National Trends in Use of Sodium-Glucose Cotransporter-2 Inhibitors and Glucagon-like Peptide-1 Receptor Agonists by Cardiologists and Other Specialties, 2015 to 2020.

Background Sodium-glucose cotransporter-2 inhibitors (SGLT2is) and glucagon-like peptide-1 receptor agonists (GLP-1RAs) mitigate cardiovascular risk in individuals with type 2 diabetes, but most eligible patients do not receive them. We characterized temporal trends in SGLT2i and GLP-1RA use by cardiologists compared with other groups of clinicians. Methods and Results We conducted a descriptive analysis of serial, cross-sectional data derived from IQVIA's National Prescription Audit, a comprehensive audit capturing ≈90% of US retail prescription dispensing and projected to population-level data, to estimate monthly SGLT2is and GLP-1RAs dispensed from January 2015 to December 2020, stratified by prescriber specialty and molecule. We also used the American Medical Association's Physician Masterfile to calculate average annual SGLT2is and GLP-1RAs dispensed per physician. Between January 2015 and December 2020, a total of 63.2 million SGLT2i and 63.4 million GLP-1RA prescriptions were dispensed in the United States. Monthly prescriptions from cardiologists increased 12-fold for SGLT2is (from 2228 to 25 815) and 4-fold for GLP-1RAs (from 1927 to 6981). Nonetheless, cardiologists represented only 1.5% of SGLT2i prescriptions and 0.4% of GLP-1RA prescriptions in 2020, while total use was predominated by primary care physicians/internists (57% of 2020 SGLT2is and 52% of GLP-1RAs). Endocrinologists led in terms of prescriptions dispensed per physician in 2020 (272 SGLT2is and 405 GLP-1RAs). Cardiologists, but not noncardiologists, increasingly used SGLT2is over GLP-1RAs, with accelerated uptake of empagliflozin and dapagliflozin coinciding with their landmark cardiovascular outcomes trials and subsequent US Food and Drug Administration label expansions. Conclusions While use of SGLT2is and GLP-1RAs by cardiologists in the United States increased substantially over a 6-year period, cardiologists still account for a very small proportion of all use, contributing to marked undertreatment of individuals with type 2 diabetes at high cardiovascular risk.

US Food and Drug Administration

glucagon‐like peptide‐1 receptor agonist

National Prescription Audit

Peptide Innovation for Early Diabetes Treatment

sodium‐glucose cotransporter‐2 inhibitor

type 2 diabetes

Clinical Perspective

What Is New?

The role cardiologists have played relative to other clinicians in disseminating cardioprotective sodium‐glucose cotransporter‐2 inhibitor (SGLT2i) and glucagon‐like peptide‐1 receptor agonist (GLP‐1RA) medications has not been well characterized.

This analysis of a near–census‐level audit of the US retail prescriptions shows that since 2015, cardiologists have increased use of SGLT2is and GLP‐1RAs 12‐fold and 4‐fold, respectively.

Nonetheless, cardiologists accounted for <2% of all SGLT2i and GLP‐1RA use in 2020.

What Are the Clinical Implications?

The marginal role of cardiologists in deploying cardioprotective SGLT2is/GLP‐1RAs may be contributing to the undertreatment of patients with diabetes and cardiovascular disease.

Deploying SGLT2i and GLP‐1RA medications to these patients will require a focused and coordinated effort across all the medical specialties that treat patients with diabetes, with an enhanced role for the cardiologist.

Further research is needed to understand the persistent barriers among clinicians to increased adoption of these evidence‐based therapies, particularly how these barriers might differ by specialty (ie, cardiology).

Cardiovascular disease (CVD) is the leading cause of death and disability among individuals with type 2 diabetes (T2D). , Of the 34 million Americans with T2D, over one‐third also have CVD, and half will eventually die of heart disease or stroke. Patients with T2D and CVD face an alarming 4‐fold risk of heart disease or stroke mortality, yet only 21% of individuals with diabetes achieve basic risk factor goals for blood pressure, blood glucose, and low‐density lipoprotein. , As the prevalence of diabetes continues to surge globally, therapies that effectively mitigate the cardiovascular risks associated with T2D have the potential to transform the landscape of diabetes care and improve health outcomes for millions of patients.

