despite technological advances associated with improved life expectancy and quality of life. As a result, longer life is associated with multiple chronic cardiovascular conditions. Clinical guidelines provide recommendations without considering prevalent scenarios of multimorbidity and health system complexities that affect practical adoption. The diversity of personal preferences, cultures, and lifestyles that make up one's social and environmental context is often overlooked in ongoing care planning for symptom management and health behavior support, hindering adoption and compromising patient outcomes, particularly in groups at high risk. The purpose of this scientific statement was to describe the characteristics and reported outcomes in existing person-centered care delivery models for selected cardiovascular conditions. We conducted a scoping review using Ovid MEDLINE, Embase.com, Web of Science, CINAHL Complete, Cochrane Central Register of Controlled Trials through Ovid, and ClinicalTrials.gov from 2010 to 2022. A range of study designs with a defined aim to systematically evaluate care delivery models for selected cardiovascular conditions were included. Models were selected on the basis of their stated use of evidence-based guidelines, clinical decision support tools, systematic evaluation processes, and inclusion of the patient's perspective in defining the plan of care. Findings reflected variation in methodological approach, outcome measures, and care processes used across models. Evidence to support optimal care delivery models remains limited by inconsistencies in approach, variation in reimbursement, and inability of health systems to meet the needs of patients with chronic, complex cardiovascular conditions.

Socioeconomic inequalities in type 2 diabetes comorbidities in different population subgroups: trend analyses using German health insurance data

While socioeconomic inequalities in the prevalence and management of type 2 diabetes (T2D) are well established, little is known about whether inequalities exist in the prevalence and the temporal development of T2D comorbidities. Previous research points towards expansion of morbidity in T2D as depicted mainly by a rising trend of T2D comorbidities. Against this background, and using German claims data, this study aims to examine whether socioeconomic status (SES) inequalities exist in the rates and the temporal development of T2D comorbidities. Since previous research indicates varying risk levels for T2D prevalence in the population subgroups: working individuals, nonworking spouses and pensioners, the analyses are stratified by these three population subgroups. The study is done on a large population of statutory insured individuals with T2D in three time-periods between 2005 and 2017. Predicted probabilities of three comorbidity groups and the number of comorbidities were estimated using logistic and ordinal regression analyses among different income, education and occupation groups. Interaction analyses were applied to examine whether potential SES inequalities changed over time. The study showed that neither the cross-sectional existence, nor the temporal development of T2D comorbidities differed significantly among SES groups, ruling out SES inequalities in the prevalence and the temporal development of T2D comorbidities in Germany. In men and women of all examined population subgroups, predicted probabilities for less severe cardiovascular (CVD) comorbidities, other vascular diseases and the number of comorbidities per individual rose significantly over time regardless of SES, but little if any change took place for more severe CVD comorbidities. Another important finding is that the population subgroup of nonworking spouses had markedly higher predicted probabilities for most of the examined outcomes compared to working individuals. The study indicates that the expansion of morbidity in T2D in Germany does not appear to be SES-dependent, and applies equally to different population subgroups. Yet, the study highlights that nonworking spouses are a susceptible population subgroup that needs to be focused upon when planning and implementing T2D management interventions.

Exploring regional and sociodemographic disparities associated with unenrollment for the disease management program for type 2 Diabetes Mellitus using Bayesian spatial modelling

BACKGROUND

The disease management program (DMP) for type 2 Diabetes Mellitus (T2DM) is the largest DMP in Germany. Our goal was to analyze regional differences in unenrollment rates, suggest areas for intervention and provide background information, which population groups in which locations are currently not enrolled in the DMP for T2DM.

METHODS

In this study, we used data of the 1.7 mil. insurants of the AOK Nordost health insurance. For the visualization of enrollment potential, we used the Besag-York-Mollie model (BYM). The spatial scan statistic (SaTScan) was used to detect areas of unusually high rates of unenrolled diabetics to prioritize areas for intervention. To explore sociodemographic associations, we used Bayesian spatial global regression models. A Spatially varying coefficient model (SVC) revealed in how far the detected associations vary over space.

