INCIDENCE, PREVALENCE, COSTS AND QUALITY OF CARE OF TYPE 1 DIABETES IN ITALY, AGE 0-29 YEARS: THE POPULATION-BASED ARNO OBSERVATORY, 2002-2012
Graziella Bruno1, MD, Eva Pagano2, BS, Elisa Rossi3, BS, Salvatore Cataudella3, BS, Marisa De Rosa3,PharmaD, Giulio Marchesini4, MD, Roberto Miccoli5, MD, PHD, Olga Vaccaro6, MD, Enzo Bonora7, MD, PHD on behalf of ARNO Diabetes Observatory
1Dept. of Medical Sciences, University of Torino, Italy
2Unit of Cancer Epidemiology, "Città della Salute e della Scienza" Hospital-University of Turin and CPO Piemonte, Turin, Italy
3CINECA Interuniversity Consortium, Health Department, Bologna, Italy
4Unit of Metabolic Diseases and Clinical Dietetics, Alma Mater Studiorum University, Bologna, Italy
5Dept. of Endocrinology and Metabolism, Section of Metabolic Diseases and Diabetes, University of Pisa, Italy
6Dept. of Clinical Medicine and Surgery, University of Napoli Federico II, Napoli, Italy
7Endocrinology, Diabetes and Metabolism, Dept. of Medicine, University and University Hospital of Verona, Italy
Corresponding author and address for reprint requests: Prof. Graziella Bruno,
Dept. of Medical Sciences, University of Torino, corso Dogliotti 14, I-10126 Torino, Italy
Tel: +39 11 633 6709 Fax: +39 11 6634 751 Email: graziella.bruno@unito.it Word count: 2303 Tables:4 Figures: 2
ABSTRACT
Background and Aims: to assess temporal trend in incidence (2003-12) and prevalence (2002-12) of type 1 diabetes in children and young adults, direct costs and selected indicators of quality of care under the coverage of the universalistic Italian National Health Service (NHS).
Methods and Results: The ARNO Observatory, a healthcare monitoring system based on administrative data, identified a population-based multiregional cohort of subjects aged 0-29 years. Type 1 diabetes was defined by at least two prescriptions of insulin over 12 months and continuous insulin-treatment in the following year. Indicators of quality of care and directs costs were assessed in persons with diabetes and in people without diabetes, individually matched for age, gender and health unit (1:4 ratio). We identified 2357 incident cases of type 1 diabetes aged 0-29 years (completeness of ascertainment, 99%). Incidence rates were similar in ages 0-14 (15.8, 95% CI 14.9-16.8) and 15-29 years (16.3, 15.4-17.2), with no significant trend. Prevalence increased from 137 to 166.9/100,000, particularly in the age 15-29 years. Direct costs accounted for € 2,117 in persons with diabetes and € 292 in control individuals. A statistically significant decreasing trend in hospitalization for acute complications was evident (p<0.001), which was almost completely due to ketoacidosis. People with at least one HbA1c measurement over the year was 48.5%.
Conclusion: we showed high incidence and increasing prevalence of type 1 diabetes in young adults in Italy, which impact on direct costs under the universalistic coverage of the NHS.
Population-based registries both in Europe and in the United States have provided evidence that the incidence of type 1 diabetes is increasing by 2.8-4% per year (1-2). As most registries limit the recruitment of data up to 14 years (3), the impact of the disease on health and direct costs of young people is probably under-estimated (4-7). The increasing burden of the disease on health care services might translate into a lower quality of care if patients are not adequately supported with devoted personnel, and this is particularly true for adolescents and young adults, who often have low compliance to intensive insulin treatment and adherence to scheduled visits (8-9). Adolescents still have high hospitalization rates for hypo- and hyper-glycemic acute diabetic complications, that remain their leading causes of mortality (10-13). Therefore, a correct estimation of prevalence and direct costs of type 1 diabetes could support health policy-makers in planning the adequate allocation of resources and in improving quality of care. Epidemiological data, however, are very limited (14-15).