Two such treatments, sodium‐glucose cotransporter‐2 inhibitors (SGLT2is) and glucagon‐like peptide‐1 receptor agonists (GLP‐1RAs), are already broadly available in the United States. Over a dozen cardiovascular outcomes trials have shown that these antihyperglycemic drugs are also robust cardioprotective agents ; meta‐analyses indicate that among patients with T2D, SGLT2is reduce cardiovascular mortality or hospitalization for heart failure by ≈23%, while GLP‐1RAs reduce a composite of cardiovascular death, stroke, or myocardial infarction by ≈12%. Additional data point toward a role for SGLT2is and GLP‐1RAs in slowing progression of diabetic kidney disease, while multiple large trials now indicate that SGLT2is are effective treatments to manage heart failure and chronic kidney disease, regardless of whether patients have diabetes. , , , Thus, SGLT2i and GLP‐1RA drugs have been heralded as paradigm‐shifting cardiometabolic therapies to curtail the chronic cardiovascular and renal consequences of T2D. Reflecting this evidence, since 2016 the US Food and Drug Administration (FDA) has approved label expansion of many SGLT2is and GLP‐1RAs to incorporate cardiovascular risk reduction indications, and recent clinical guidelines , , , now strongly recommend the use of an SGLT2i or GLP‐1RA in patients with T2D who have established CVD or additional cardiovascular risk factors.

Despite the promise of SGLT2is and GLP‐1RAs, most individuals who are eligible for these treatments do not receive them. Estimates from multiple cross‐sectional analyses of large T2D cohorts suggest that only 2% to 10% and 3% to 8% of patients with T2D and CVD are receiving an SGLT2i or GLP‐1RA, respectively. , , , , , Meanwhile, over a third of patients with T2D have concomitant CVD, , , making them strong candidates for an SGLT2i or GLP‐1RA. Because cardiologists outnumber most other specialists that regularly manage the microvascular and macrovascular consequences of T2D and have more frequent encounters than other specialists with high‐risk patients with T2D with comorbid CVD, they may be well positioned to deploy these treatments. However, limited available data from individual academic medical centers and 1 pharmacy benefits manager before 2019 suggest that cardiologists may be playing a limited role in prescribing these drugs. , , , Many landmark cardiovascular outcomes trials occurred after this period, and prominent clinical guidelines and awareness campaigns have since specifically called upon cardiologists to take a leading role in deploying these drugs. , , , Therefore, we used a large, nationally representative data set to characterize prescribing patterns of cardiologists compared with noncardiologists, as well as to understand how these patterns have evolved in response to key events from 2015 to 2020.

Methods

The data that support the findings of this study are available from the corresponding author upon reasonable request. This study was exempt from the Johns Hopkins University School of Medicine institutional review board, as it did not constitute human subjects research.

Data Sources

We analyzed IQVIA’s National Prescription Audit (NPA) from January 2015 to December 2020. The NPA represents a gold‐standard source of information about retail pharmacy transactions and provides dispensing estimates based on a sample of ≈90% of all retail prescriptions dispensed in the United States, including 90% of retail pharmacies, 80% of nursing home pharmacies, and 70% of mail‐service pharmacies. , This sample is then weighted, using projection factors such as distance between sample and nonsample stores, to estimate 100% of retail transactions in the United States. IQVIA links the NPA to the American Medical Association’s Physician Masterfile and other professional organization records to confirm the primary specialty of prescribers. We also combined the NPA‐generated projections of total dispensed prescription volume with aggregate physician census data for each specialty from the American Medical Association’s Physician Masterfile and the American Osteopathic Association’s Osteopathic Medical Profession Report to estimate the average number of dispensed prescriptions per clinician in 2020 within each specialty.

For our analysis, we categorized physician assistants and nurse practitioners as advanced practice providers, while grouping a cluster of general practitioners as primary care physicians (PCP)/internists. This latter category includes family practice, general practice, general preventive medicine, geriatrics, internal medicine, internal medicine/pediatrics, osteopathic medicine, and pediatrics, and is based on prior publications that reported specialty‐specific data from the NPA.

From the NPA, we extracted monthly dispensed prescriptions, prescriber specialty, and brand names of the drugs used among all patients, regardless of their diabetes status. All SGLT2is and GLP‐1RAs were included, except for Saxenda, a liraglutide dose marketed for weight loss and not indicated for T2D or cardiovascular risk reduction. We examined total, new, and refill prescriptions in our initial analyses; these analyses each yielded substantively similar findings, and only total prescriptions are reported herein. Thus, use of the term prescriptions throughout the article refers to total dispensed prescriptions.

