Patients with type 2 diabetes treated with SGLT2 inhibitors are more likely to receive a diabetic ketoacidosis diagnosis than those not taking the glucose-lowering medication despite not meeting international criteria, a study suggests.
Australian researchers say their findings have implications for patients who might be unnecessarily missing out on the “important” class of drugs.
The retrospective observational study involved 164 adult patients with type 2 diabetes from two teaching hospitals in Queensland who were diagnosed with diabetic ketoacidosis (DKA) between 2015 and 2022.
At the time of diagnosis, 94 patients were taking SGLT2 inhibitors (45% female, mean age 63) and 70 were not on the medication (53% female, mean age 62).
For “true” diagnoses of DKA, the researchers relied on the published Joint British Diabetes Society (JBDS) and American Association of Clinical Endocrinology/American College of Endocrinology (AACE/ACE) criteria, which they said had been recently updated to account for the physiological effects of SGTL2 inhibitors.
Findings published in the Journal of Diabetes Research [link here] revealed a significantly smaller proportion of patients in the SGLT2i cohort met JBDS (56% vs 72%) and AACE/ ACE (63% vs 82%) criteria compared to non-SGLT2i patients.
However, both groups had a large number of patients who did not fall under the DKA criteria, noted the researchers from Gold Coast Health and Griffith University.
Euglycaemia was observed in 39% and 48%, respectively, of patients in the SGLT2i group versus 3% in the non-SGLT2i group based on JBDS (<11 mmol/L) and AACE/ACE/American Diabetes Association (<13.9 mmol/L) criteria.
Other findings showed significant differences in pH, serum bicarbonate, and eGFR between the SGLT2i and non-SGLT2i groups. Significant differences were also observed in blood glucose levels (median 14.35 mmol/L vs 26.25 mmol/L).
The researchers said the proportion of patients diagnosed using both JDBS and AACE/ACE criteria were comparable, suggesting a reasonable degree of agreement. However, they raised concerns about overdiagnosis in both groups.
“Acidosis is the sine qua non of diabetic ketoacidosis, and the criterion of pH < 7.3 is common to the major society guidelines. This criterion was present in just 79% of the non-SGLT2i group and 68% of SGLT2i cohort, suggesting a tendency toward over-/misdiagnosis of DKA in both groups,” they wrote.
“The differences observed in the severity of the acidaemia between groups may at least partly be explained by several patients being diagnosed in the periprocedural setting, with different criteria employed by the Australian and New Zealand College of Anaesthetists (ANZCA) to diagnose DKA.”
They said this underscored “the need for a more universal, interdisciplinary definition of DKA to avoid potential uncertainty and risk of unnecessary deferral of procedures”.
Hyperglycaemia and ketonuria were no longer reliable tools for the diagnosis of DKA in patients taking SGLT2 inhibitors, they added, because the medications resulted in a “paradoxical” increase in ketonuria, giving the “erroneous impression of worsening ketoacidosis and failure of treatment”.
“As the indications for SGLT2i expand beyond diabetes into chronic kidney disease and heart failure, rigorous adherence to existing diagnostic criteria for DKA will be necessary to avoid overdiagnosis with its attendant risks, including unnecessary deferral of procedures and adverse clinical outcomes associated with overtreatment (e.g., hypokalaemia, hypoglycaemia),” they concluded.
“Indeed, an erroneous label of ‘SGLT2i-related DKA’ or ‘euglycaemic DKA secondary to SGLT2i’ in a patient’s medical history may result in reticence on the part of clinicians to continue or resume treatment with this important class of medications in the affected patients.”