The wider availability of electronic health records is providing new opportunities to study the epidemiology of type 1 diabetes in the adults (3). Continuous insulin-treatment is the hallmark of type 1 diabetes in children and adolescents and administrative data linked with demographic registries provide accurate population-based data up to young adulthood. In this study we examined temporal trend in incidence (period 2003-2012) and prevalence (period 2002-2012) of type 1 diabetes in children and young adults living in Italy, using the unique population-based ARNO Observatory (16). Finally, we analyzed data on direct costs and selected indicators of quality of care of the disease in the 2012 cohort.
Methods
Identification of incident and prevalent subjects with type 1 diabetes
The population-based multiregional ARNO Observatory is a healthcare monitoring system, based on administrative data (demographics, disease specific exemptions, outpatient drug prescriptions, inpatient hospital discharges, lab tests prescriptions and instrumental examination) from more than 30 Local Health Units (LHUs) scattered throughout Italy. It was organized by
CINECA to provide local and national reports on the main epidemiologic features and drug consumption of representative subgroups of Italian people, such as children, the elderly and those with chronic diseases (16-20). Prescriptions are provided by all participating LHUs, whereas hospitalizations and health services utilization data are provided by a subgroup of LHUs. In Italy, the identification of all people using health care services is facilitated by the universalistic National Health System (NHS), which allows all residents to receive care, irrespective of citizenship, social class and income. Moreover, individuals with diabetes receive free access (without any co-payment) to drugs, diagnostic procedures and hospitalizations, so that the use of private care resources is almost negligible in the diabetes area.
This report refers data of the 2002-2012 population-based cohort of people aged 0-29 years (n=1,713,196 in 2012) who were resident in 18 Italian areas covered by LHUs participating to the ARNO Observatory. They were distributed as follows: 57% in the North, 15% in the Centre, and 28% in the South. All data referring to one subject were linked by means of a unique encrypted identification code. Most LHU (14/18) had prescriptions data covering the whole 2002-2012 period, whereas 4 of them only had shorter time period data (2004-2012, 2006-2012, 2008-2012 and 2010-2012, respectively). Subjects were considered to be affected by type 1 diabetes if they were aged 0-29 years, were residents in the LHUs, had received over 12 months at least two prescriptions of insulin (Anatomical Therapeutic Chemical Classification System [ATC], code A10A), and were continuously insulin-treated in the following year. Given these criteria, both gestational diabetes and type 2 diabetes may be excluded. Incident cases were identified by excluding subjects identified as having diabetes during the previous year. Prevalence and incidence data were collected for period 2002-2012 and 2003-2012, respectively, to allow for assessment of continuous insulin treatment for at least 24 months.
Direct costs
Data referring to prescriptions, hospitalizations, and health care services utilization in 2012 were analyzed, including consultations, laboratory tests, reactive strips, glucometers and insulin
pumps. Data sources were linked to the patients’ fiscal code, an encrypted identity code generated by name, date and place of birth. Five per cent only of prescriptions were excluded from analyses because of incomplete patient identification. A case-cohort study design was employed to compare prescriptions, hospitalization rates, services use and direct costs in persons with diabetes (n=2,905) and in those without diabetes individually matched for age, gender and LHU with a 1:4 ratio (n=11.620). For every patient, healthcare costs were estimated as the sum of all the resources supplied during the year. For examinations, consultations and hospitalizations, regional tariffs, which represent the reimbursement paid by the LHUs to healthcare providers, were used as standard costs. Drugs were valued with the public prices reimbursed by NHS. Costs were expressed in Euros.
Quality of care
We assessed indicators of quality of care in 2009-2012 in the prevalent cohort of diabetic patients aged 0-29 years (n=700,479 individuals in 2012). All individuals with type 1 diabetes (n=995 in 2009; n=1016 in 2010; n=1033 in 2011; n=1162 in 2012) were individually matched to 4 individuals without diabetes for age, gender and LHU. As process indicator we assessed the frequency of people with at least one outpatient HbA1c measurement in the last 12 months. As outcome indicator we assessed the hospitalization rate for acute diabetes complications. Ketoacidosis and hyperosmolar coma were defined as hospital discharge records reporting the 250.1, 250.2, or 250.3 ICD-9-CM codes as primary or secondary diagnosis. Hypoglycemic coma was defined by ICD-9-CM code: 251.0.
