Comorbidity Patterns Analysis in Patients with Thyroid Disease Using Large-Scale Electronic Medical Records: Network-Based Study (Preprint)

Abstract

BACKGROUND Thyroid disease (TD) is a prominent endocrine disorder and an increasing worldwide public health concern, yet the phenotypic comorbidity pattern of TD patients remains unclear. OBJECTIVE We aimed to use the network-based method to systematically analyze and provide a complete picture of the comorbidity patterns for TD patients. METHODS In this retrospective observational study, we extracted comorbidities of nearly 20 thousand adults diagnosed with TD from a tertiary hospital in China, from 2018 to 2022. All comorbidities were identified by ICD-10 (International Classification of Diseases, 10th Revision) codes at 3 digits, and we focused on comorbidities with >2% prevalence. We divided patients into several subgroups by sex, age, and disease type (thyroid cancer [TC] or benign TD [BTD]). Phenotypic comorbidity network (PCN), where comorbidities were used as nodes and their significant correlations as edges, was constructed across all TD patients and different subgroups. The comorbidity associations and differences in the PCN of each subgroup were analyzed and compared. RESULTS The final cohort included 18,311 TD patients with 72 comorbidities. Most patients were female (61.1%), and 6.2% had TC. The mean age of TD patients was 57.2 years. Approximately one-third (31.2%) of patients had less than three comorbidities, and more than half (54.0%) had at least five comorbidities. Except nontoxic goiter and hypothyroidism that with the highest prevalence (67.8% and 34.0%, respectively), other most prevalent comorbidities were hypertension (35.1%), liver disease (22.0%), and diabetes (16.8%). The prevalence of single comorbidity was different in each subgroup. TD patients’ PCN contained 72 comorbidities and 492 comorbid disease pairs. Comorbidities were closely associated with cardio-cerebrovascular diseases and diabetes. Males and females shared 103 disease pairs with 55 comorbidities. Male-female disparities occurred in diseases coexisting. Patients with BTD had more comorbidities with higher prevalence and more complex disease coexistence relationships, whereas those with TC had a few co-existing severe diseases. CONCLUSIONS TD patients had complex comorbidity patterns, especially with cardio-cerebrovascular diseases and diabetes. Comorbidities’ associations among different TD subgroups were various. This study is expected to improve the understanding of TD patients’ comorbidity patterns and enhance the integrated management of patients.