[Application of the Chinese Expert Consensus on Diabetes Classification in clinical practice]

Current diabetes classification practices lack standardized implementation. Prior research has shown that diagnostic subtyping remains inconsistent across clinical settings. No prior work had resolved the gap between expert guidelines and real-world diagnostic testing rates. Existing knowledge suggests that C-peptide and autoantibody testing are essential for accurate diabetes classification. However, the extent of guideline adherence remains unclear. This gap motivated a retrospective analysis of diagnostic practices. The study aimed to assess how well clinical practice aligns with expert consensus recommendations. The findings could help improve diagnostic accuracy in diabetes classification.

This study aimed to evaluate how well clinical practice aligns with the 2022 Chinese Expert Consensus on Diabetes Classification. The goal was to assess diagnostic testing rates for diabetes subtypes. The researchers focused on newly diagnosed diabetes patients. They wanted to determine if clinical features matched consensus criteria. The study also aimed to test the diagnostic accuracy of the consensus-based screening strategy. The researchers sought to identify barriers to guideline implementation. They analyzed the use of C-peptide and autoantibody testing in real-world settings. The findings could inform improvements in diabetes diagnostic protocols.

The study used a retrospective case series design. Researchers collected data from electronic medical records at a single hospital. They included patients with new-onset diabetes diagnosed within one year. The sample size totaled 3,384 patients. Patients were categorized based on age, BMI, and T1DM suspicion. The chi-square test compared subgroups for diagnostic markers. A logistic regression model assessed the diagnostic value of clinical features. The area under the receiver-operating curve (AUC) evaluated model accuracy.

Only 36.6% of patients underwent C-peptide testing. A similar 37.5% completed GADA antibody testing. Younger onset and non-obese patients showed higher GADA detection rates. Patients suspected of T1DM had higher antibody positivity rates. The consensus diagnostic pathway classified only 57.4% of patients. Many required additional testing for definitive subtyping. Logistic regression identified young onset and ketosis as significant predictors. The AUC of 0.77 suggested good diagnostic accuracy for T1DM screening.

The study found a mismatch between clinical practice and consensus guidelines. Low testing rates for subtyping indicators were observed. The consensus diagnostic pathway classified only half of the patients. Improving detection rates for C-peptide and autoantibodies is essential. The T1DM screening strategy showed promising diagnostic value. Clinical features like young age and ketosis were strong predictors. The logistic regression model supported the consensus criteria. These findings suggest a need for better guideline implementation.