(A-316) Risk of Diabetic Ketoacidosis Across Stages of Chronic Kidney Disease in Patients with Type 1 Diabetes in the U.S. Food and Drug Administration’s Sentinel System

Background: Diabetic ketoacidosis (DKA) is a life-threatening complication of type 1 diabetes (T1D). There is a paucity of information in literature describing the risk of DKA in patients with T1D with different chronic kidney disease (CKD) stages.

Objectives: To (1) describe patient characteristics and incidence rate of DKA and (2) estimate risk of DKA in patients with T1D across CKD stages.

Methods: We identified outpatients with T1D who received short/rapid-acting insulin from March 2013 to February 2024 among 6 Data Partners in the U.S. Sentinel System. Patients had (1) medical and drug coverage, (2) a plurality (>50%) of diagnosis codes specific to T1D, and (3) no dispensing of noninsulin antidiabetic drug (except metformin) in the 365 days before the index dispensing of the insulin.
DKA events were identified using a diagnosis code (ICD-9-CM 250.1x or ICD-10-CM E1x.1x) in any diagnosis position in inpatient or emergency departments.
Using a code-based algorithm that was validated against estimated glomerular filtration rate (eGFR, mL/min/1.73 m2), CKD stage at baseline was classified into stages 1 or 2 (eGFR ≥60), 3 (eGFR 30-59), and 4 or 5 (eGFR < 30, including dialysis).
Follow-up began at the index dispensing and ended at the first occurrence of end of exposure, death, end of available data, a DKA event, switch to a different CKD stage, or day 365 since follow-up.
We report the incidence rate of DKA across CKD stage. Using CKD stage 1 or 2 as reference, we estimated crude hazard ratios (95% confidence intervals [CIs]) for DKA in CKD stage 3 and in CKD stage 4 or 5.

Results: Of the 683,436 patients with T1D, 50.2% were male, 89.1% had CKD stage 1 or 2 (mean [±standard deviation] age 39 ±16 years), 5.6% had CKD stage 3 (mean age 62 ±14 years), and 5.3% had CKD stage 4 or 5 (mean age 54 ±15 years). Baseline insulin pump use was recorded in 22.0% (CKD stage 1 or 2), 10.8% (CKD stage 3), and 6.7% (CKD stage 4 or 5) of the cohort; and history of DKA in 13.5% (CKD stage 1 or 2), 12.8% (CKD stage 3), and 19.1% (CKD stage 4 or 5).
We identified 17,530 DKA events in 608,462 patients (167,338 person-years [PYs]) with CKD stage 1 or 2; 1,381 events in 38,583 patients (9,882 PYs) with CKD stage 3; and 2,384 events in 36,391 patients (8,687 PYs) with CKD stage 4 or 5. Incidence rate of DKA was 10.5 (CKD stage 1 or 2), 14.0 (CKD stage 3), and 27.4 events (CKD stage 4 or 5) per 100 PYs.
Compared to patients with CKD stage 1 or 2, those with CKD stage 3 had a 1.56-fold (95% CI: 1.47, 1.65) increased risk of DKA, and those with CKD stage 4 or 5 had a 2.84-fold (95% CI: 2.71, 2.97) increased risk.

Conclusions: In patients with T1D, the crude risk of DKA in one-year follow-up increased numerically with advancing CKD stage. The contribution of CKD to DKA risk in patients with T1D merits further study.