Are there disparities in prescribing of sodium-glucose co-transporter-2 (SGLT2) inhibitors in those at high risk of cardiovascular events? A cross-sectional analysis of a national primary care database

Problem

In the Type 2 diabetes population, there exists a higher risk sub-group of cardio vascular patients. These include older people, those with a longer duration of the disease, and impaired renal function. This high risk group has been the focus of several large cardiovascular outcome trials (CVOTs), which have reported cardiovascular benefit in addition to safety, for the use of sodium-glucose co-transporter-2 inhibitors when compared to placebo added to standard care.

The aims of this study were to identify the proportions of people prescribed sodium-glucose co-transporter-2 (SGLT2) inhibitor real world use based on the eligibility criteria for each of these trials. We also sought to explore differences in prescribing within these groups according to gender, ethnicity, and socioeconomic status.We describe three facets of mismatch: (1) Selectivity of the trial population, particularly previous cardiovascular disease; (2) Difference in characteristics between trial and real world population with equivalent cardiovascular risk; and (3) The proportion of eligible patients that were actually prescribed SGLT2 inhibitors and any disparity between those prescribed SGLT2s.

Approach

A cross-sectional analysis was performed of the Royal College of General Practitioners (RCGP) Research and Surveillance Centre (RSC) network database. This is a nationally representative primary care database, English primary care is a registration based system with complete prescribing data. We report: (1) Differences between trial eligible populations; (2) The characteristics of the eligible population registered in primary care compared with trial participants reporting demographics, disease course, and vascular risk; and (3) Any disparities within the eligible population of who is actually prescribed.

Findings

(1) There is variability between the CVOTs, principally on the degree of vascular risk; (2) The proportion people with type 2 diabetes meeting the inclusion criteria for each trial were: CANVAS 14,227 (16.9%); DECLARE 30,498 (36.1%); EMPA-REG 5,628 (6.7%); VERTIS 6,119 (7.3%). The practice populations were older by 6 to 8 years; (3) Less than 10% of people that met the inclusion criteria of each trial had been prescribed an SGLT2 inhibitor. Male gender (OR 1.30, 95% C.I. 1.131 - 1.489) and white ethnicity (OR 1.23, 95% C.I. 1.009 - 1.501) were associated with greater odds odd of prescribing across all CVOTs, with socioeconomic status associated with increased odds in two trials , (OR 1.23, 95 % C.I. 1.115 - 1.500). Black ethnicity however, was associated with reduced odds of SGLT2 inhibitor prescribing (OR 0.33, 95% C.I. 0.175 - 0.561).

Consequences

There is variation in the stringency of recruitment into CVOTs for SGLT2s; this links to the size of the eligible real world population. However, despite evidence of superiority or non-inferiority , uptake is low and there are disparities in the population included.