In people with type 2 diabetes (T2D) and high cardiovascular (CV) risk, sodium–glucose cotransporter 2 inhibitors (SGLT2i) provide CV and kidney clinical benefits, with reduced rates of death and hospitalization for heart failure (1). As a result, the most recent guidelines from the European Society of Cardiology recommend the initiation of an SGLT2i and a glucagon-like peptide 1 receptor agonist (GLP-1 RA) as first-line therapies for people with T2D and with or at high risk of atherosclerotic CV disease, independent of hemoglobin A1c level, hemoglobin A1c target, or background use of metformin, with similar recommendations from the American Diabetes Association and European Association for the Study of Diabetes (2,3). Despite these recommendations, real-world evidence has been lacking in assessment of the impact of SGLT2i use on health care resource utilization and cost among people with T2D in the U.S. health care system, potentially impacting the use of these therapies (4).
The current analyses of the Empagliflozin Comparativeness and Safety (EMPRISE) study, with data from U.S. Medicare fee-for-service beneficiaries with T2D, include estimation of associations between use of the SGLT2i empagliflozin and short-term health care resource utilization and cost, with propensity score matching to compare costs between individuals initiated on either empagliflozin or a dipeptidyl peptidase 4 inhibitor (DPP-4i), a common antihyperglycemic alternative for management of T2D in case of a neutral CV risk/benefit profile (Fig. 1). In their article in this issue of Diabetes Care, Htoo et al. (5) report the results from analyses of data from >46,000 older U.S. adults with T2D, assessing 1) health care utilization, including inpatient hospitalizations, emergency department visits, and outpatient office visits, and no. of prescribed drugs, as well as 2) costs associated with inpatient medical care, outpatient medical care, and prescription drug use and total costs across the spectrum of care provided. The authors find significantly less health care utilization overall among individuals initiating empagliflozin rather than a DPP-4i, including fewer days spent in inpatient care, fewer emergency department visits, and fewer office visits. Furthermore, empagliflozin was associated with lower inpatient and outpatient costs, as well as lower cost of care overall. These results were driven by individuals with known CV conditions, though individuals without CV conditions also had a numeric reduction in overall health care costs.
Health care resource utilization with empagliflozin: findings from the EMPRISE study. ED, emergency department.
Health care resource utilization with empagliflozin: findings from the EMPRISE study. ED, emergency department.
The authors are to be congratulated on their work, which supports the use of empagliflozin for both clinical benefit and in the management of health care resource utilization and cost. As expected, much of the improvement seen in such metrics associated with use of empagliflozin seems to be derived from data of individuals with CV conditions. Given that older adults with T2D are at high risk for CV and kidney complications, this population might derive the maximal absolute benefit from SGLT2i, although they are less likely to be prescribed these medications (6–8). In line with results from secondary analyses of randomized trials of SGLT2i with consistent demonstration of the efficacy and safety of SGLT2i in older adults with T2D, the results from this real-world evidence study underscore the clinical benefits of empagliflozin in individuals aged ≥65 years with T2D (6,9).
It is noteworthy that even individuals without prevalent CV conditions tended to have lower costs of care overall, even over a relatively short time horizon; it is possible that with longer follow-up times (and higher accrued CV risk) more robust favorable associations may have emerged across both subgroups. It is also possible that due to limitations in the use of claims data, outcomes may have been missed or inaccurately captured. Though real-world evidence is inherently more prone to confounding than data generated through randomized trials, the authors have leveraged both propensity score matching and multiple sensitivity analyses to perform robust analyses of a large cohort.
The average cost of antihyperglycemic medications was higher among individuals initiating empagliflozin, despite the fact that, as the authors note, empagliflozin and most DPP-4i are priced similarly. Prior to propensity score matching, individuals initiating empagliflozin were almost 10 times more likely to be on a baseline GLP-1 RA (12.8% vs. 1.3%), an imbalance contributing significantly to assessment of drug cost in crude analyses. As a class, GLP-1 RA have established clinical benefit, including significant reductions in CV death, nonfatal myocardial infarction, and nonfatal stroke and improved kidney outcomes among people with T2D (10). Though there were nonsignificant differences in baseline GLP-1 RA use after propensity score matching, the cost of diabetes medications remained higher for those treated with SGLT2i over time. It is possible and even probable that individuals initiated on empagliflozin were more likely to be initiated on a GLP-1 RA over the course of follow-up, an imbalance that would not be accounted for by propensity adjustments based on baseline data yet still contributes to the analysis of drugs costs by group over time; if this is the case, then some of the differences in outcomes by SGLT2i status may be attributable not only to SGLT2i use but also to higher frequency of use of GLP-1 RA, which have proven CV efficacy, among the empagliflozin group. Though SGLT2i and GLP-1 RA have been shown to offer clinical benefits of similar magnitude, and with similar outcomes, their mechanisms of efficacy are distinct and likely offer incremental or even additive CV benefit (11). More broadly, this question underscores some of the inherent limitations of nonrandomized evidence generation: some degree of residual confounding remains likely, even after propensity score matching.
Htoo et al. have provided compelling real-world evidence that the use of empagliflozin among people with T2D is associated with both lower health care resource utilization and cost. The use of empagliflozin therefore appears to be associated not only with meaningfully lower CV risk for individuals with T2D but also with reduced strain on health care resources. The observed lower health care resource utilization and cost, particularly among individuals with baseline CV conditions, underscore the potential benefits of empagliflozin in routine clinical practice. Moving forward, research should focus on addressing the limitations of the current study, including longer follow-up periods and consideration of additional factors, e.g., medication drop-in, to further elucidate the impact of empagliflozin on health care outcomes and costs. Additionally, comparative effectiveness observational studies featuring a broader range of medical therapies would provide additional insights into the optimal management of T2D. The work by Htoo et al. therefore represents an important step toward the generation of real-world evidence to help guide clinical decision-making to enhance patient care and optimize resource allocation in the management of T2D.
See accompanying article, p. 1900.
Article Information
Duality of Interest. D.K.M. reports personal fees from Boehringer Ingelheim, Sanofi US, Merck & Co., Merck Sharp & Dohme, Lilly USA, Novo Nordisk, AstraZeneca, Lexicon Pharmaceuticals, Eisai, Pfizer, Metavant Sciences, Applied Therapeutics, Afimmune, Bayer, CSL Behring, and Esperion Therapeutics; research support for clinical trials leadership from Boehringer Ingelheim, Pfizer, AstraZeneca, Novo Nordisk, Esperion Therapeutics, Lilly USA, and CSL Behring; and honoraria for consultancy from Lilly USA, Pfizer, Boehringer Ingelheim, Lexicon Pharmaceuticals, Novo Nordisk, Applied Therapeutics, Altimmune, CSL Behring, Bayer, Intercept, and New Amsterdam. No other potential conflicts of interest relevant to this article were reported.
Handling Editors. The journal editors responsible for overseeing the review of the manuscript were Elizabeth Selvin and Naveed Sattar.