Background: Despite growing evidence that diabetes is associated with worse COVID-19 outcomes, there remains little information on inpatient glycemic response among individuals with diabetes and uncontrolled hyperglycemia hospitalized with the novel coronavirus SARS-CoV-2.

Methodology: This is a multicentric retrospective study aimed to explore the role of hyperglycaemia (uncontrolled HbA1c) in predicting length of hospital stay and ventilatory support in diabetic patients hospitalised for COVID-19. We excluded all patients with newly diagnosed diabetes within last 3 months and/or those who are only on lifestyle modification.

Results: A total of 133 patient records admitted to three tertiary care hospitals between 15/05/2020 to 15/11/2020 were analysed retrospectively. The 62 patients requiring ventilatory support had a significantly higher HbA1c (9.49±2.73%) as compared to those without ventilatory support (8.09±1.32%), p=0.002. Multiple logistic regression also indicated that the HbA1c was an independent risk factor for ventilatory support in DM patients with COVID-19 (Odds Ratio [OR] = 1.310, 95% confidence interval [CI]: 1.063-1.615), p=0.011. The odds of requiring ventilatory support is 35% lower in patients with good glycemic control (HbA1c<7.5%) vs. those with HbA1c≥7.5% (OR-0.65, 95%CI 0.32-0.88). The mean duration of hospital stay was 10.5±3.45 days. Multiple linear regression analysis revealed HbA1c to predict the duration of hospital stay, p=0.001. For every 1% increase in HbA1c, the length of hospital stay was found to increase by 0.461 days.

Conclusion: The present study demonstrated that a high level of HbA1c was an independent risk factor for ventilatory support and prolonged hospital stay in COVID-19 patients with DM.

Disclosure

S. Roy chaudhuri: None. A. Majumder: None. D. Sanyal: None. A. Biswas: None. K. Bhattacharjee: None.

Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. More information is available at http://www.diabetesjournals.org/content/license.