A-079 Evaluation of American Diabetes Association diagnostic criteria for diabetic ketoacidosis in emergency room patients using point-of-care ketone measurements

Abstract Background Recent guidelines from the American Diabetes Association suggest that a glucose = 200 mg/dL, the presence of metabolic acidosis (pH =7.3), and a beta hydroxybutyrate (B-OHB) = 3 mmol/L are diagnostic for diabetic ketoacidosis (DKA). The determination of B-OHB is especially useful in identifying patients in DKA that are euglycemic (glucose < 200 mg/dL). Known instigators of euglycemic DKA include the use of GLP1 and SGL2 inhibitors, two antihyperglycemic agents that are increasingly prescribed to diabetic patients. The purpose of this study was to assess both the frequency in which patients with concern for DKA meet the new clinical criteria, and to assess the agreement relative to physician diagnosed DKA. Methods Patients presenting to the ED from 1/1/25-1/31/25 to with physician ordered POC ketones were included. B-OHB was assessed on an Abbott Precision Pro at the point of care. Results were extracted from the laboratory information system. Chart review was performed for all patients for diabetes diagnosis, and physician diagnosed acidosis at the time of their encounter. Clinical diagnosis of DKA was made at the discretion of the treating physician. Lab values closest to B-OHB were extracted including anion gap, glucose, HbgA1c, pH, and HCO3. Any use of medications known to influence glucose and ketone concentrations were included (i.e GLP1 inhibitors, SGL2 inhibitors, diuretics, etc.,). DKA criteria from clinical guidelines were assessed relative to physician diagnosed of DKA. Results There were 278 patients enrolled, of which 32 had type 1 diabetes (T1D), 243 had type 2 diabetes (T2D), and 3 had unspecific diabetes type. There was a correlation between B-OHB and anion gap (spearman r = 0.48; 0.37 to 0.58) HCO3 (-0.5; -0.64 to -.032) and glucose (0.27; 0.06 to 0.46). There were 34 patients that were diagnosed clinically with DKA with a median B-OHB of 4.3 (IQR:1.3-5.7) and 244 without DKA with a median of 0.2 (0.1-0.4). B-OHB had an ROC of 0.95 (95% CI: 0.92-0.98) for DKA. At a threshold of 3.0 mmol/L, the positive percent agreement with clinical adjudication was 61.8% (45.0-76.1), the negative percent agreement was 98.4% (95.9-99.4), the PPV was 84% and the NPV was 94.9%. In total there were 14 patients (5.3%) that met guideline definitions for DKA, all of which had a physician diagnosis of DKA. There were additional 20 patients that were diagnosed with DKA but did not fulfill criteria. Of these, 4 patients were diagnosed with euglycemic DKA, with an average glucose of 116 mg/dL and ketone concentrations of 6.1, 4.2, 1.3, and 0.8 mmol/L respectively. 3 of the 4 were observed to be on Empagliflozin. Conclusion The prevalence of DKA as defined by clinical guidelines is relatively rare in patients with clinical suspicion. However, many cases that were adjudicated clinically as DKA did not meet guideline definitions. Euglycemic DKA is rare, and of those diagnosed clinically, only 2 met the ketone threshold of 3 mmol/L, suggesting different thresholds may be necessary.