Health Economic and Socio Economic Impact of Diabetes Mellitus in Denmark: A Register Based Investigation

Diabetes Mellitus is one of our time’s greatest threats towards public health with both developed and
developing countries experiencing increasing prevalence (1). On a global level it is estimated that
around 9% of adults aged 18+ are diagnosed with diabetes. World Health Organization (WHO)
projects, that diabetes will be the 7th leading cause of death in 2030(2). Diabetes is a chronic disease
with major impact on patient’s lives in terms of morbidity, mortality and reduced quality of life(2, 3).
Hence, diabetes patients pose high demands on health care systems for continuous and quality health
care and nursing services in a lifetime perspective. In a societal perspective, diabetes also poses a
substantial economic burden in terms of productivity loss due to reduced labour market performance
and premature mortality(4, 5).

This Ph.D. dissertation draft investigates health economic and socio economic impact of Diabetes
Mellitus in Denmark. The Ph.D. dissertation draft rests upon a large-scale register based observational
investigation, the Diabetes Impact Study(6), which investigates epidemiological, health economic and
socioeconomic aspects of diabetes in Denmark. Denmark has a particularly strong tradition for register
based research(7), which has been further enhanced with the establishment of the Danish National
Diabetes Register (NDR) (8, 9). The study population in Diabetes Impact Study includes all Danish
diabetes patients registered in NDR. Individual patient data are combined, through the unique Danish
personal identification number, with data from the Danish National Patient Register (10), the Danish
National Prescription Registry (11), the Danish National Health Service Register (12), the Danish Civil
Registration System (13), the National Register of Cause of Death(14) and social registers at Statistics
Denmark (SD). This results in a unique, comprehensive data pool on diabetes patients’ morbidity,
health care usage and socio-demographic characteristics, ideal for research inquiries. The analysis of the
economic and socioeconomic impact of diabetes is further based upon an extensive epidemiological
framework, including initial validation of the study population obtained from NDR(15), and analyses of
the epidemiological drivers of diabetes and trends over time(16).

Diabetes epidemiology is modelled in the BOX-model, (16) a simple multi-state transition model
reflecting patient flows, based on individual data from the entire Danish diabetes population. The
model categorizes patients according to their disease progression, into three health states: “no”,
“minor” and “severe” complications. This is a novel approach within the field of chronic disease
research. The rationale behind this approach is to be able to compare well-regulated patients, without complications, to patients who have developed complications, in relation to both epidemiological
indicators, resource usage and socio-demographic characteristics. This enables easy interpretable
insights, to inform decision makers concerning where diabetes is heading, and which patients bear the
greatest burden of disease.

The first paper: “Societal costs of diabetes mellitus in Denmark 2011” investigates societal costs of
diabetes in 2011 in Denmark, applying state of the art methods within Cost of Illness (COI) studies(5).
Costs are analyzed according to health care sector services, pharmaceutical consumption, nursing
services, productivity losses and additional cost components including conservative estimates of i.e.
patients’ own time for managing their disease. Diabetes attributable costs are calculated as the cost
difference between a diabetes patient’s total costs and the expected total costs, given total costs of a
relevant control population. The paper concludes that costs attributable to diabetes in 2011 were
estimated to be at least 4.27 billion EUR, corresponding to 14,349 EUR per person-year. A steep
increase in attributable costs for patients with major complications compared with patients without
complications across all cost components is evidenced, also when potential confounders are controlled
for. This underlines an economic rationale for secondary prevention. Furthermore, impact of diabetes
on patients’ labour market participation is illustrated through significant different labour market
patterns and lower annual income, among patients with diabetes compared with the control subjects,
also after control for educational level. Increasing loss in productivity among patients with a higher
degree of complications and higher educational level is evidenced.

We apply the estimated costs attributable to diabetes to the BOX-model, along with observed
epidemiological trends based on 14 years of data. From this, the model forecasts the future Danish
diabetes population and corresponding cost level. The second paper: “Societal costs of diabetes
mellitus 2025 and 2040 – forecasts based on real world cost evidence and observed
epidemiological trends in Denmark” presents results of this projection, given three different
epidemiological scenarios. A number of suggested economic hypotheses are analysed according to the
projections, to accommodate that forecasts mirror the observed cost structure and level in 2011,
whereas the future will imply new investments, treatments and cost structures. Results indicate that
increasing prevalence of diabetes and, hence, costs of diabetes, cannot be dramatically altered within
the shorter time span. The diabetic population will approximately double in the next 10 years, primarily
due to already achieved improvements in diabetes mortality and morbidity. Hereafter, it seems realistic
to assume a 2.5 or tripling of the patient population in 2040. A consequence of increasing prevalence due to prognosis improvements is a shift in the relative distribution of “expensive” patients, with
severe complications, compared to well-regulated patients, without complications. This implies a lower
cost per PYRS, and hence also a lower cost increase relative to prevalence increase. An important
conclusion therefore is that prevalence is an inadequate measure of disease control when it comes to
chronic diseases. Lower cost per patient year might be more desirable than lower prevalence, as this
means that each patient is living better with his or her disease, contributing with more patient years of a
higher quality. This scenario is not only economically efficient in the health care system, where
increased resources invested in secondary prevention and better compliance to treatment, will save
resources for later care of patients with expensive severe complications. In a societal perspective, better
regulated patients will also perform better on the labor market reducing the evidenced productivity loss
of patients with diabetes. Finally, for the individual patient it is crucial to be able to live a long life of
high quality despite suffering from a chronic disease. The inevitable prevalence increase just underpin
the importance of the goal of a diabetes population, which might count more heads, but with the
majority of patients living without complications of their disease, and where the societal costs of
diabetes are minimized.