Statistical Analysis

We used descriptive statistics to perform our analyses, with main outcomes including monthly/yearly prescriptions for each drug, monthly/yearly change in prescription volumes, monthly shifts in market share of each medication within its drug class, and annual prescriptions dispensed per physician within each specialty. We plotted these outcomes over time to assess for notable trends and inflection points, then used the Wald test for structural breaks to determine whether key clinical trial e‐publications and FDA drug label expansions coincided with statistically significant inflections in our data. We analyzed the data using Stata 15 (StataCorp, College Station, TX) and Microsoft Excel (Microsoft, Redmond, Washington).

Results

Growth in Prescribing of SGLT2is and GLP‐1RAs

From January 2015 to December 2020, ≈63.2 million SGLT2i and 63.4 million GLP‐1RA medications were dispensed in the United States (Table and Table S1). Annual prescriptions doubled for SGLT2is and tripled for GLP‐1RAs from 2015 to 2020. The average annual growth rate during this 6‐year period was 15.6% for SGLT2is and 25.2% for GLP‐1RAs.

Table 1

SGLT2i and GLP‐1RA Prescriptions Dispensed and Year‐Over‐Year Growth by Specialty

SGLT2i	2015	2016	2017	2018	2019	2020	Total from 2015–2020
Cardiology	35 136	48 759 (+39%)	61 883 (+27%)	87 883 (+42%)	135 476 (+54%)	221 948 (+64%)	591 085 (+532%)
Nephrology	14 616	21 105 (+44%)	24 597 (+17%)	28 350 (+15%)	40 031 (+41%)	69 507 (+74%)	198 206 (+376%)
Endocrinology	1 332 436	1 623 451 (+22%)	1 656 326 (+2%)	1 786 542 (+8%)	1 971 564 (+10%)	2 220 876 (+13%)	10 591 195 (+67%)
PCP/Internist	4 505 408	5 666 536 (+26%)	5 835 253 (+3%)	6 190 060 (+6%)	6 954 341 (+12%)	8 281 097 (+19%)	37 432 695 (+84%)
APP	1 198 857	1 685 227 (+41%)	1 906 972 (+13%)	2 218 806 (+16%)	2 723 749 (+23%)	3 452 394 (+27%)	13 186 005 (+188%)
Other	147 671	181 212 (+23%)	171 557 (−5%)	186 469 (+9%)	210 610 (+13%)	310 215 (+47%)	1 207 734 (+110%)
Total SGLT2i	7 234 124	9 226 290 (+28%)	9 656 588 (+5%)	10 498 110 (+9%)	12 035 771 (+15%)	14 556 037 (+21%)	63 206 920 (+101%)

Values for each year are absolute number of prescriptions dispensed that year, with percentages in parentheses indicating change in prescription volumes from the prior year. The rightmost column displays the total number of prescriptions dispensed over the study period and the total percentage increase from 2015 compared with 2020. APP indicates advanced practice providers; GLP‐1RA, glucagon‐like peptide‐1 receptor agonist; PCP, primary care physician; and SGLT2i, Sodium‐glucose cotransporter‐2 inhibitor.

33 671

(+25%)

34 867

(+4%)

40 571

52 229

(+29%)

70 055

(+34%)

258 290

(+160%)

24 378

27 699

(+14%)

31 942

40 010

(+25%)

51 719

(+29%)

195 583

(+161%)

1 735 004

(+30%)

1 895 542

2 276 274

(+20%)

2 793 210

3 314 692

13 350 745

(+148%)

4 034 407

(+25%)

4 495 686

(+11%)

5 649 762

7 215 560

9 004 492

(+25%)

33 626 778

(+179%)

1 395 344

1 736 788

(+24%)

2 411 028

3 421 990

4 602 282

(+34%)

14 514 321

(+386%)

165 942

185 324

239 404

(+29%)

303 255

441 287

(+46%)

1 466 175

(+237%)

7 388 746

(+30%)

8 375 906

10 648 981

13 826 254

(+30%)

17 484 527

63 411 892

(+207%)

Among cardiologists, monthly dispensed prescriptions increased ≈12‐fold for SGLT2is (2228 in January 2015 to 25 815 in December 2020) and 4‐fold for GLP‐1RAs (from 1927 to 6981) (Figure 1). Compared with the annual trend among all prescribers, cardiologists outpaced the overall increase for SGLT2is (532% increase among cardiologists versus 101% overall) but slightly lagged the overall increase for GLP‐1RAs (160% versus 207%) (Table). This observed increase in prescription volumes was largely not explained by an increase in the number of cardiologists in the United States, which grew 15% from 2015 to 2020. While cardiologists increased prescribing for these medications, they represented a marginal fraction of national monthly totals (SGLT2is: 0.5% in January 2015 versus 1.9% in December 2020; GLP‐1RAs: 0.5% versus 0.4%) (Figure 1).

