080 Screening for diabetes and steroid-induced hyperglycaemia in patients with rheumatoid arthritis is done poorly

Abstract

Background: Recent UK epidemiological research suggests a hazard ratio of 1.5 for diabetes in RA patients on prednisolone ≤5mg or daily. National guidelines produced by NICE, ARMA and BSR set clear standards for screening for diabetes and other cardiovascular risk factors in patients with RA. Further, the Joint British Diabetes Societies have recently produced guidelines for the management of hyperglycaemia in receiving steroid therapy during hospital admission. We therefore assessed our current levels of screening for diabetes against these standards. Methods: The audit was conducted in two rheumatology departments in a district general hospital outpatient setting. Data was collected prospectively using a proforma over a 3 month period in randomly selected patients with RA either commencing or established on oral prednisolone. Results: Data on a total of 82 patients from both hospitals was analysed. Of these 34% were female; mean age 66yr (30-89); 97% were white Caucasian; 14/82 (17%) were know diabetics (all type II, with the exception of two patients). Twenty patients were either commencing prednisolone for the first time or had taken it for <6months whilst 37/82 (45%) had been established on prednisolone for >12months. The mean starting dose for prednisolone was 9.5mg (range 5-30mg). Glycated haemoglobin (HbA1c) was measured in 35/ 82 (43%) of patients before and in 40/82 (48%) established on prednisolone. The mean HbA1c prior to treatment was 47.3mmol/mol (normal <48) and the mean whilst on prednisolone was 45.6mmol/mol. Serum glucose was measured in 34/82 (41%) of patients prior to commencing prednisolone but in only 15/82 (18%) of patients established on prednisolone. The mean pre-steroid serum glucose was 6.3 and post-steroid was 6.9mmol/l. Measurement of HbA1c and plasma glucose before and during steroid therapy was undertaken in all patients with known diabetes. No new cases of steroid-induced diabetes were diagnosed as per current HbA1c criteria, although several cases of impaired glucose (HbA1c 42-47) were observed. Conclusion: This two centre audit suggests that screening for steroid induced diabetes in our non-diabetic RA patients is done poorly, with less than half of patients being assessed. This has clinical and medicolegal implications, particularly in view of current guidelines suggesting the early use of corticosteroids in early inflammatory arthritis. Testing was better in patients with known diabetes, as might be expected. However, no new cases of steroid-induced diabetes were identified in screened patients, although this was the minority of patients. Surprisingly blood glucose control, as assessed using HbA1c, was little affected in patients receiving long-term low-dose oral prednisolone, perhaps due to more intensive blood glucose monitoring and dose adjustment of oral hypoglycaemics. Despite these observations, there is a need for much more vigilance in looking for for the development of hyperglycaemia in patients with RA and other inflammatory arthropathies.