Inequality in health is an important topic globally and nationally. With diabetes being one of the leading
chronic diseases, inequality in diabetes is relevant to investigate. The unique data foundation of the
Diabetes Impact Study allows for novel insights into inequality in diabetes morbidity patterns and
health care usage as well as the determining factors behind inequality. These new insights can guide
future strategies aiming to reduce inequality in diabetes, and within chronic disease in general.
Differences in compliance to treatment and persistence of life style changes among chronic patients(17-
19), and the economic potential of hindering complication development among diabetes patients, stress
an economic potential of targeting efforts towards specific patient groups. The third paper:
“Decomposing inequality in Diabetes Patients’ Morbidity Patterns, Survival and Health Care
Service Usage in Denmark”, unfold novel inequality aspects in diabetes through investigation of
socioeconomic inequality in diabetes morbidity patterns, survival rates and time before complication
development plus in health care utilization patterns. Inequalities are decomposed into individual sociodemographic
and morbidity determinants, and the effect on inequality estimates, of applying
educational level or income level as proxy for patients’ SES is investigated. Inequality in clinical diabetes
morbidity patterns from diagnosis to death is demonstrated, with patients of lower SES being
diagnosed later and in a worse state, living shorter lives with their disease and experiencing more severe complications. Results further indicate differences in realized access to health care, with patients of
higher SES, especially higher educational level, to a greater extent enjoying out-patient services,
especially rehabilitation, and also specialist services in primary care. Health care usage hereby not always
corresponds to patients’ need. These patterns of health care usage differentiates from previously
demonstrated patterns of health care usage within the Danish population(20), showing that other
inequality aspects might be at stake within a chronic disease as diabetes. To reduce inequality in health,
it is therefore important to invest in efforts targeted at socially vulnerable groups in relation to early
detection, secondary prevention and compliance of treatment. Elderly people, people living alone,
people outside the labor-market and immigrants are especially vulnerable when belonging to lower SES.
These patient groups would benefit from being targeted directly. Methodologically, our findings
underpin important differences of using respectively income and educational level as proxy for SES.
Results indicate that education is more decisive than income for patients’ realized access whereas
income-related inequality in health care usage to a higher extent is explained by morbidity. Higher
estimates of income-related inequality compared to education related inequality is an expression of
reversed causality between diabetes and SES. This aspect of inequality is in itself a valuable message of
diabetes’ impact on patients’ labour market participation. More of these findings may underpin
universal structures behind inequality in diabetes and in chronic disease in general, valuable beyond
Denmark.

Altogether, this Ph.D. dissertation draft brings forward novel real world evidenced estimates of the
economic and socioeconomic impact of diabetes, by bringing together comprehensive individual data
from national registries, a solid epidemiological framework, and up to date statistical methodology. The
findings provide important innovative, universal knowledge, valuable beyond Denmark. With the
challenges of an increasing diabetes population urgent in all countries globally, it is crucial that diabetes
is dealt with in a way, where continued increases in diabetes prevalence will not lead to marked
decreased population health, uncontrolled costs and increased inequality in health. Secondary
prevention, rehabilitation and treatment compliance already from diagnosis, to delay or hinder the
onset of expensive late complications are strategies, which are likely to be cost efficient. Especially, if
they are also targeted vulnerable patient groups, who must receive more attention, to be able to take
advantage, and gain success, of the services provided, even in a universal health care system. To meet
the challenge of diabetes, societies must shift focus of attention from a fear of high numbers of
diabetes to a goal of well-regulated patients experiencing a minimum of severe late complications.

Diabetes prevalence increase is not solely to be seen as a defeat of public health preventive strategies,
but also as a success of current treatment expressed through prognosis improvements. Hence, a healthy
diabetes population is a success criterion, which should be prioritized, measured through patients’
quality and quantity of life, lower cost per patient year and reduced impact on patients’ labour market
participation. New investments and structural improvements within diabetes must prioritise weak
patient groups, with focus on realised access, compliance and equity in relation to outcomes of
prevention and treatment. In this way, the increasing burden of chronic diabetes patients can be met by
a prepared health care system better capable of ensuring quality care and equal access to health care.