RESULTS

The proportion of diabetics currently not enrolled in the DMP T2DM was 36.8% in 2019 and varied within northeastern Germany. Local clusters were detected mainly in Mecklenburg-West-Pomerania and Berlin. The main sociodemographic variables associated with unenrollment were female sex, younger age, being unemployed, foreign citizenship, small household size and the proportion of persons commuting to work outside their residential municipality. The SVC model revealed important spatially varying effects for some but not all associations.

CONCLUSION

Lower socioeconomic status and foreign citizenship had an ubiquitous effect on not being enrolled. The DMP T2DM therefore does currently not reach those population groups, which have a higher risk for secondary diseases and possible avoidable hospitalizations. Logically, future interventions should focus on these groups. Our methodology clearly suggests areas for intervention and points out, which population group in which locations should be specifically approached.

Deprived social status is associated with decreased use of oral chemotherapy in patients with metastatic colorectal cancer: A retrospective cohort study on administrative databases in a French University Hospital

Factors associated with the choice of oral versus intravenous CT are not clearly established. Our purpose was to evaluate the influence of social status and home distance to hospital on the use of oral CT in patients with metastatic colorectal cancer (mCRC). This retrospective single-center study included mCRC patients between 2011 and 2017. Patient social status was assessed by European Deprivation Index (EDI) and home distance to the hospital was calculated. Univariable and multivariable logistic regression analyses were performed. One hundred and seventy-five mCRC patients were included, with 71 receiving oral CT. Most deprived patients received less oral CT (OR 0.5 [0.26, 0.96], p = .039). No association was found for road distance. Previous use of adjuvant oral CT was associated with oral CT in mCRC (OR 2.65 [1.06, 6.66], p = .038). Our results suggest that deprived social status is a factor associated with decreased use of oral CT in patients with mCRC. Clinical trial registration: no registration.

[Social inequalities in healthcare provision for patients with coronary heart disease: Results from the GEDA (German Health Update) study 2014/2015]

INTRODUCTION

Little is known about social inequalities in outpatient long-term care of coronary heart disease (CHD) in Germany.

METHODS

Regression analyses are based on the responses of women and men who participated in the national cross-sectional study "German Health update" (GEDA) 2014/2015 and had self-reported CHD (N=920). Outpatient healthcare of CHD was analysed on the basis of the self-reported administration of antihypertensive and cholesterol-lowering drugs, and the frequency of general practitioner (GP) contacts.

RESULTS

On average, respondents visited their GP 7.5 times a year (mean). 46 % did not receive guideline-consistent treatment, i. e. both antihypertensive and cholesterol-lowering drugs. Respondents of lower social status consulted their GP more frequently (approx. two visits per year) than those of higher social status (AME: 1.94; 95% CI 0.56 to 3.31). Regarding treatment with antihypertensive and cholesterol-lowering drugs, there were no significant differences for either gender or social status. Nevertheless, the probability that respondents with increased levels of blood lipids or cholesterol took only one or none of the two medications recommended for long-term treatment of CHD was reduced by 54 percentage points (AME: -0,54; 95% CI -0,61 to -0,48).

DISCUSSION

There are no social inequalities in the treatment of CHD patients with antihypertensive and cholesterol-lowering drugs, but inequalities exist in the frequency of visits to the GP who is more often consulted by the more socially disadvantaged patients.

CONCLUSION

With about 7.5 consultations per year, CHD patients visit their general practitioner more often than average, but in about half of these patients the medication supply is less than optimal. This may indicate a deficit in the medical treatment of CHD that cannot be explained by social inequalities. A possible starting point for improving healthcare, especially for patients without other risk factors, is to focus more strongly on a guideline-based approach to prescribing medication for CHD patients.

Participation in disease management programs and major adverse cardiac events in patients after acute myocardial infarction: a longitudinal study based on registry data

Background

Cardiovascular diseases are still the main cause of death in the western world. However, diminishing mortality rates of acute myocardial infarction (AMI) are motivating the need to investigate the process of secondary prevention after AMI. Besides cardiac rehabilitation, disease management programs (DMPs) are an important component of outpatient care after AMI in Germany. This study aims to analyze outcomes after AMI among those who participated in DMPs and cardiac rehabilitation (CR) in a region with overall increased cardiovascular morbidity and mortality.

