gms | German Medical Science

64. Jahrestagung der Deutschen Gesellschaft für Medizinische Informatik, Biometrie und Epidemiologie e. V. (GMDS)

Deutsche Gesellschaft für Medizinische Informatik, Biometrie und Epidemiologie

08. - 11.09.2019, Dortmund

Analyzing Relations between Antidiabetic Drugs using the Diagnostic Spectrum of the Related Patients with a Graph Theoretic Method and a Markov Model

Meeting Abstract

  • Reinhard Schuster - MDK Nord, Lübeck, Germany; Universität Lübeck, Lübeck, Germany
  • Marc Heidbreeder - MDK Nord, Lübeck, Germany; Universität Lübeck, Lübeck, Germany
  • Timo Emcke - Kassenärztliche Vereinigung Schleswig-Holstein, Bad Segeberg, Germany
  • Martin Schuster - Christian-Albrechts-Universität zu Kiel Institut für Epidemiologie, Kiel, Germany

Deutsche Gesellschaft für Medizinische Informatik, Biometrie und Epidemiologie. 64. Jahrestagung der Deutschen Gesellschaft für Medizinische Informatik, Biometrie und Epidemiologie e.V. (GMDS). Dortmund, 08.-11.09.2019. Düsseldorf: German Medical Science GMS Publishing House; 2019. DocAbstr. 99

doi: 10.3205/19gmds145, urn:nbn:de:0183-19gmds1452

Published: September 6, 2019

© 2019 Schuster et al.
This is an Open Access article distributed under the terms of the Creative Commons Attribution 4.0 License. See license information at http://creativecommons.org/licenses/by/4.0/.


Outline

Text

We consider a neighborhood relation between antidiabetic drugs classified with respect to the international ATC (Anatomic-Therapeutic-Chemical) system and the related therapeutic spectrum. In order to determine increased patient risks we use those diagnoses on three digit ICD-10 level (International Statistical Classification of Diseases and Related Health Problems) which occur more often for diabetes patients compared to the whole population in the dataset. We analyze all treatments and prescriptions of physicians for patients of the statutory health insurance (SHI) by SHI physicians in Schleswig-Holstein in the first and second quarter of 2018. The largest increase of relative risks for diabetes patients in comparison with all patients of the population are given for “Polyneuropathy in diseases classified elsewhere” (G63), “Obesity” (E66), “Disorders of purine and pyrimidine metabolism” (E79), “Chronic ischaemic heart disease” (I25) and “Disorders of lipoprotein metabolism and other lipidaemias” (E78). Multimorbidity and polypharmacy are a major problem for elderly patients. Above 70 years one has more than seven drug groups at four digit ATC level on average. Using the vector of the fraction of patients with those diagnoses with respect to the related antidiabetic drug we use the Manhattan distance in order to determine the therapeutic most similar drugs. The drugs are used as nodes of a graph and edges are given by therapeutic neighborhood. We determine graph clusters using the modularity method. The related algorithm lead to an ILP (integer linear program) which is in general NP-hard (NP: no computations are possible in polynomial time). Using a related LP (linear program) and post processing applying Mathematica (Wolfram Research) we get graph communities with different levels of resolution. Community structures of graphs give new insights in therapeutic backgrounds of prescribed drugs. This offers the possibility for improved health care decisions at the negotiation level, improved medical decisions from a patient centered point of view and adapted national and international guidelines from a unified point of view. We consider changes of drug therapy between 2013 and 2018 and analyze the time stability of the drug therapy in 2018 the by a Markov equilibrium using data of successive quarters.

The authors declare that they have no competing interests.

The authors declare that an ethics committee vote is not required.


References

1.
International Diabetes Federation. Diabetes Atlas. 8. Auflage. Brüssel: International Diabetes Federation; 2017.
2.
Shi Y, Hu FB. The global implications of diabetes and cancer. The Lancet. 2014 Jun 7;383(9933):1947-8.
3.
World Health Organization. Report on Diabetes. 2016 [Accessed 19 July 2019]. Available from: https://apps.who.int/iris/bitstream/handle/10665/204871/9789241565257_eng.pdf External link
4.
Songer TJ, Ettaro L. Economics of Diabetes Project Panel: Studies on the cost of diabetes. Atlanta, GA: Centers for Disease Control and Prevention; 1998.
5.
American Diabetes Association. Economic consequences of diabetes mellitus in the US in 1997. Diabetes care. 1998 Feb 1;21(2):296-309.
6.
Kobrin Klein BE. Overview of epidemiologic studies of diabetic retinopathy. Ophthalmic epidemiology. 2007 Jan 1;14(4):179-83.
7.
Agarwal G, Kempe D. Modularity-maximizing graph communities via mathematical programming. The European Physical Journal B. 2008 Dec 1;66(3):409-18.
8.
Charikar M, Guruswami V, Wirth A. Clustering with qualitative information. Journal of Computer and System Sciences. 2005 Oct 1;71(3):360-83.
9.
Schuster R, Emcke T. Graph-theoretic network analysis of interactions between patients, physicians and prescribed drugs. Submitted.
10.
Schuster R, Ostermann T, Emcke T. Modeling (Multi-)Morbidity and (Poly-)Pharmacy in Outpatient Treatment with Gamma Distributions. In: Roman M, Fred A, Gamboa H, editors. Proceedings of the 12th International Joint Conference on Biomedical Engineering Systems and Technologies, 2019 Feb 22-24, Prague, Czech Republic. Groningen: SciTePress; 2019. p. 353-358.