Statistical analyses
Denominators of incidence/100,000 person-years and prevalence/1000 were residents in each LHU by each year. The 95% confidence intervals (CIs) of rates were estimated assuming the Poisson distribution of the cases. From the database, we derived total drug packages and the cumulative cost of treatment, the number of hospital admissions and consultations/year. The mean cost of drug treatment was calculated by multiplying the number of prescribed packages by their unit cost. Completeness of ascertainment was assessed in 2012 using the two-sample
capture-recapture method, comparing cases ascertained through prescriptions (primary data source) and hospital discharges (secondary data source). To assess trend in incidence and prevalence during the study period, we used the Cochrane-Armitage test. Statistical analyses were conducted using Stata version 10.0 A P value less than 0.05 was considered significant. The confidence intervals of direct costs had been computed by the bootstrap method with software R, using a level of significance equal to 95%.
Results
Incidence
Over the 2003-2012 period, 2357 incident cases of type 1 diabetes were identified in the age group 0-29 years (estimated completeness of ascertainment, 99%), giving an incidence rate of 16.1/100,000 person-years (95% CI 15.4-16.7) (Table 1). Rates were similar in males (16.8, 95% CI 15.9-17.8) and females (15.3, 95% CI 14.4-16.2) and in the age group 0-14 years (15.8, 14.9-16.8) compared to the age group 15-29 years (16.3, 15.4-17.2). No significant trend was evident over the study period, although the lowest risk was found in 2012. Large year-by-year variations in risk were evident among five-years age group at diabetes onset, with the lowest risk in the 0-4 years age-group (Figure 1).
Prevalence
Prevalence of type 1 diabetes in age group 0-29 increased over time from 137 to 166.9/100,000 (Table 2). Although the increase was present in both 0-14 and 15-29 age groups, it was particularly relevant in the latter, with an absolute increase of 50 individuals over 100,000 residents from 2002 to 2012 (Figure 2).
Direct Costs
Costs were computed in 2012 in the prevalent cohort of 2905 people with type 1 diabetes, including 556 children and 2349 young adults. As shown in table 3, direct costs accounted for €
2,117 in persons with diabetes and € 292 in control individuals, matched for age, gender and LHU, Compared to non diabetic individuals, the excess of cost was 6-fold in diabetic individuals and higher among children than among young adults. Individuals who had no healthcare cost were 2.7% (n=77) among diabetic people and 30% (n=3420) among non-diabetic people. As shown in Table 4, when we compared costs in diabetic and non diabetic people after having excluding individuals with zero-costs, we found higher mean costs in both groups and, consequently, lower differences among them.
Quality of care
The frequency of people with at least one outpatient HbA1c measurement over the year 2012 was only 48.5%. The distribution among age groups was as follows: 37.2% in subjects aged 5-9, 57.7% in the age 10-14, 61.5% in the age 15-15-9, 53.2% in the age 20-24, and 44.1% in the age 25-29.
Overall hospitalization rates significantly decreased over period 2009-2012 (27.1%, 9.3%, 5.7% and 12%, p<0.001). Respective rates in control subjects showed a similar pattern, although rates were five/ten-fold lower (4.6%, 0.5%, 0.3% and 1.2%). A statistically significant decreasing trend in hospitalization for acute complications was also evident over the study period (11.3%, 5.2%, 4.0% and 5.8%, p<0.001), which was almost completely due to ketoacidosis. Rates were higher in children than in young adults, and the decreasing trend was more clearly observed in the latter.
Conclusions
The present analyses of the ARNO Observatory provides unique epidemiological data of type 1 diabetes in children and young adults. Firstly, incidence rates in young adults up to age 29 years were similarly high compared to children, with stable rates over the ten-years study period in both groups. Second, the prevalence of the disease increased steadily, with almost 50 additional new young adults out of 100,000 residents in the 10-year period. Third, compared to individuals
without diabetes of similar age, the young people with diabetes up to 29 years had more than 6-times higher direct costs under the universalistic coverage of the National Health Service (NHS) and this excess was mainly due to insulin treatment and reagent strips. Finally, overall hospitalization rates and discharges due to acute diabetes complications, an indicator of quality of care, decreased over 2009-12, suggesting that outpatients services provided a good level of care. However, the frequency of HbA1c measurements over the year was still unsatisfactorily low, with almost half of the population-based cohort having no measurement. Such epidemiological data are relevant, as literature is limited on this issue but data are mandatory for comprehensive economic evaluation studies and cost-of-illness modelling on type 1 diabetes.