Figure 1

Cardiologist contribution to use of SGLT2is and GLP‐1RAs.

SGLT2i (A) and GLP‐1RA (B) prescriptions by cardiologists. The primary axis in orange shows the percentage of overall monthly dispensed prescriptions for each drug class that are attributed to cardiologists; the secondary axis in blue shows the absolute number of dispensed prescriptions by cardiologists each month. While cardiologists have substantially increased absolute numbers of prescriptions, they continue to account for a very small proportion of overall use. GLP‐1RA indicates glucagon‐like peptide‐1 receptor agonists; and SGLT2i, sodium‐glucose cotransporter‐2 inhibitor. Source: IQVIA National Prescription Audit, 2015–2020.

Relative Prescription Volumes Across Specialties

Most prescriptions for both SGLT2is and GLP‐1RAs were provided by PCPs/internists (Figure 2). In 2020, this group accounted for 57% and 52% of national SGLT2i and GLP‐1RA prescriptions, respectively, followed by advanced practice providers (24% and 26%), endocrinologists (15% and 19%), cardiologists (1.5% and 0.4%), and nephrologists (0.5% and 0.3%). Using a different metric, average number of dispensed prescriptions per physician in 2020 within each specialty, endocrinologists prescribed more than other clinicians (272 SGLT2is and 405 GLP‐1RAs), with cardiologists (6 SGLT2is and 2 GLP‐1RAs) again lower (Figure 3).

Figure 2

SGLT2i and GLP‐1RA use across clinician specialties.

Monthly SGLT2i (A) and GLP‐1RA (B) dispensed prescriptions by specialty of the prescriber. PCP/internist prescribers accounted for the majority of dispensed prescriptions. Arrows delineate a subset of relevant clinical trial publications and key regulatory actions. APP indicates advanced practice provider; CREDENCE, Canagliflozin and Renal Events in Diabetes with Established Nephropathy Clinical Evaluation; CV, cardiovascular; DAPA‐HF, Dapagliflozin and Prevention of Adverse Outcomes in Heart Failure; DECLARE‐TIMI, Multicenter Trial to Evaluate the Effect of Dapagliflozin on the Incidence of Cardiovascular Events; EMPA‐REG OUTCOME, BI 10773 (Empagliflozin) Cardiovascular Outcome Event Trial in Type 2 Diabetes Mellitus Patients; FDA, US Food and Drug Administration; GLP‐1RA, glucagon‐like peptide‐1 receptor agonists; HF, heart failure; LEADER, Liraglutide Effect and Action in Diabetes: Evaluation of Cardiovascular Outcome Results; PCP/Int Med, primary care physicians and internists; SGLT2i, sodium‐glucose cotransporter‐2 inhibitor; SQ, subcutaneous; and SUSTAIN‐6, Trial to Evaluate Cardiovascular and Other Long‐Term Outcomes With Semaglutide in Subjects With Type 2 Diabetes 6. Source: IQVIA National Prescription Audit, 2015–2020.

Figure 3

Prescriptions per physician in 2020.

Estimated average SGLT2i and GLP1‐RA prescriptions dispensed per physician by specialty from January to December 2020. Endocrinologists outpace all other physician specialties in per‐physician use. GLP‐1RA indicates glucagon‐like peptide‐1 receptor agonists; PCP/Int Med primary care physicians and internists; and SGLT2i, sodium‐glucose cotransporter‐2 inhibitor. Source: IQVIA National Prescription Audit, 2015–2020.

Drug Choice Among Cardiologists versus Noncardiology Clinicians

While both cardiologists and noncardiology clinicians used SGLT2is and GLP‐1RAs in approximately equal ratios in 2015, cardiologists increasingly used SGLT2is over the course of the study (Figure 4A). The ratio of SGLT2i prescriptions to GLP‐1RA prescriptions among cardiologists was 1.2:1 in January 2015. By December 2020, this ratio had shifted to 3.7:1. In contrast, noncardiologists trended slightly in the opposite direction (Figure 4B), starting the study at 1.1:1 in January 2015 and ending at 0.8:1 in December 2020.