Methods

Based on data from a regional myocardial infarction registry and a 2-year follow-up period, we assessed the occurrence of major adverse cardiac events (MACE) in relation to participation in CR and DMP, risk factors for complications and individual healths well as lifestyle characteristics. Multivariable Cox regression was performed to compare survival time between participants and non-participants until an adverse event occurred.

Results

Of 1094 observed patients post-AMI, 272 were enrolled in a DMP. An association between DMP participation and lower hazard rates for MACE compared to non-enrollees could not be proven in the crude model (hazard ratio = 0.93; 95% confidence interval = 0.65–1.33). When adjusted for possible confounding variables, these results remained virtually unchanged (1.03; 0.72–1.48). Furthermore, smokers and obese patients showed a distinctly lower chance of DMP enrollment. In contrast, those who participated in CR showed a lower risk for MACE in crude (0.52; 0.41–0.65) and adjusted analysis (0.56; 0.44–0.71).

Conclusions

Participation in DMP was not associated with a lower risk of MACE, but participation in CR showed beneficial effects. Adjustment only slightly changed effect estimates in both cases, but it is still important to consider potential effects of additional confounding variables.

Cross-classified multilevel models (CCMM) in health research: A systematic review of published empirical studies and recommendations for best practices

Recognizing that health outcomes are influenced by and occur within multiple social and physical contexts, researchers have used multilevel modeling techniques for decades to analyze hierarchical or nested data. Cross-Classified Multilevel Models (CCMM) are a statistical technique proposed in the 1990s that extend standard multilevel modeling and enable the simultaneous analysis of non-nested multilevel data. Though use of CCMM in empirical health studies has become increasingly popular, there has not yet been a review summarizing how CCMM are used in the health literature. To address this gap, we performed a scoping review of empirical health studies using CCMM to: (a) evaluate the extent to which this statistical approach has been adopted; (b) assess the rationale and procedures for using CCMM; and (c) provide concrete recommendations for the future use of CCMM. We identified 118 CCMM papers published in English-language literature between 1994 and 2018. Our results reveal a steady growth in empirical health studies using CCMM to address a wide variety of health outcomes in clustered non-hierarchical data. Health researchers use CCMM primarily for five reasons: (1) to statistically account for non-independence in clustered data structures; out of substantive interest in the variance explained by (2) concurrent contexts, (3) contexts over time, and (4) age-period-cohort effects; and (5) to apply CCMM alongside other techniques within a joint model. We conclude by proposing a set of recommendations for use of CCMM with the aim of improved clarity and standardization of reporting in future research using this statistical approach.

Patients with coronary artery disease after acute myocardial infarction: effects of continuous enrollment in a structured Disease Management Program on adherence to guideline-recommended medication, health care expenditures, and survival

OBJECTIVE

Acute myocardial infarction (AMI) carries increased risk of mortality and excess costs. Disease Management Programs (DMPs) providing guideline-recommended care for chronic diseases seem an intuitively appealing way to enhance health outcomes for patients with chronic conditions such as AMI. The aim of the study is to compare adherence to guideline-recommended medication, health care expenditures and survival of patients enrolled and not enrolled in the German DMP for coronary artery disease (CAD) after an AMI from the perspective of a third-party payer over a follow-up period of 3 years.

METHODS

The study is based on routinely collected data from a regional statutory health insurance fund (n = 15,360). A propensity score matching with caliper method was conducted. Afterwards guideline-recommended medication, health care expenditures, and survival between patients enrolled and not enrolled in the DMP were compared with generalized linear and Cox proportional hazard models.

