Healthcare claims data are a valuable source, particularly for the epidemiological investigation of chronic conditions including diabetes. Despite these data being designed to collect information on resource utilization, they are frequently used as repositories for research purposes. These databases, including a large sample size, and understanding the advantages, feasibility, accessibility, and low cost could all be factors in estimating the burden of diabetes [1, 2]. However, straightforward analysis and interpretation of healthcare claims data could produce biased information [1, 2].
Authors Huber et al. [3] recently reported a high medical and economic burden of diabetes in Switzerland. In using the healthcare claims data of a large health insurance group in Switzerland, based on prescription data for diabetes, 49,757 patients were identified with diabetes. Further, it was determined that diabetes-related costs increased from 5,036 Euros per patient in 2006 to 5,331 Euros per patient in 2011 in services covered by mandatory healthcare insurance. Before accepting the comments of Huber et al., consideration of the fragility of the prescription database to identify the prevalence and incidence of diabetes should be undertaken, as follows:
Diabetes, particularly Type 2, is a silent disease and individuals may have diabetes for many years before being diagnosed. Diabetes management often begins with life-style modification, and pharmacological treatments are initiated if lifestyle interventions fail to control a blood glucose level. Several studies indicate that patients achieve the recommended glucose level with life-style modification [4, 5].
Some anti-diabetic medications, such as metformin, are prescribed for other conditions. Metformin is occasionally prescribed to delay/prevent Type 2 diabetes in pre-diabetes conditions, e.g., individuals with impaired glucose tolerance or impaired fasting glucose [6]. Metformin is also an effective treatment for Polycystic Ovarian Syndrome [7].
Among patients with diabetes, anti-diabetic medications, including insulin, are sometimes prescribed for a short term in cases of pregnancy or to adjust a glucose level in medical interventions [4]; incidence estimation could be influenced in such cases.
Sometimes, mandatory healthcare insurance systems do not cover all required medications and supplies for diabetes management; without such coverage, both the case identification and estimated cost may be affected [8, 9].
The accuracy of a case definition to identify patients with diabetes, including both prevalent and incident cases of diagnosed diabetes, is a significant aspect of this study. This aspect could explain the differences reported in the prevalence and incidence of diabetes in the researched population compared to other studies in Switzerland, and thus may cause a difference in the estimated cost of diabetes.
In conclusion, when interpreting the findings of this study in light of its limitations, the retrospective medical insurance database of one company in Switzerland, including the mandatory health plan, and the accuracy of the method to identify patients with diabetes using medications ought to be taken into consideration. Furthermore, in highlighting the importance of using a valid procedure to estimate the incidence and prevalence of diabetes using medical databases, it is necessary to consider the medication prescription pattern and the diabetes management guidelines when interpreting the results.
Contribution Statement
Significant contributions to the conception and interpretation of this manuscript were made by all the authors; S. A. is credited for drafting the manuscript, and both O.H. and M.M. are credited for providing revisions and comments. The final version of the manuscript was approved by all authors.
Acknowledgement
The authors of this manuscript acknowledge Ms. Kathleen Murphy for language editing of the manuscript.
Huber CA1, Schwenkglenks M, Rapold R, Reich O. Epidemiology and costs of diabetes mellitus in Switzerland: an analysis of health care claims data, 2006 and 2011. BMC Endocr Disord. 2014 Jun 3; 14:44. doi: 10.1186/1472-6823-14-44.
Canadian Diabetes Association Clinical Practice Guidelines Expert Committee. Canadian Diabetes Association 2013 Clinical Practice Guidelines for the Prevention and Management of Diabetes in Canada. Can J Diabetes 2013;37(suppl 1):S1-S212
Comments on Epidemiology and Costs of Diabetes Mellitus in Switzerland: An Analysis of Healthcare Claims Data for 2006 and 2011
24 July 2014
S. Asghari, O. Hurley, M. Mahdavian
Memorial University of Newfoundland
300 Prince Philip Drive, St. John’s, NL A1A 4N8
Corresponding Author:
Dr. Shabnam Asghari
Memorial University of Newfoundland
300 Prince Philip Drive, St. John’s, NL
A1A 4N8
709-777-2142
Shabnam.asghari@med.mun.ca
Keywords: diabetes, healthcare claims database
Healthcare claims data are a valuable source, particularly for the epidemiological investigation of chronic conditions including diabetes. Despite these data being designed to collect information on resource utilization, they are frequently used as repositories for research purposes. These databases, including a large sample size, and understanding the advantages, feasibility, accessibility, and low cost could all be factors in estimating the burden of diabetes [1, 2]. However, straightforward analysis and interpretation of healthcare claims data could produce biased information [1, 2].
Authors Huber et al. [3] recently reported a high medical and economic burden of diabetes in Switzerland. In using the healthcare claims data of a large health insurance group in Switzerland, based on prescription data for diabetes, 49,757 patients were identified with diabetes. Further, it was determined that diabetes-related costs increased from 5,036 Euros per patient in 2006 to 5,331 Euros per patient in 2011 in services covered by mandatory healthcare insurance. Before accepting the comments of Huber et al., consideration of the fragility of the prescription database to identify the prevalence and incidence of diabetes should be undertaken, as follows:
The accuracy of a case definition to identify patients with diabetes, including both prevalent and incident cases of diagnosed diabetes, is a significant aspect of this study. This aspect could explain the differences reported in the prevalence and incidence of diabetes in the researched population compared to other studies in Switzerland, and thus may cause a difference in the estimated cost of diabetes.
In conclusion, when interpreting the findings of this study in light of its limitations, the retrospective medical insurance database of one company in Switzerland, including the mandatory health plan, and the accuracy of the method to identify patients with diabetes using medications ought to be taken into consideration. Furthermore, in highlighting the importance of using a valid procedure to estimate the incidence and prevalence of diabetes using medical databases, it is necessary to consider the medication prescription pattern and the diabetes management guidelines when interpreting the results.
Contribution Statement
Significant contributions to the conception and interpretation of this manuscript were made by all the authors; S. A. is credited for drafting the manuscript, and both O.H. and M.M. are credited for providing revisions and comments. The final version of the manuscript was approved by all authors.
Acknowledgement
The authors of this manuscript acknowledge Ms. Kathleen Murphy for language editing of the manuscript.
References
Competing interests
The authors of this manuscript declare that no conflict of interest exists.