Steroid-induced diabetes commonly remains undiagnosed until it becomes symptomatic or acute hyperglycaemia develops, while steroid-induced hyperglycaemia in those with pre-existing diabetes is often identified earlier, and usually requires uptitration of glucose-lowering medication. The May 2021 update to the Management of Hyperglycaemia and Steroid (Glucocorticoid) Therapy from the Joint British Societies for Inpatient Care strengthens guidance on managing people who have been discharged on steroids and provides clarity around identifying and managing these conditions in people started on long-term oral steroids in primary care.
When planning long-term oral steroids in practice, in those at high risk of but without diagnosed diabetes, an HbA1c test prior to initiation is recommended to exclude pre-existing type 2 diabetes; however, HbA1c is not reliable for monitoring hyperglycaemia during treatment, as glucose levels may rise rapidly. For once-daily doses of prednisolone >5 mg taken in the morning, once-daily capillary blood glucose testing is recommended prior to or 1–2 hours after lunch or the evening meal, as this is when hyperglycaemia is most likely. If blood glucose is >12 mmol/L, increase capillary glucose monitoring to four times per day. If two or more readings are >12 mmol/L in 24 hours, consider starting gliclazide 40 mg and uptitrate the once-daily dose as required to 240 mg taken in the morning, with an additional evening dose if required, to a maximum daily dose of 320 mg. Reduce gliclazide or insulin as steroids are reduced or discontinued, and continue using capillary glucose monitoring, taking care to avoid hypoglycaemia.
In people with pre-existing type 2 diabetes, when starting oral steroids, consider four-times-daily self-monitoring of capillary blood glucose (SMBG), and if two consecutive readings are >12 mmol/L in 48 hours, additional glucose-lowering treatment should be considered.
Post discharge following an admission, if >5 mg daily prednisolone is planned over the longer term and the patient is on insulin, SMBG should be undertaken at least once daily and before driving. As the steroid dose decreases, reduce glucose-lowering treatment to avoid hypoglycaemia. For example, a 5-mg prednisolone reduction from 20 mg may need a 20–25% insulin reduction or a 40 mg gliclazide reduction. If there was no type 2 diabetes prior to steroids, check HbA1c 3 months after the steroids have been stopped and capillary glucose values have returned to normal, to screen for type 2 diabetes. If diabetes needs confirmation earlier than when the HbA1c test is reliable – for example, if someone is symptomatic – use a fasting blood glucose or oral glucose tolerance test.
The treatment algorithm in those taking oral prednisolone recommends gliclazide as first-line therapy, followed by once-daily night-time insulin or twice-daily insulin if hyperglycaemia is not controlled in type 2 diabetes, with twice-daily or basal–bolus insulin in people with type 1 diabetes. There is no evidence of benefit for metformin, DPP-4 inhibitors, GLP-1 receptor agonists or SGLT2 inhibitors in the management of acute steroid-induced diabetes/hyperglycaemia. Evidence for pioglitazone is weak, and the drug takes several weeks to lower blood glucose. Dexamethasone causes hyperglycaemia throughout the 24-hour period, so insulin is usually required.
Dexamethasone significantly reduces mortality in those with severe COVID-19 who require ventilation or oxygen therapy. These people already have increased insulin resistance and decreased secretion of insulin from beta-cells due to COVID-19, so the added hyperglycaemia from the long-acting dexamethasone requires insulin therapy rather than gliclazide, aiming for glucose levels <10 mmol/L during the acute hospital phase. Following discharge, shared care with the diabetes specialist team to monitor and reduce insulin will be required if this has not been achieved in hospital. There may be an increased risk of type 2 diabetes in the longer term post-COVID-19, so HbA1c monitoring is recommended annually (National Inpatient Diabetes COVID-19 Response Group, 2020).
Click here to access the full guideline. The document is dynamic and will be reviewed and updated as new evidence and feedback becomes available to the authoring team.