RESULTS

The propensity score matching resulted in 3870 pairs of AMI patients previously and continuously enrolled and not enrolled in the DMP. In the 3-year follow-up period the proportion of days covered rates for ACE-inhibitors (60.95% vs. 58.92%), anti-platelet agents (74.20% vs. 70.66%), statins (54.18% vs. 52.13%), and β-blockers (61.95% vs. 52.64%) were higher in the DMP group. Besides that, DMP participants induced lower health care expenditures per day (€58.24 vs. €72.72) and had a significantly lower risk of death (HR: 0.757).

CONCLUSION

Previous and continuous enrollment in the DMP CAD for patients after AMI is a promising strategy as it enhances guideline-recommended medication, reduces health care expenditures and the risk of death.

[Indices of Multiple Deprivation for the analysis of regional health disparities in Germany : Experiences from epidemiology and healthcare research]

BACKGROUND

Deprivation indices allow material and social differences at the regional level to be described in a statistically efficient and concise manner and to use these in health analyses. Following the British example, Indices of Multiple Deprivation (IMDs) are now available for Germany, the German Index of Multiple Deprivation (GIMD) as well as its regional versions. In this study, empirical experiences based on the use of these indices in health studies will be presented.

METHOD

The German IMDs consist of seven deprivation domains, which represent single aspects of deprivation (income, employment, and educational deprivation, municipal revenue deprivation, social capital deprivation, environment and security deprivation). Specific indicators were generated from data of official statistics and assigned to the deprivation domains. The weighted single domains were finally combined to an overall index. The German IMDs are available at a municipal level and at a district level.

RESULTS

Analyses using the IMDs showed significant associations between regional deprivation and mortality, morbidity and aspects of health services research. Multilevel analyses showed significant associations with regional deprivation, independent of individual factors.

CONCLUSIONS

The German IMDs are valid and efficient tools for the use in epidemiology and health services research, but also for health policy. When constructing deprivation indices, several methodological challenges have to be considered.

Trends in diagnosis-specific work disability before and after ischaemic heart disease: a nationwide population-based cohort study in Sweden

OBJECTIVES

We examined trends of diagnosis-specific work disability before and after ischaemic heart disease (IHD).

DESIGN

Participants were followed 4 years before and 4 years after an IHD event for diagnosis-specific work disability (sickness absence and disability pension).

SETTING AND PARTICIPANTS

A Swedish population-based cohort study using register data on all individuals aged 25-60 years, living in Sweden, and who suffered their first IHD event in 2006-2008 (n=23 971) was conducted.

RESULTS

Before the event, the most common diagnoses of work disability were musculoskeletal disorders (21 annual days for men and 44 for women) and mental disorders (19 men and 31 for women). After multivariable adjustments, we observed a fivefold increase (from 12 to 60 days) in work disability due to diseases of the circulatory system in the first postevent year compared with the last pre-event year among men. Among women, the corresponding increase was fourfold (from 14 to 62 days). By the second postevent year, the number of work disability days decreased significantly compared with the first postevent year among both sexes (to 19 days among men and 23 days among women). Among women, mean days of work disability due to diseases of the circulatory system remained at a higher level than among men during the postevent years. Work disability risk after versus before an IHD event was slightly higher among men (rate ratio (RR) 2.49; 95% CI 2.36 to 2.62) than among women (RR 2.29, 95% CI 2.12 to 2.49). When pre-event long-term work disability was excluded, diseases of the circulatory system were the most prevalent diagnosis for work disability after an IHD event among both men and women.

CONCLUSIONS

An IHD event was strongly associated with an increase in work disability due to diseases of the circulatory system, especially among men and particularly in the first postevent year.

Permanent work disability before and after ischaemic heart disease or stroke event: a nationwide population-based cohort study in Sweden

OBJECTIVES

We examined the risk of disability pension before and after ischaemic heart disease (IHD) or stroke event, the burden of stroke compared with IHD and which factors predicted disability pension after either event.

DESIGN

A population-based cohort study with follow-up 5 years before and after the event. Register data were analysed with general linear modelling with binary and Poisson distributions including interaction tests for event type (IHD/stroke).

