An Early Assessment of the Real-World Treatment Patterns of Type 2 Diabetes: A Comparison to the 2018 ADA/EASD Consensus Report Recommendations

Recent studies of drugs in the glucagon-like peptide-1 receptor agonist (GLP-1 RA) and sodium-glucose co-transporter 2 inhibitor (SGLT-2i) classes provided evidence of glycemic efficacy and safety, along with strong evidence of benefits for cardiovascular outcomes and improved outcomes in patients with heart failure (HF) or renal disease [1, 2]. Key SGLT-2i cardiovascular outcomes trials (CVOTs) showed that treatment with this drug class reduced major adverse cardiovascular events (MACE) by up to 20% and reduced the incidence of hospitalization for HF by up to 39% [3‐5]. Key GLP-1 RA CVOTs showed that treatment with this drug class reduced MACE by up to 26% [6‐9]. Significant benefits of these drug classes on renal outcomes, obesity, and minimizing hypoglycemia were also observed in numerous clinical trials [1].

These studies prompted changes in the type 2 diabetes treatment guidelines from major diabetes associations, such as The American Diabetes Association (ADA) and the European Association for the Study of Diabetes (EASD). Together, these organizations released a consensus report in December 2018, which identified targeted clinical groups of interest where GLP-1 RA and SLGT-2i were the preferred agents for treatment [10, 11]. These clinical groups included patients with high risk or history of atherosclerotic cardiovascular disease (ASCVD), chronic kidney disease (CKD), history of HF, a need to minimize hypoglycemia, and/or a need to minimize weight gain or to promote weight loss. These recommendations were later adopted by the official guidelines.

Evaluating RTUR in the year of publication of the guideline allows an assessment of the cumulative adoption in prior years and provides a baseline for future evaluations of guideline alignment. The guideline is based on accumulating data that may have driven adoption prior to the actual publication. The current data likely reflect primarily this early adoption rather than guideline-driven changes in the pattern of treatment. Knowledge about the degree of guideline alignment can motivate efforts to improve it. Furthermore, evaluating factors associated with treatment choices can help guide interventions to improve guideline alignment. Therefore, we set out to provide a description of the use rates and the factors associated with utilization (or not) of guideline-aligned therapy in real-world data derived from Commercial- and Medicare-supported practices.

Based on these recommendations, we retrospectively investigated the medication use patterns in real-world patients with type 2 diabetes in the US. Data were drawn from a large health insurance claims database, evaluating medication utilization patterns and alignment with these clinical groupings.

This was a retrospective observational study of patients with type 2 diabetes in a large health insurance claims database (IBM^® MarketScan^®, including Commercial and Medicare Supplemental Databases). The objectives of this study were to provide an early post-guideline update of patterns of medication use among patients with type 2 diabetes using the latest healthcare claims data available from 2019 and to examine how different classes of diabetes treatments are used in clinical groups defined in the ADA Pharmacotherapy Prescribing Guidelines [2].

The database provided information on patient enrollment, demographic characteristics, inpatient and outpatient services, and prescription drug use over time. The Commercial cohort contained data from 475,613 patients and the Medicare cohort contained data from 105,128 patients with type 2 diabetes. Inclusion criteria for this study were patients identified as having type 2 diabetes by two separate diagnoses during the study period, were at least 18 years of age in midyear 2019, and filled diabetes prescription claims in 2019. All patients were required to have continuous enrollment for medical and pharmacy coverage during the immediate year before (2018) and the year (2019) of prescribing pattern analysis.

This study utilized the full calendar year 2019 data to evaluate treatment utilization patterns. Prior data from 2007 to 2018 were used to establish disease history and comorbidities of each patient. Laboratory data, such as hemoglobin A1c (HbA1c), estimated glomerular filtration rate (eGFR), and lipid levels, were incorporated for a small proportion of patients based on availability.

All database records used in the study are de-identified and fully compliant with the US Health Insurance Portability and Accountability Act (HIPAA) of 1996. This study was exempted from Institutional Review Board approval as it did not involve collection, use, or transmittal of individually identifiable data. Eli Lilly has the User License with the owner of the data for a fee. This is an enterprise-wide license, which gives Lilly the right to access and analyze data while the license is in effect.

Five clinical groups based on the ADA/EASD consensus statement were identified and defined using the following criteria. Corresponding recommended therapies in the consensus statement are noted in parentheses:

Data from the Commercial and Medicare insurance cohorts were analyzed separately.

The current study applied data from 2019 and attempted an early evaluation of utilization rates of guideline-recommended treatments for diabetes in relation to recommendations in the 2018 ADA/EASD consensus report on management of hyperglycemia and the subsequent 2019 guidelines update. The results showed that a large proportion of patients with type 2 diabetes in both the Commercial and Medicare cohorts had qualifying conditions for the recommendation, and this was more common in the Medicare cohort. Overall, GLP-1 RA and SLGT-2i were underutilized in several clinical groups.

