Insulin Use and Poor COVID-19 Outcomes among Diabetes Patients: Association Not Necessarily Causation.

Abstract

Our study looked at the relationship between insulin use and clinical outcomes in COVID-19. A response to our article, written by Dr. Chia Sing Kow and Dr. Syed Shahzad Hasan raised a few questions. They mentioned our use of hemoglobin A1c may be inaccurate as the patients in our study had high rates of CKD or ESRD which could alter the hemoglobin A1c levels. However due to the limitations of our patient population and perhaps in a lot of other sample populations in the real-world setting, it was the most feasible way to represent glucose control.The writers also suggested that the use of metformin, a potential confounder, was also not adjusted for. This should be considered in future research but addition of too many variables in a regression model may lead to less reliability of results for our study.The letter writers also suggested that the results of our paper may lead to misinterpretation by readers and may influence providers to not use insulin therapy for their patients when necessary due to fear of worse outcomes in the setting of COVID-19. We reiterated that it is very important that the data not be misinterpreted, and that nowhere in our paper did we imply or suggest that patients who need insulin therapy to treat their diabetes should not receive proper therapy due to the association we delineated in our paper. Instead, more careful surveillance of patients with advanced diabetes is needed especially when admitted with COVID-19.