Sodium–glucose co‐transporter 2 inhibitors (SGLT2Is) – The latest residents on the block: Impact on glycaemic control at a general practice level in England

Aims Despite increasing spend on new Type 2 diabetes mellitus (T2DM) therapies, the proportion of people with T2DM achieving target glycaemia outcomes is declining. Our aim was to determine, using published General Practice level data, how differences in T2DM prescribing patterns relate to glycaemic target achievement levels.MethodsMultiple linear regression modelling was used to link practice characteristics and defined daily dose (DDD) of class of medication in 2015/16 and changes to 2014/15 in medication to proportions achieving target glycaemic control (TGC; glycated haemoglobin A1c (HbA1c) ≤7.5%, 58 mmol/mol) and high glycaemic risk (HGR; HbA1c >10.0%, 86 mmol/mol) for practices in the National Diabetes Audit (NDA) with >100 T2DM on their register.ResultsOverall, HbA1c outcomes were not different between the years studied. Although in percentage terms most practices increased their use of Sodium–glucose co‐transporter 2 inhibitors (SGLT‐2Is) (96%), Dipeptidyl peptidase‐4 inhibitors (DPP‐4Is) (76%) and glucagon‐like peptide 1 (GLP‐1) analogues (53%), there was wide variation in use of older and newer therapies. For example, 12% of practices use >200% national average of some newer agents. In cross‐sectional analysis: greater prescribing of metformin and analogue insulin were associated with a higher proportion achieving HbA1c <58 mmol/mol; SGLT‐2Is and metformin were associated with reduced proportion of HbA1c >86 mol/mol; otherwise associations for sulphonylureas, GLP‐1 analogues, SGLT‐2Is and DPP‐4Is were neutral or negative. In year‐on‐year analysis there was ongoing deterioration in glycaemic control which was offset to some extent by increased use of SGLT‐2Is and GLP‐1 analogues, which were associated with a greater proportion achieving HbA1c <58 mmol/mol and a less proportion of people at >86mmol/mol. SGLT‐2I prescribing was associated with significantly greater improvements than those found for GLP‐1 analogues.ConclusionIncreased use of newer agents was associated with improvement of glycaemic outcomes but not sufficient to compensate for prevailing decline. This may reflect wide variability in the prescribing of newer agents. We have found that SGLTIs may display advantage vs other oral agents in relation to HbA1C outcome. Serious consideration should be given to their use.