JMIR Publications

ABSTRACT

Background:

Diabetes mellitus prevalence is increasing among adults and children around the world. Diabetes care is complex. Examining the diet, type of medication, diabetes recognition, and willingness to use self-management tools are just a number of the challenges faced by diabetes clinicians who should make decisions about them. The appropriate decisions will reduce the cost of treatment, decrease the mortality rate of patients with diabetes, and improve the life quality of patients with diabetes. Taking appropriate decisions belongs to MCDM techniques. In this paper, we provide a review of the majority of reports on the application of MCDM methods in diabetes and propose a new classification of MCDM techniques in diabetes.

Objective:

In this paper, we provide a review of the majority of reports on the application of MCDM methods in diabetes and propose a new classification of MCDM techniques in diabetes.

Methods:

The PRISMA method was applied to the selection of articles. 56 articles were examined between 2003 and 2022.

Results:

After reviewing, we concluded that applications of MCDM methods for diabetes can be classified into six groups: selection of diabetes medications (16), diagnosis of diabetes (12), meal recommendations (8), diabetes management (11), diagnosis of complications caused by diabetes (7), and estimation of diabetes prevalence (2).

Conclusions:

Our review showed that, although artificial intelligence and machine learning methods could be more accurate than other techniques, MCDM methods are easier to implement and still provide acceptable results and can be used for a wider variety of diabetes challenges.