Presentation (CTI)

OBJECTIVES: This study evaluates the rate of delayed diagnosis of adult-onset type 1 diabetes mellitus (T1DM) and compares the risk of diabetic ketoacidosis (DKA) with direct diagnosis of T1DM.
METHODS: This retrospective cohort study used electronic health records from a large integrated delivery network in Texas. Adults (≥18 years) newly diagnosed with diabetes between January 1, 2016, and January 1, 2025, were included. Patients were classified as delayed if they were initially coded as type 2 diabetes mellitus (T2DM) and subsequently diagnosed with T1DM. T1DM and T2DM status were defined using ICD-10 codes, with patients requiring at least two outpatient or one inpatient encounters with the respective condition. The index date was the first documented diabetes diagnosis, and patients were followed from the index date until lost to follow-up or study end (January 1, 2026). Outcomes were compared between patients with delayed diagnosis vs. direct diagnosis of T1DM. The primary outcome was the frequency of delayed diagnosis. The secondary outcome was the incidence rate of DKA. Descriptive statistics were used to report the outcomes.
RESULTS: Among 1541 patients, 29.7% of T1DM patients had a delayed diagnosis. Delayed diagnosis patients were older (48.3 vs. 40.7 years) and had slightly more males (50.0% vs. 47.3%). The delayed diagnosis group had a higher proportion of non-Hispanic Black individuals (18.0%) compared to the direct diagnosis group (11.5%). Delayed diagnosis patients had a higher rate of DKA events (27.8%) compared to directly diagnosed patients (22.9%). The average follow-up period for the delayed diagnosis group was longer (1973 vs. 1396 days).
CONCLUSIONS: Adult-onset T1DM is less commonly recognized, which may lead to a higher risk of DKA, a life-threatening complication. These findings highlight the increased risk of DKA present in adult patients with a delayed diagnosis of T1DM.