This study has several limitations. First, this was a retrospective study. Second, we did not cover HbA1c, a long-term glycaemic control indicator that helps distinguish patients with poor long-term glycaemic control from those with stress hyperglycaemia. Finally, because our results were premised only on the glucose levels at admission, we might have underestimated the risks associated with hyperglycaemia (assuming that patients with the highest glucose levels are more likely to be treated in the hospital), and we did not have sufficient data to study the effect of glucose-lowering treatment (e.g. insulin, metformin) on the outcome of our patients. However, we believe that acute hyperglycaemia is more important than long-term glycaemic control in predicting the prognosis of hospitalised COVID-19 patients.