Endocrine Abstracts

Introduction: In the UK, the North West of England is one of the worst affected regions by COVID 19, with over 275 cases/100,000. We retrospectively looked at the diabetes characteristics of consecutive patients at our trust admitted with COVID 19 and had known diabetes, and assessed whether there was a correlation with the outcome, Clinical Frailty Score (CFS) and age.

Methods: 88 continuous patients with a prior diagnosis of diabetes and COVID 19 swab positive requiring inpatient stay were included. Primary care records provided information on HbA1c. Inpatient records were interrogated to obtain information on CFS (1–9), diabetes drugs and outcome. A Cox’s proportional hazards regressional model was used to evaluate the results.

Results: One patient was excluded. Male:female ratio 54:33. Mean age 75.9 years. Mean CFS 5.06. At time of analysis, 43 patients had died, 39 discharged and 5 still inpatients. 68 patients had a history of hypertension. No obesity data was available. Mean HbA1c was 53.2 mmol/mol. 38 patients took metformin, 22 DPP4-inhibitors, 12 sulphonylureas, 6 GLP-1 analogues, 5 SGLT2-inhibitors, 1 pioglitazone, and 12 insulin. 2 patients were admitted with glycaemic emergency (1 DKA, 1 hypoglycaemia). An increase in HbA1c (P = 0.034) and a CFS < 5 (P = 0.0285) were associated with an increased risk of mortality. A 1 mmol/mol increase in HbA1c was associated with a 10.8% increase in mortality (CI 1.008–1.2192). No correlation between age and HbA1c, and CFS and HbA1c were found.

Discussion: This study adds to the evidence base regarding the importance of diabetes control and COVID outcome. A number of questions arise however, including the role of non-diabetic hyperglycaemia and outcome, and the balance of very tight control vs hypoglycaemia. Prospective interventional studies are also required to look the impact of various levels of inpatient glucose control and outcomes. There was insufficient data to look at the impact of individual diabetes drug classes. The unexpected finding of a lower CFS score (ie less frail people) having a higher mortality (converse to expectation) may be explained by a number of factors, including a low number of patients with low CFS, subjectivity regarding the CFS by many different clinicians on assessment, and a small sample size. It is also possible that those less frail patients who need hospital admission are generally very unwell, as compared to older multimorbid frailer patients who may be admitted with lower levels of illness.

Many thanks to Luigi Sedda, Alison Unsworth & Abdul Ashish