Figure 4

Drug choice among cardiologists and noncardiologists.

Proportion of monthly SGLT2i/GLP‐1RA dispensed prescriptions from cardiologists (A) or noncardiology clinicians (B) attributed to each molecule in the 2 drug classes. Within each graph, SGLT2i drugs are shaded in blue hues, while GLP‐1RA drugs are shaded in brown hues. The gray series (other) corresponds to ertugliflozin and albiglutide, which each accounted for <1% of the total prescriptions. Arrows delineate a subset of relevant clinical trial publications and key regulatory actions. CV indicates cardiovascular; DAPA‐HF, Dapagliflozin and Prevention of Adverse Outcomes in Heart Failure; EMPA‐REG OUTCOME, BI 10773 (Empagliflozin) Cardiovascular Outcome Event Trial in Type 2 Diabetes Mellitus Patients; EMPEROR‐Reduced, Empagliflozin Outcome Trial in Patients With Chronic Heart Failure With Reduced Ejection Fraction; FDA, US Food and Drug Administration; GLP‐1RA, glucagon‐like peptide‐1 receptor agonists; HF, heart failure; LEADER, Liraglutide Effect and Action in Diabetes: Evaluation of Cardiovascular Outcome Results; PCP/Int Med, primary care physicians and internists; SGLT2i, sodium‐glucose cotransporter‐2 inhibitor; and SQ, subcutaneous. Source: IQVIA National Prescription Audit, 2015–2020.

SGLT2i Trends

The predominance of SGLT2is over GLP‐1RAs among cardiologists began in late 2015 and early 2016, mainly consisting of accelerated use of the SGLT2i empagliflozin (Figure 4A). Empagliflozin eventually became the most widely used SGLT2i among cardiologists and noncardiologists, accounting for 67% and 58%, respectively, of SGLT2is dispensed in 2020, although its growth was more rapid and pronounced among cardiologists (Figure 4A versus Figure 4B). The timing of empagliflozin’s early rise coincided with publication of the results from the EMPA‐REG OUTCOME (BI 10773 [Empagliflozin] Cardiovascular Outcome Event Trial in Type 2 Diabetes Mellitus Patients) trial and the FDA expansion of empagliflozin’s drug label to reflect a cardiovascular risk reduction indication in T2D (Figure 4A).

The average monthly change in empagliflozin’s share of SGLT2i prescriptions among cardiologists was 1.6 times higher in the 6 months following publication of its trial than the 6 months prior, and 3.8 times higher in the 6 months following FDA indication expansion than the 6 months prior. The Wald test for structural breaks, which can be used to detect abrupt changes in time‐series data, showed a statistically significant structural break in December 2016 (P<0.001), when the FDA expanded empagliflozin’s indications but not in September 2015 (P=0.16), when EMPA‐REG OUTCOME results were published online. While noncardiology clinicians also increased their empagliflozin use following this FDA activity, the shift was less pronounced (Figure 4B and Figure 5B; P=0.41 in September 2015 after EMPA‐REG OUTCOME and P=0.043 in December 2016 after the FDA activity).

Figure 5

Share of SGLT2i prescriptions comprised by dapagliflozin and empagliflozin.

Share of monthly SGLT2i dispensing attained by dapagliflozin (A) and empagliflozin (B) among cardiologists compared with all other clinicians. Arrows delineate a selection of landmark cardiovascular outcomes trial publications and key FDA indication expansions which may have influenced prescription choice among clinicians. CV indicates cardiovascular; DAPA‐HF, Dapagliflozin and Prevention of Adverse Outcomes in Heart Failure; EMPA‐REG OUTCOME, BI 10773 (Empagliflozin) Cardiovascular Outcome Event Trial in Type 2 Diabetes Mellitus Patients; EMPEROR‐Reduced, Empagliflozin Outcome Trial in Patients With Chronic Heart Failure With Reduced Ejection Fraction; FDA, US Food and Drug Administration; GLP‐1RA, glucagon‐like peptide‐1 receptor agonists; HF, heart failure; and SGLT2i, sodium‐glucose cotransporter‐2 inhibitor. Source: IQVIA National Prescription Audit, 2015–2020.

