How Often Are Emergency Patients Diagnosed With Diabetic Ketoacidosis Despite Not Meeting Laboratory Criteria?
Emergency department (ED) patients with hyperglycemia, acidosis, and/or ketosis may have diabetic ketoacidosis (DKA) or any of several well-described conditions. Though the diagnosis of DKA is based on specific laboratory criteria, DKA often presents with mixed clinical pictures, making a strictly laboratory diagnosis problematic. When laboratories fail to meet criteria, patients may nevertheless be diagnosed with DKA. The extent of this is not reported.
We conducted a retrospective observational study at an urban academic ED (01/01/2019-02/28/2023) of adult patients with point of care (POC) glucose > 300 mg/dL and POC ketone > 1.1 mmol/L and/or an ED diagnosis of DKA. We present the proportion whose initial laboratories met DKA criteria based on one of two laboratory definitions, the proportion whose initial laboratories did not meet criteria but who were nevertheless diagnosed with DKA, and the frequencies of non-DKA diagnoses potentially explaining these laboratory abnormalities. Analyses were descriptive.
Of 1676 patients included, 883 (53%, 95% CI: 50, 55) met lab criteria for DKA. Of 740 screening positive by POC testing whose initial labs did not meet DKA criteria, 229 (31%, 95% CI: 28, 34) were diagnosed with DKA. Primary ED diagnoses of the remaining 511 included: hyperglycemia (196, 38%), starvation ketosis (58, 11%), hyperosmotic hyperglycemic state (11, 2%), and other ketosis (9, 2%), while 67 (13%) had a primary diagnosis of infection, 1 (< 1%) metabolic acidosis and 169 (33%) an unrelated diagnosis.
In this single center study of patients screening positive for DKA or given an ED diagnosis of DKA, 53% met laboratory criteria for DKA, and of those not meeting criteria, 31% were nevertheless diagnosed with DKA. This suggests that emergency physicians use criteria beyond laboratory values to diagnose DKA and supports the idea that DKA is ultimately a clinical rather than a purely laboratory-diagnosed condition.
We conducted a retrospective observational study at an urban academic ED (01/01/2019-02/28/2023) of adult patients with point of care (POC) glucose > 300 mg/dL and POC ketone > 1.1 mmol/L and/or an ED diagnosis of DKA. We present the proportion whose initial laboratories met DKA criteria based on one of two laboratory definitions, the proportion whose initial laboratories did not meet criteria but who were nevertheless diagnosed with DKA, and the frequencies of non-DKA diagnoses potentially explaining these laboratory abnormalities. Analyses were descriptive.
Of 1676 patients included, 883 (53%, 95% CI: 50, 55) met lab criteria for DKA. Of 740 screening positive by POC testing whose initial labs did not meet DKA criteria, 229 (31%, 95% CI: 28, 34) were diagnosed with DKA. Primary ED diagnoses of the remaining 511 included: hyperglycemia (196, 38%), starvation ketosis (58, 11%), hyperosmotic hyperglycemic state (11, 2%), and other ketosis (9, 2%), while 67 (13%) had a primary diagnosis of infection, 1 (< 1%) metabolic acidosis and 169 (33%) an unrelated diagnosis.
In this single center study of patients screening positive for DKA or given an ED diagnosis of DKA, 53% met laboratory criteria for DKA, and of those not meeting criteria, 31% were nevertheless diagnosed with DKA. This suggests that emergency physicians use criteria beyond laboratory values to diagnose DKA and supports the idea that DKA is ultimately a clinical rather than a purely laboratory-diagnosed condition.
Authors
Griffey Griffey, Haas Haas, Schneider Schneider, Suarez Suarez, Kline Kline, Ancona Ancona, Cruz-Bravo Cruz-Bravo
View on Pubmed