The strength of our study derives from a population-based multiregional study design, and it is one of the few European studies providing epidemiological and economic cost data of type 1 diabetes in both children and young adults. Data were not affected by the selection bias of specialized centres; the ARNO-CINECA Observatory covers the entire population and is characterized by a high level of standardization and reproducibility, allowing for an easy updating of the analyses and feasible time-trend assessments, previously used for several studies on diabetes care quality and costs (16-20). As we examined a population under the universalistic coverage of the NHS, considerations regarding selection bias due to the impact of cost of the disease on people with a low socioeconomic level do not apply.
The study has limitations too. The administrative data gathered in the ARNO Observatory lack of individual clinical data. We considered as incident cases only subjects who had not received insulin prescriptions in the previous year and were continuously insulin treated for at least one year, thus excluding subjects who were temporarily residents in the LHU. Although we cannot exclude that any prevalent case might have been misclassified as incident case, these numbers, if any, were presumably very low due to the validated identification procedures. Unfortunately, individual validation could not be performed due to the restrictive rules applied to administrative data in the context of the Observatory. As quality indicator, we assessed the frequency of outpatient HbA1c
measurements. A subgroup of individuals might have had HbA1c measurement during their hospitalization, however in the index year 12% only of the cohort was hospitalized, thus slightly affecting our estimate. Finally, childhood registries suggested heterogeneities in risk of the disease within Italy (21-24). Our data did not include the high risk Sardinia Region and covered mainly Northern Italy. Moreover, due to the limited numbers of cases in each LHU, we did not performed geographical comparisons of risk within Italy.
We did not find evidence of an increasing temporal trend in incidence of type 1 diabetes, and results were similar even stratifying data by 5-years age groups. Previous analyses from the EURODIAB study suggested that incidence increased more in children aged 0-4 years (1), but no evidence of a differential pattern by age group was found in a previous analysis of Italian registries (23). Data on incidence of type 1 diabetes in adults are very limited, but they consistently pointed out the numerical relevance of the disease up to 29 years (3). An increasing prevalence of the disease has been observed in the age cohort 0-14 in Germany and in the age 0-19 in the United States (14-15). The SEARCH study showed an increase from 148/100,000 in 2001 to 193/100,000 in 2009 in subjects aged 0-19, which compares to our increase from 137/100,000 in 2002 to 166.9/100,000 in 2012 in age 0-29 (15). In our data, the increase was particularly relevant among young adults (+50/100,000); accordingly, we estimate that almost 5000 more young adults with type 1 diabetes are now cared for by diabetes clinics at national level compared to ten years ago. A parallel increase in devoted specialized care services should also be planned.
Care of type 1 diabetes has profoundly changed over time, favorably affecting quality of life and indicators of quality of care, such as hospitalizations for acute diabetic complications. Our data shows that ketoacidosis requiring hospitalization was low and that it further decreased over time. This finding is consistent with previous observation using national dataset of hospital discharges and clinical data (26-28). However, epidemiological studies have pointed out that despite significant improvements in health care of people with type 1 diabetes, life expectancy at birth remains twelve years lower than in people without diabetes (11). Type 1 diabetes is a chronic disease requiring
self-management, disease-focused behaviors, and frequently scheduled controls. Health system barriers in terms of access to care and its financial costs may have detrimental effects on subjects with chronic conditions such as type 1 diabetes (29-30). Under the universalistic coverage of the NHS we provide evidence that that the costs associated with type 1 diabetes are substantial, with an overall annualized median cost of almost € 2100, which was 6-fold higher than young individuals without diabetes of similar age. These estimates can be used to assess the economic burden of diabetes for publicly insured children and young adults in other countries. Recently, large variability in mortality risk of type 1 diabetes has been pointed out, together with an observed association between greater healthcare expenditure and reduced mortality excess in a subgroup of countries, which would suggest that adequate health-care provision has a positive impact on subjects with type 1 (12).