SETTING AND PARTICIPANTS

All people living in Sweden, aged 25‒60 years at the first event year, who had been living in Sweden for 5 years before the event and had no indication of IHD or stroke prior to the index event in 2006‒2008 were included, except for cases in which death occurred within 30 days of the event. People with both IHD and stroke were excluded, resulting in 18 480 cases of IHD (65%) and 9750 stroke cases (35%).

PRIMARY OUTCOME MEASURES

Disability pension.

RESULTS

Of those going to suffer IHD or stroke event, 25% were already on disability pension a year before the event. The adjusted OR for disability pension at first postevent year was 2.64-fold (95% CI 2.25 to 3.11) for people with stroke compared with IHD. Economic inactivity predicted disability pension regardless of event type (OR=3.40; 95% CI 2.85 to 4.04). Comorbid mental disorder was associated with the greatest risk (OR=3.60; 95% CI 2.69 to 4.83) after an IHD event. Regarding stroke, medical procedure, a proxy for event severity, was the largest contributor (OR=2.27, 95% CI 1.43 to 3.60).

CONCLUSIONS

While IHD event was more common, stroke involved more permanent work disability. Demographic, socioeconomic and comorbidity-related factors were associated with disability pension both before and after the event. The results help occupational and other healthcare professionals to identify vulnerable groups at risk for permanent labour market exclusion after such an event.

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

Background

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.

Methods

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.

Results

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.

Conclusion

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.

Age-related differences in persistence in women with breast cancer treated with tamoxifen or aromatase inhibitors in Germany

AIMS

To study age-related persistence in postmenopausal women with endocrine-responsive breast cancer treated with tamoxifen (TAM) and aromatase inhibitors (AI).

METHODS

Data on 29,245 patients diagnosed with metastatic or non-metastatic breast cancer (BC) and initially treated with TAM or AI between 2004 and 2013 were included. The primary outcome measure was the age-dependent rate of discontinuation of endocrine treatment within 5years after initiation. Discontinuation of therapy was defined as a period of at least 90days without treatment. A multivariate Cox regression model was created to determine the influence of age on the risk of discontinuation. Health insurance type (private/statutory), type of care (gynecological/general), region (West/East Germany), concomitant diagnoses (depression, osteoporosis, and diabetes), and Charlson Comorbidity Score were included as covariates.

RESULTS

The mean ages of the women in the <70 and ≥70 groups were 55.9 (SD: 9.7) and 77.4 (SD: 5.4) years, respectively. Within 5years after treatment initiation, 88.8% of women <70 of age and 82% of women ≥70 years of age had terminated treatment (p-value<0.001). Patients aged ≥70 exhibited a lower risk of treatment discontinuation than patients aged <70 (HR=0.75, 95% CI: 0.66-0.85). Furthermore, gynecological practices, disease management programs, and high Charlson scores increased persistence.

CONCLUSIONS

Overall, the present study indicates that persistence rates are low in both women with BC aged <70 and those aged ≥70 years. We also found that younger women with BC are at a higher risk of treatment discontinuation than older women.

Forgone care among chronically ill patients in Germany-Results from a cross-sectional survey with 15,565 individuals

The decision not to seek health care although one feels that care is needed (forgone care), is influenced by various factors. Within the study "Responsiveness in ambulatory care" 15,565 chronically ill (coronary heart disease and/or type 2 diabetes) patients in Germany were surveyed in 2013. The survey included questions on forgone care, perceived discrimination when seeking care, net-income, subjective health status and subjective socioeconomic status (subSES). Survey data were linked on patient-level with administrative claims data by a German sickness fund. We applied multivariate binomial logistic regression analyses to assess the association between age, sex, comorbidities, living area, subjective health status, subSES, experienced discrimination, net-equivalent income and reported forgone care. The majority in the sample are men (71.4%), the average age is 69.4 (SD: 10.2) years and 14.1% reported forgone care. In the multivariate model, we find that younger age, female gender, perceived discrimination, depression, and a poor subjective health status increase the odds of reporting forgone care. Overall, our results suggest that a negative experience with the health care system, i.e. perceived discrimination/unfair treatment, are strong predictors of forgone care among the chronically ill.