During the study period, there was a declining trend in the relative use of canagliflozin, from a maximum of 65% of all SGLT2i prescriptions in the first month of the study to a nadir of 3.0% in the most recent month (Figure 4). Our analysis did not detect any structural breaks in this trend coinciding with FDA communications regarding potential amputation risks with this medication (Figure S1). , ,

From September 2019 to December 2020, dapagliflozin’s share of SGLT2i prescriptions among cardiologists increased 23%, following a decrease since 2015 (Figure 5A), while empagliflozin’s share declined 18%, following an increase since 2015 (Figure 5B). This shift from empagliflozin to dapagliflozin among cardiologists was not observed among noncardiologists, for whom empagliflozin’s share of SGLT2i prescriptions continued to increase and dapagliflozin’s share remained flat. This swing in prescribing trends among cardiologists coincided with the publication of the DAPA‐HF (Dapagliflozin and Prevention of Adverse Outcomes in Heart Failure) trial in September 2019 (P<0.001), followed in May 2020 (P<0.001) by FDA drug label expansion for the treatment of heart failure in adults with and without T2D.

GLP‐1RA Trends

Among the GLP‐1RA medications, we did not observe similar shifts in prescription choice following publication of trial results, including LEADER (Liraglutide Effect and Action in Diabetes: Evaluation of Cardiovascular Outcome Results; liraglutide), SUSTAIN‐6 (Trial to Evaluate Cardiovascular and Other Long‐Term Outcomes With Semaglutide in Subjects With Type 2 Diabetes; subcutaneous semaglutide), PIONEER‐6 (Peptide Innovation for Early Diabetes Treatment; oral semaglutide), and REWIND (Researching Cardiovascular Events With a Weekly Incretin in Diabetes; dulaglutide), or their associated FDA drug label changes (Figure 4). We did observe a marked decline in the share of daily injection GLP‐1RAs (76% in January 2015 to 19% in December 2020), along with a reciprocal increase toward weekly injection GLP‐1RAs (Figure 6). Among both cardiologists and noncardiologists, the most dispensed GLP‐1RAs in 2020 were dulaglutide (39% of cardiology GLP‐1RAs and 45% of noncardiology GLP‐1RAs) followed by semaglutide (37% of cardiology GLP‐1RAs and 28% of noncardiology GLP‐1RAs), both of which are weekly injections. We did not detect any pronounced change in trends for the prescribing of semaglutide following approval of its oral formulation in September 2019 as the only GLP‐1RA available as an oral medication (Figure 4). However, the growth in oral semaglutide prescriptions has consistently exceeded the growth of the GLP‐1RA category since entrance into the market (Figure S2).

Figure 6

Prescriptions of daily vs weekly injection GLP‐1RAs.

Since January 2015, the absolute number of dispensed prescriptions for daily injection GLP‐1RAs has stagnated, while weekly injection GLP‐1RA prescriptions have increased substantially, accounting for the majority of GLP‐1RA prescriptions in 2020 (A). This shift from daily to weekly GLP‐1RA injections is also reflected in each category’s relative proportion of GLP‐1RA prescriptions (B). Daily injection GLP‐1RAs are lixisenatide, liraglutide, and short‐acting exenatide (Byetta). Weekly injections are albiglutide, dulaglutide, subcutaneous semaglutide, and long‐acting exenatides (Bydureon, Bydurean BCise, and Bydureon Pen). CV indicates cardiovascular; FDA, US Food and Drug Administration; GLP‐1RA, glucagon‐like peptide‐1 receptor agonists; LEADER, Liraglutide Effect and Action in Diabetes: Evaluation of Cardiovascular Outcome Results; REWIND, Researching Cardiovascular Events With a Weekly Incretin in Diabetes; SGLT2i, sodium‐glucose cotransporter‐2 inhibitor; SQ, subcutaneous; and SUSTAIN‐6, Trial to Evaluate Cardiovascular and Other Long‐Term Outcomes With Semaglutide in Subjects With Type 2 Diabetes 6. Source: IQVIA National Prescription Audit, 2015–2020.