In conclusion, the ARNO Observatory showed high incidence and increasing prevalence of type 1 diabetes in young adults in Italy, and more than 6-times higher direct costs in diabetic than in non diabetic young people under the universalistic coverage of the NHS.
Acknowledgments: The analyses of ARNO Observatory have been supported by a grant from the Italian Diabetes Society (SID) to CINECA.
Guarantors: Graziella Bruno, Elisa Rossi
Conflict of interest statement: The Authors have no conflict of interest to declare
Authors Contribution: E.P. and G.B. contributed to the study concept and design, researched and interpreted the data. GB drafted the manuscript.
E.R. and E.C. contributed to data analysis and reviewed the manuscript. M.D.R., G.M., R.M., O.V., and E.B. oversaw the progress of the project, contributed to the discussion and reviewed the manuscript.
ER. and G.B. are the guarantors of this work and, as such, had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
References
1. Patterson CC, Gyürüs E, Rosenbauer J, Cinek O, Neu A, Schober E et al. Trends in childhood type 1 diabetes incidence in Europe during 1989-2008: evidence of non-uniformity over time in rates of increase. Diabetologia 2012;55:2142-2147
2. Lawrence JM, Imperatore G, Dabelea D, Mayer-Davis EJ, Linder B, Saydah S et al. Trends in incidence of type 1 diabetes among non-Hispanic white youth in the U.S., 2002-2009. Diabetes 2014;63:3938-3945
3. Bruno G, Gruden G, Songini M. Incidence of type 1 diabetes in age groups above 15 years: facts, hypothesis and prospects for future epidemiologic research. Acta Diabetol 2016; 53:339-472
4. Cameron FJ, Wherrett DK. Care of diabetes in children and adolescents: controversies, changes, and consensus. Lancet 2015;385:2096-2106
5. Trento M, Panero F, Porta M, Gruden G, Barutta F, Cerutti F et al. Diabetes-specific variables associated with quality of life changes in young diabetic people: the Turin Registry of Type 1 Diabetes, Italy. Nutr Metab Cardiovasc Dis 2013;23:1031-1036
6. Kruger J, Brennan A. The cost of Type 1 diabetes mellitus in the United Kingdom: a review of cost-of-illness studies. Eur J Health Econ. 2013;14:887-899
7. Franciosi M, Lucisano G, Amoretti R, Capani F, Bruttomesso D, Di Bartolo P et al. Costs of treatment and complications of adult type 1 diabetes. Nutr Metab Cardiovasc Dis. 2013;23:606-611.
8. Brink SJ, Miller M, Moltz KC. Education and multidisciplinary team care concepts for pediatric and adolescents diabetes mellitus. J Pediatr Endocrinol Metab 2012:15:1113-1130 9. Kakleas K, Kandiyla B, Karayianni C, Karavanaki K. Psychological problems in
adolescents with type 1 diabetes mellitus. Diabetes Metab 2009;35:339-350
10. Maahs DM, Hermann JM, Holman N, Foster NC, Kapellen TM, Allgrove J et al. Rates of Diabetic Ketoacidosis: International Comparison With 49,859 Pediatric Patients With Type 1 Diabetes From England, Wales, the U.S., Austria, and Germany. Diabetes Care 2015;38:1876-1882
11. Huo L, Harding JL, Peeters A, Shaw JE, Magliano DJ. Life expectancy of type 1 diabetic patients during 1997-2010: a national Australian registry-based cohort study. Diabetologia 2016;59:1177-1185
12. Morgan E, Cardwell CR, Black CJ, McCance DR, Patterson CC. Excess mortality in Type 1 diabetes diagnosed in childhood and adolescence: a systematic review of population-based cohorts. Acta Diabetol. 2015;52:801-807
13. Bruno G, Cerutti F, Merletti F, Novelli G, Panero F, Zucco C et al. Short-term mortality risk in children and young adults with type 1 diabetes: the population-based Registry of the Province of Torino, Italy. Nutr Metab Cardiovasc Dis 2009;19:340-344
14. Bendas A, Rothe U, Kiess W, Kapellen TM, Stange T, Manuwald U et al. Trends in Incidence Rates during 1999-2008 and Prevalence in 2008 of Childhood Type 1 Diabetes Mellitus in Germany--Model-Based National Estimates. PLoS One 2015;10:e0132716. 15. Dabelea D, Mayer-Davis EJ, Saydah S, Imperatore G, Linder B, Divers J et al. Prevalence
of type 1 and type 2 diabetes among children and adolescents from 2001 to 2009. JAMA 2014; 311:1778-1786
16. Marchesini G, Forlani G, Rossi E, Berti A, De Rosa M. The direct economic cost of pharmacologically-treated diabetes in Italy-2006. The ARNO observatory. Nutr Metab Cardiovasc Dis 2011;21:339-346.