At the end of the first year following the Management of Hyperglycemia in Type 2 Diabetes Consensus Report (2018), RTURs were < 40% in each clinical group. The Medicare cohort had a measurably lower RTUR in each clinical group (~ 20%). This difference between cohorts could be due to demographic, clinical, economic, or other factors that are associated with Medicare coverage. For example, as evident in the higher CCI, patients in the Medicare cohort had more chronic comorbidities and therefore were likely require more medications, with the attendant medical and financial burdens. Features of Medicare coverage, such as formulary listings and copays for branded drugs like these new medication classes, may have also contributed.

Recent studies examined the clinical characteristics and prescribing preferences for SGLT-2is and GLP-1 RAs in the years prior to the release of the 2018 ADA/EASD consensus statement. Dave et al. [13] concluded that over 5 years between 2013 and 2018 for SGLT-2i, shifts in preference for an SGLT-2i with proven CV benefit did occur following changes in drug labels and guidelines, while for GLP-1 RAs, shifts in preference for GLP-1 RAs with proven CV benefit following changes in drug labels and guidelines were not evident, and other factors, such as price or ease of administration, may have been the cause. Ganz et al. [14] reported that despite recent increases, use of GLP-1 RA and SGLT-2i among patients with type 2 diabetes and CVD in the Optum claims database remained similar and low across 2014 to 2018.

Surprisingly, SGLT-2i utilization rate in the HF patient group was below the overall cohort average in 2019. There were several CV outcomes studies demonstrating clear benefits in HF for a period of time [3‐5]; however, definitive trials with HF improvement as the primary endpoint have only recently been published [15, 16]. This may suggest delayed utilization pending availability of fully supportive data. Other considerations may include comorbidities, polypharmacy, or renal dysfunction complicating medication decision making. The current available data provide stronger support for preferential utilization of SGLT-2i in HF, as reflected in the guideline. Our observed RTUR and differences in utilization between Medicare and Commercial cohorts represent an opportunity for meaningful improvement in care.

In patients with a history of severe hypoglycemia (prior hospitalization or emergency room visit for hypoglycemia), no preferential utilization of GLP-1 RA and SLGT-2i was evident in the Medicare cohort; however, there was a small preference for these drug classes in patients with a history of severe hypoglycemia in the Commercial cohort, and the utilization of SU only was comparable to that without severe hypoglycemic event history in both cohorts. The utilization rate of SU was also higher in the Medicare cohort than in the Commercial cohort, where both increased age and the declining renal status of this cohort point to greater risk of SU-related hypoglycemia events [17], magnifying the risk associated with prior hypoglycemic events. In patients with a compelling need to minimize hypoglycemia, with or without ASCVD, CKD or HF, secretagogues (e.g., SUs) are the least preferred therapy [2]. Our data highlight the persisting use of SU among patients with a history of severe hypoglycemia as an opportunity for meaningful improvement in care.

Interestingly, the obesity group in both cohorts achieved relatively favorable alignments with the guidelines. This may have been influenced by patients’ desire for weight management and the physicians’ perception of obesity as a risk modifier, which together might increase the likelihood of selecting an SGLT2i or GLP-1RA. The multivariate regression analysis also showed that only the obesity group achieved a higher guideline alignment; all other groups (ASCVD, HF, CKD, and hypo) lagged behind the reference group. The regression further revealed that many individual patient factors, including demographics, past and current type 2 diabetes treatment regimens, and comorbidities contribute to the RTUR in the real world.

There are some limitations to our study. Although studies demonstrating CV and renal benefits had been published in years prior, the recommendation that suggested preferential utilization of SGLT2i and GLP-1RA among specified clinical groups was first published in December 2018. The current evaluation was based on the first 12 months of prescription claims data following the release of the recommendation. Longer term evaluation is needed. The guidelines apply to patients who are not successfully managed by metformin and diet. However, actual and target HbA1c data were not available to determine whether first-line metformin and lifestyle therapy were sufficient, potentially contributing to an incorrect perception of treatment inertia. Although this study attempted to identify patients with specific risks using all available information and well-established algorithms in the database, the clinical groups may fail to accurately include all eligible patients. For example, patients with obesity or mild CKD may not have been diagnosed. It is also possible that the current prescriber may not be aware of relevant clinical history (e.g., prior hypo). Some covariates, such as race/ethnicity and socioeconomic status of the patients, were not available. This study was based on large samples of a relatively well-insured population (Commercial insurance and Medicare with supplemental insurance). Thus, patients may have fewer access problems. The results may not be extended to other special populations, such as the uninsured or those on Medicaid.

In conclusion, prevalence of guideline-aligned treatment use in 2019 was low, particularly since many patients fit into multiple risk groups with established treatment benefits. This research in the real world reveals that SGLT-2is and GLP-1RAs are underutilized and are only partially aligned with current treatment recommendations. Lower RTUR and larger alignment gaps were seen in the Medicare population. Utilization of recommended therapies are lagging in some clinical and demographic groups, and they represent potential opportunities to make meaningful improvement in the care for patients with type 2 diabetes. Continued monitoring and evaluations of real-world use data will help inform and improve implementation of guideline-based treatment in actual prescribing practice and perhaps identify where further consensus is needed to enhance effectiveness in diabetes care.