Discussion

Despite the cardiovascular benefits of SGLT2is and GLP‐1RAs, these therapies have not reached the majority of patients for whom they are guideline recommended. , , , , , , We analyzed a comprehensive audit of the US retail prescription drug market to examine use of SGLT2is and GLP‐1RAs across medical specialties between 2015 and 2020. Total use of both drug classes increased substantially during this 6‐year period, including among cardiologists, whose monthly prescribing of SGLT2is increased 12‐fold and prescribing of GLP‐1RAs increased 4‐fold. Over the 6‐year study period, the relative ratio of SGLT2i to GLP‐1RA use increased for cardiologists but not noncardiologists, with accelerated uptake of empagliflozin and dapagliflozin coinciding with their landmark cardiovascular outcomes trials and resulting FDA label expansions. However, cardiologists accounted for <2% of prescriptions for these drugs in 2020. Rather, PCPs/internists accounted for the majority of prescriptions in that year (57% of SGLT2is and 52% of GLP‐1RAs), while endocrinologists far outpaced other specialties in prescriptions per physician. In light of prior estimates showing that 40% to 70% of patients with T2D and CVD see a cardiologist each year, , that an individual with T2D and CVD is equally as likely to encounter a PCP and 4 times less likely to encounter an endocrinologist as they are a cardiologist in a given year, , and that one‐third of the patients that cardiologists manage have diabetes, our results demonstrating a limited role of cardiologists in disseminating SGLT2i and GLP‐1RA therapies present an opportunity to optimize access to these cardiometabolic therapies.

Comparison to Prior Studies

The results of our analysis build upon prior investigations of cardiologist prescribing of SGLT2is/GLP‐1RAs. Studies within the Partners HealthCare system demonstrated that cardiologists accounted for ≈5% of total prescribing of these medications in that academic medical institution, with endocrinologists accounting for a plurality of prescriptions. , A study within the University of Mississippi Medical Center as well as another study analyzing data from a pharmacy benefits manager also found low relative use among cardiologists, with a majority of prescriptions originating from PCPs/internists. Our study also suggests that PCPs/internists account for most prescriptions, with the additional insight that endocrinologists lead prescribing on a per‐physician basis, but account for fewer total prescriptions than PCPs/internists because of their smaller total population.

Because prior studies were restricted to prescriptions at individual academic medical centers or a single pharmacy benefits manager, their generalizability may be limited because of institutional formularies, institution‐specific cultures and physician distribution across specialties, narrow patient populations, lack of all payer types, and small samples. By examining near–census‐level data of total, all‐payer dispensing of prescriptions each month, our analysis provides a unique opportunity to analyze highly generalizable national patterns and to compare the evolving drug choices of cardiologists to those of noncardiologists in the context of recent clinical trials and FDA announcements.

Possible Influence by Key Events

While both cardiologists and noncardiologists started the study period with approximately equal use of SGLT2is versus GLP‐1RAs, there were prominent increases in SGLT2i use relative to GLP‐1RA use among cardiologists. This shift was driven primarily by an accelerated uptake of empagliflozin and may have been influenced by publication in 2015 of the EMPA‐REG OUTCOME trial, the first cardiovascular outcomes trial to demonstrate cardiovascular benefit of an SGLT2i, and the FDA’s addition of an indication for cardiovascular risk reduction in T2D in 2016. While later trials (CANVAS [Canagliflozin Cardiovascular Assessment Study], DECLARE‐TIMI [Multicenter Trial to Evaluate the Effect of Dapagliflozin on the Incidence of Cardiovascular Events], etc) demonstrated cardioprotective benefits for other SGLT2is, we did not observe any inflections in our data corresponding to those trials.

In September 2019, the DAPA‐HF trial demonstrated for the first time that an SGLT2i, dapagliflozin, could be effective not just in the primary prevention of heart failure in T2D, but also in secondary prevention as a treatment for heart failure with reduced ejection fraction, regardless of a patient’s diabetes status. This marked expansion in the population of patients who may benefit from an SGLT2i is reflected in our time‐series data after publication of this trial, with use of dapagliflozin accelerating and reducing empagliflozin’s market share among cardiologists. Although the EMPEROR‐Reduced (Empagliflozin Outcome Trial in Patients With Chronic Heart Failure With Reduced Ejection Fraction) trial later showed a similar heart failure result for empagliflozin, the impact of this is not yet reflected in our data. As with the prior shift toward empagliflozin, this prominent shift toward dapagliflozin was observed among cardiologists, but not among noncardiology prescribers.

