Continuous glucose monitoring (CGM) has emerged as a promising approach for improving glycemic management in hospitalized individuals, with accumulating evidence supporting its role in reducing hypoglycemia and improving glycemic control. However, the inpatient setting presents unique challenges, including hemodynamic instability, surgical interventions, radiologic procedures, and potential drug interference, which may affect CGM accuracy and safety. International consensus guidelines on the use of CGM in hospitalized patients have not yet been established, and data from Korean inpatient settings are even more limited, highlighting the need for locally relevant evidence synthesis and guideline development. Importantly, the clinical benefits of CGM are not consistently observed when used in isolation. Evidence indicates that CGM is most effective when integrated with CGMbased insulin-titration protocols and electronic medical record systems and supported by dedicated glycemic management teams. In this position statement, the Korean Diabetes Association Digital Health Committee reviews the current evidence on the efficacy and accuracy of CGM in hospitalized patients and highlights key considerations for its safe implementation, including patient selection, hybrid monitoring strategies, and procedural safety. We propose practical recommendations and an operational framework to support the safe and effective use of CGM in inpatient care, thereby establishing a foundation for its systematic adoption in Korea and informing institutional implementation and future policy development.
With the widespread adoption of CGM (continuous glucose monitoring), glycemic control can now be evaluated across a continuous time axis rather than at isolated time points. This shift has led to the emergence of time-based glycemic metrics, which quantify the proportion of time glucose levels remain within predefined ranges and provide a more comprehensive assessment of glycemic exposure. Among these metrics, time in range (TIR; 70–180 mg/dL) has become the standard indicator for overall glycemic control, as it integrates both hypoglycemic and hyperglycemic exposure and shows robust associations with microvascular complications independent of HbA1c. However, TIR alone may be insufficient to discriminate qualitative differences in glycemic control among individuals who have already achieved stable or near-target glucose levels. In this context, time in tight range (TITR; 70–140 mg/dL), reflecting glucose levels closer to normal physiology, has been proposed as a complementary metric. This article discusses the conceptual framework and clinical implications of time-based glycemic metrics, with a particular focus on the distinct and complementary roles of TIR and TITR.
Alongside time in range (TIR, 70–180 mg/dL), an established continuous glucose monitoring (CGM) metric, time in tight range (TITR, 70–140 mg/dL) was originally proposed to reflect glucose exposure closer to physiological normoglycemia, although its clinical applicability has often been considered limited to intensive glycemic control settings. Physiological CGM data show that individuals with normoglycemia spend most of their time below 140 mg/dL, and longitudinal studies in high-risk populations demonstrate that time above 140 mg/dL predicts progression to diabetes. While concerns have been raised that increases in TITR may coincide with higher glycemic variability, particularly in hyperglycemic states, this interpretation is context dependent. In recent analyses, at an HbA1c level of approximately 7.0%, TIR varied widely according to glycemic variability, whereas TITR converged within a relatively narrow range (about 47~48%), indicating a variability-independent reference zone. Moreover, TITR outperformed TIR in predicting attainment of the HbA1c 7.0% target. TITR is already widely recognized as a superior metric in settings of tight glycemic control. Beyond this established role, emerging evidence suggests that TITR also provides complementary and clinically meaningful information across common glycemic targets and hyperglycemic states, thereby enhancing interpretation of glucose distribution beyond conventional CGM metrics.
Type 2 diabetes is increasingly recognized as a heterogeneous syndrome rather than a single disease. Clinical phenotype-based clustering, exemplified by the Ahlqvist model, has demonstrated that subgroups defined by simple variables exhibit distinct complication risks and treatment responses, yet this approach is limited by information loss due to categorization, static classification, and poor transferability across ethnicities. Genome-wide association studies have identified more than 1,200 independent risk signals and revealed genetically driven pathophysiological clusters enriched in cell-type-specific regulatory regions. In addition, partitioned polygenic risk scores show significant associations with vascular complications across ancestries. Multi-omics integration—spanning metabolomics, transcriptomics, epigenomics, and the gut microbiome—has further uncovered subtype-specific signatures that single-omics approaches cannot capture, though the translational gap between complex omics panels and clinically actionable biomarkers remains substantial. Artificial intelligence (AI) and machine learning methods, including ensemble models, deep survival analysis, foundation models trained on continuous glucose monitoring data, and multimodal fusion architectures, are enabling the integration of high-dimensional, heterogeneous data for complication prediction with performance surpassing conventional risk calculators. Recent studies combining polygenic risk scores with clinical variables in deep-learning frameworks have demonstrated significant improvements in predicting cardiovascular and renal complications, with evidence that genetic risk modifies the benefit of standard interventions. However, most models have been developed in European-ancestry populations, and their predictive accuracy diminishes substantially when applied to East Asian populations, where nonobese phenotypes and beta-cell dysfunction predominate. Korea possesses unique strengths to address these challenges, including nationwide health-insurance data, population-based genomic cohorts, and regulatory-approved AI-based retinal biomarkers that could serve as platforms for integrating genomic information. Realizing AI-driven precision diabetes care will require the concurrent development of population-specific prediction models, prospective multi-institutional validation, and robust data-governance frameworks.
While many healthcare professionals and diabetes educators recognize the critical need to screen for and intervene in health-related social needs (HRSNs) among patients with diabetes, the practical implementation of screening and referral to community social services remains limited in clinical settings. Various obstacles hinder effective screening, including a lack of community resources, time constraints for educators, limited reimbursement systems, workforce shortages, and a lack of specialized training. Beyond these systemic issues, a significant barrier is the shame and stigma patients may experience when discussing sensitive socioeconomic vulnerabilities. The process of screening for HRSNs should be viewed not merely as data collection, but as a therapeutic intervention designed to deeply understand the patient’s life and build a foundation of trust. This article introduces the “empathic inquiry” approach, developed by the Oregon Primary Care Association (OPCA), as a strategy to avoid shame and stigma while mitigating psychological resistance during the screening and referral process. Engaging in empathetic dialogue regarding social needs is a valuable therapeutic tool in its own right, fostering patient engagement and strengthening therapeutic alliances even when immediate external resource linkage is not possible.
