Abstract Background Hemoglobin A1c (HbA1c) is a widely used lab test to assist in diagnosing and managing diabetes. However, in patients with known disorders of red blood cell turnover and certain hemoglobin variants, this value is inaccurate. Current American Diabetes Association guidelines state that HbA1c should not be used to diagnose diabetes in these individuals. A 2021 publication by Sivasanskar showed there is a significant quality gap with inappropriate HbA1c test orders among those with sickle cell disease (SCD) at hundreds of healthcare facilities across the United States. We suspect this is also the case for glucose 6-phosphate dehydrogenase (G6PD) deficiency and thalassemias known to affect red blood cell turnover. Multiple recent publications have shown that HbA1c levels at all levels of glucose are much lower for those with G6PD deficiency than for controls. Since both SCD and G6PD deficiency are known to occur in African-Americans to a greater extent than Caucasians, we believe this could be a significant issue in our health care system leading to delayed diagnosis or inappropriate management of diabetes in this population. Methods Using retrospective electronic health record data from one children’s hospital, one additional hospital and several outpatient clinics in our health system, we examined 40,000 orders for glucose and alternate markers from 1955 unique patients with an ICD-10 code for SCD, thalassemia, or anemia caused by enzyme deficiency. The protocol for the study was approved by the Washington University IRB. The majority of these patients were diagnosed with sickle-cell disease or trait (79%) while only 11 (1%) were diagnosed with G6PD deficiency and the remainder (20%) with thalassemia. Results Of the total orders examined, 95% were for glucose (either alone or as part of a panel with additional tests). For the remaining 5% of orders, 447 were for glucose tolerance testing (gestational), 21 for glucose tolerance testing (non-gestational), 1487 for HbA1c, and 55 for fructosamine. Twenty-nine percent of patients with SCD or trait had at least one HbA1c order and this percentage was higher (51%) in patients with a thalassemia. For the 11 patients with G6PD deficiency, 36% had a HbA1c order. Fructosamine orders occurred for only 25 unique patients, none of whom had G6PD or other anemia of enzyme disorders. Less than 1% patients with thalassemia and only 1.4% of patients with SCD or trait had an order for fructosamine. For the patients in whom fructosamine was ordered, 80% also had one or more HbA1c orders, including nine instances with both tests ordered during the same week. Conclusion The percentage of patients with a disorder of red blood cells and HbA1c orders is significant while fructosamine is rarely ordered. This suggests opportunities for electronic ordering interventions and/or additional provider education regarding appropriate use of alternate markers of glucose control in this group of patients.
Carson et al. (Wed,) studied this question.