Insights Into Barriers to Uptake

All major American and European clinical guidelines for cardiovascular management in diabetes now strongly recommend an SGLT2i or GLP‐1RA for patients with T2D who have a history of or high risk for CVD. , , However, only a fraction of these patients receive these medications. , , , Such shortcomings are the result of many potential barriers, including traditional boundaries in diabetes care between medical specialties, cost, clinical inertia, lack of clinician familiarity with the evidence base and potential adverse effects, and, in the case of GLP‐1RAs, concerns and preferences on the part of patients to avoid injected therapies. , , While our study cannot directly evaluate the relative importance of each factor, it does provide some important insight.

The growth in dapagliflozin use among cardiologists coinciding with the DAPA‐HF trial, the first cardiovascular outcomes trial to demonstrate the effectiveness of an SGLT2i outside of T2D, suggests that one possible factor holding back cardiologist uptake of these drugs could be the perception that diabetes care is beyond their specialty’s purview. Second, the lagging use of GLP‐1RAs relative to SGLT2is among cardiologists, despite the opposite trend among noncardiologists, may be consistent with the hypotheses that a preference to avoid needles or concerns regarding the potential gastrointestinal side effects of GLP‐1RAs are hindering use by cardiologists. Future initiatives to foster uptake among cardiologists should specifically consider these 2 potential barriers. These efforts to increase use by cardiologists might include conducting clinical trials to assess the effectiveness of additional SGLT2is/GLP‐1RAs for a growing number of applications outside the context of T2D, investigating the cardiovascular benefits of oral GLP‐1RA formulations, and avoiding reference to SGLT2is and GLP‐1RAs solely as antihyperglycemic or “diabetes” drugs, but rather promoting cross‐disciplinary, coordinated models of care that extend across multiple medical specialties.

Study Limitations

While the NPA provides robust, all‐payer national dispensing estimates from US outpatient retail pharmacies, it excludes nonretail channels such as hospital pharmacies or dispensing physicians. This registry captures data on drug dispensing by pharmacies, rather than clinician prescribing, and the 2 metrics may be slightly discordant. , Additionally, the NPA does not include any demographic or diagnostic information about individuals, nor does it include CIs for its projections, although the audit provides a near census of the US retail market. Furthermore, because the NPA categorizes both newly initiated therapy as well as a newly written prescription to continue existing therapy after a patient is completely out of refills as new prescriptions, it was not possible to fully limit our analysis to only those prescriptions that are confirmed new initiations. In addition, the COVID‐19 pandemic overlapped with the final 10 months of our study period and may have had additional impacts on prescribing that are not reflected in our analysis (reduced doctor or pharmacy visits, shifts to mail‐order dispensing, etc). Finally, although our analyses suggest inflection points in the use of certain medications that coincided with important clinical trials or regulatory actions, our study was not designed to determine whether a particular event was directly responsible for these inflections. Other simultaneous events, such as increased physician marketing, knowledge dissemination at conferences, changes in preferred drug formularies, or lightening of prior authorization requirements, may influence prescribing patterns.

Conclusions

Our study demonstrates that while prescriptions from cardiologists for SGLT2i and GLP‐1RA medications have increased many‐fold since 2015, they continue to account for a very small share of the use of these therapies. Most prescriptions were provided by PCPs/internists, while endocrinologists accounted for the most prescriptions per provider. Favorable results from landmark SGLT2i cardiovascular outcomes trials and the subsequent FDA drug indication expansions coincided with evolving prescribing preferences, especially among cardiologists. Reaching the millions of individuals whose quality and duration of life could be significantly improved with these treatments will require a concerted and coordinated effort across a diverse range of clinicians as well as a reimagining of the traditional boundaries between medical specialties.

Sources of Funding

None.

Disclosures

Dr Alexander is past chair and a current member of the FDA Peripheral and Central Nervous System Advisory Committee; has served as a paid advisor to IQVIA; is a cofounding principal and equity holder in Monument Analytics, a health care consultancy whose clients include the life sciences industry as well as plaintiffs in opioid litigation; and is a member of OptumRx’s National P&T Committee. Dr Blaha has received research/grant support from the National Institutes of Health, FDA, American Heart Association, Aetna Foundation, Amgen, and Novo Nordisk and has served on advisory boards for Amgen, Sanofi, Regeneron, Novartis, Novo Nordisk, Bayer, Akcea, 89Bio, Kaleido, Inozyme, and Kowa. These arrangements have been reviewed and approved by Johns Hopkins University in accordance with its conflict‐of‐interest policies. The remaining authors have no disclosures to report.

Table S1

Figures S1–S2

Click here for additional data file.