17. Marchesini G, Bernardi D, Miccoli R, Rossi E, Vaccaro O, De Rosa M et al. Under-treatment of migrants with diabetes in a universalistic health care system: the ARNO Observatory. Nutr Metab Cardiovasc Dis 2014;24:393-399
18. Clavenna A, Rossi E, Berti A, Pedrazzi G, De Rosa M, Bonati M. Inappropriate use of anti-asthmatic drugs in the Italian paediatric population. Eur J Clin Pharmacol 2003;59: 565-569
19. CINECA-SID. Osservatorio ARNO Diabete. Il profilo assistenziale della popolazione con diabete. In: scientifiche CS-E (ed) Collana Rapporti Arno, Bologna, 2011 20. Pagano E, De Rosa M, Rossi E, Cinconze E, Marchesini G, Miccoli R et al. The
relative burden of complications on healthcare cost of diabetes: the population-based ARNO Diabetes Observatory, Italy. Nutr Metab Cardiovasc Dis in press doi: 10.1016/j.numecd.2016.05.002
21. Carle F, Gesuita R, Bruno G, Coppa GV, Falorni A, Lorini R et al. Diabetes incidence in 0-to 14-year age-group in Italy: a 10-year prospective study. Diabetes Care 2004;27:2790-2796
22. Vichi M, Iafusco D, Galderisi A, Stazi MA, Nisticò L. An easy, fast, effective tool to monitor the incidence of type 1 diabetes among children aged 0-4 years in Italy: the Italian Hospital Discharge Registry (IHDR). Acta Diabetol. 2014;51:287-294
23. Bruno G, Maule M, Merletti F, Novelli G, Falorni A, Iannilli A et al. Age-period-cohort analysis of 1990-2003 incidence time trends of childhood diabetes in Italy: the RIDI study. Diabetes 2010;59:2281-2287
24. Bruno G, Novelli G, Panero F, Perotto M, Monasterolo F, Bona G et al. The incidence of type 1 diabetes is increasing in both children and young adults in Northern Italy: 1984-2004 temporal trends. Diabetologia 2009;52:2531–2535
25. López-de-Andrés A, Jiménez-García R, Hernández-Barrera V, Isabel Jiménez-Trujillo M, Gil-de-Miguel A, Carrasco-Garrido P. Trends in hospitalizations among children with type 1 diabetes in Spain, 2001-2009. J Pediatr Endocrinol Metab. 2013;26:85-89
26. De Berardis G, D'Ettorre A, Graziano G, Lucisano G, Pellegrini F, Cammarota S et al. The burden of hospitalization related to diabetes mellitus: a population-based study. Nutr Metab Cardiovasc Dis. 2012;22:605-612
27. Lombardo F, Maggini M, Gruden G, Bruno G. Temporal trend in hospitalizations for acute diabetic complications: a nationwide study, Italy, 2001-2010. Plos One 8(5): e63675, 2013 28. Cherubini V, Pintaudi B, Rossi MC, Lucisano G, Pellegrini F, Chiumello G et al. Severe
hypoglycemia and ketoacidosis over one year in Italian pediatric population with type 1 diabetes mellitus: A multicenter retrospective observational study. Nutr Metab Cardiovasc Dis 2014; 24:538-546.
29. Ying AK, Lairson DR, Giardino AP, Bondy ML, Zaheer I, Haymond MW et al. Predictors of direct costs of diabetes care in pediatric patients with type 1 diabetes. Pediatr Diabetes 2011;12:177-182
30. Bächle CC, Holl RW, Straßburger K, Molz E, Chernyak N, Beyer P et al. Costs of paediatric diabetes care in Germany: current situation and comparison with the year 2000. Diabet Med 2012;29:1327-1334
