Errors in the diagnosis of types of diabetes mellitus: causes and prevention strategies (literature review and own research results)

Abstract

The article provides current information regarding medical errors in diagnosing diabetes mellitus (DM), analyzes their factors, and outlines preventive measures. The causes of the most common diagnostic errors in diabetology include limited access to quality healthcare, insufficient training and number of qualified personnel, low quality of teamwork and medical information exchange, inadequate availability of diagnostic tests, poor coordination of care and follow-up, lack of medical informatics resources, human factors, and cognitive biases (such as misidentification of DM and its type, incorrect context creation, overestimation/underestimation of incidental findings, premature termination of the diagnostic process, and lack of knowledge and skills in diabetology). The most typical diagnostic discrepancies involve misidentifying type 1 DM (including latent autoimmune diabetes in adults), type 2 DM, and other specific types of DM. This is due to the increasing heterogeneity of DM, blurring of the boundaries between its types, atypical disease course, the decreased diagnostic value of the essential criteria for DM types (age, presence of metabolic syndrome signs, ketosis, dependency on insulin therapy), presence of comorbid conditions, and limited availability of diagnostic tests to specify the type of diabetes. To optimize diagnosis and prevent diagnostic errors, we have developed a Telegram bot DiaType based on a multilevel algorithm for the differential diagnosis of various types of DM. The testing of this Telegram bot has shown its high effectiveness in identifying DM variants. The advantages of DiaType include accessibility, interactivity, accuracy, and support for medical professionals, which helps improve diagnostic efficiency, simplify the diagnostic process, especially for primary care physicians, prevent diagnostic errors, and, consequently, improve treatment outcomes.

Keywords

• diabetes mellitus
• types
• heterogeneity
• diagnosis
• medical errors
• informational technologies
• chatbot