Medicare formulary restrictions for glucagon-like peptide 1 receptor agonists and sodium glucose cotransporter 2 inhibitors used in type 2 diabetes mellitus: 2019-2023

Abstract

BACKGROUND:

Glucagon-like peptide 1 receptor agonists (GLP1-RAs) and sodium glucose cotransporter 2 inhibitors (SGLT2is) have proven benefits in patients with type 2 diabetes mellitus related to decreasing cardiovascular events and heart failure hospitalizations as well as preventing the progression of kidney disease. This led the American Diabetes Association (ADA) to update their guidelines in 2022 to recommend GLP1-RAs and SGLT2is as potential first-line options in patients with cardiorenal conditions. Formulary restrictions, such as step therapy and prior authorizations, can limit access to these beneficial medications.

OBJECTIVE:

To evaluate trends in Medicare formulary restrictions of GLP1-RAs and SGLT2is following the recommendations by the ADA for first-line use.

METHODS:

We conducted a retrospective cohort study of Quarter 1 Medicare formulary restrictions on GLP1-RAs and SGLT2is from 2019 to 2023. We analyzed changes in formulary restrictions by year, before and after the ADA first-line recommendations, single vs combination products, and products with and without other indications on a medication level (ie, each unique National Drug Code number on each unique formulary) and formulary level (ie, any prior authorization or step therapy requirements yield formulary restriction).

RESULTS:

We analyzed 12,212 unique Medicare formularies. GLP1-RAs were 12.5% less likely (relative risk [RR] = 0.87; 95% CI = 0.84-0.91; P < 0.001) and 33.2% less likely (RR = 0.67; 95% CI = 0.60-0.74; P < 0.001) to have restrictions after the ADA guideline changes on a medication and formulary level, respectively. SGLT2is were 87.8% less likely (RR = 0.12; 95% CI = 0.11-0.13; P < 0.001) and 53.0% less likely (RR = 0.47; 95% CI = 0.27-0.80; P = 0.005) to have restrictions after the ADA guideline changes on medication and formulary levels, respectively.

CONCLUSIONS:

Decreasing trends of medications and formularies with restrictions may indicate payers are recognizing GLP1-RAs and SGLT2is as potential first-line agents, aligning with ADA recommendations.

Plain language summary

In 2021, guidelines changed to recommend glucagon-like peptide 1 receptor agonists and sodium glucose cotransporter 2 inhibitors as a first-line therapy in people with type 2 diabetes mellitus. Since then, there has been a significant decrease in restrictions on these 2 medication classes. It is unclear if this is because of the updated recommendations or other factors, such as increased US Food and Drug Administration–approved indications.

Implications for managed care pharmacy

Medicare formulary restrictions on glucagon-like peptide 1 receptor agonists and sodium glucose cotransporter 2 inhibitors decreased following the 2021 American Diabetes Association guideline updates to recommend these classes as first-line options; however, restrictions still exist, with step therapy requirements outnumbering prior authorization requirements. Although these formulary restrictions may be necessary to reduce costs and ensure appropriate medication use, formularies should also consider how prior authorization and step therapy restrictions may limit access, especially if evidence of cost-effectiveness were to emerge.

Approximately 10% of Americans have diabetes, with an estimated 90%-95% having type 2 diabetes mellitus (T2DM). Patients with T2DM are at an increased risk of macrovascular complications, such as cardiovascular disease (CVD), as well as microvascular complications, including neuropathy, nephropathy, and retinopathy.¹ These complications contribute to significant morbidity and mortality; CVD accounts for 50% of all deaths in patients with T2DM, and patients with comorbid chronic kidney disease (CKD) have a 10-year cumulative all-cause mortality of 31.1%.^2,3 Additionally, patients with diabetes experience heart failure (HF) at a rate 4 times higher than the general population.⁴ Evidence suggests that comorbid HF with T2DM has a profound impact on patient mortality, leading to a median survival time of 4.1 years after HF diagnosis.⁵ Because of the high prevalence and serious complications of T2DM, it is the most expensive chronic condition in the United States with approximately $1 out of every $4 in US health care costs spent on treating or managing diabetes-related complications. This amounts to $237 billion in direct medical costs and $90 billion in indirect costs.⁶ Evidence-based diabetes management can improve patients’ health outcomes and reduce costs.⁷

Two newer medication classes, glucagon-like peptide 1 receptor agonists (GLP1-RAs) and sodium glucose cotransporter 2 inhibitors (SGLT2is), have become the standard of care for patients with or at high risk of macrovascular and microvascular diabetes-related complications.⁸ Current literature has shown that GLP1-RAs reduce the risk of a major adverse cardiovascular event by 13%-14%, hospital admissions for HF by 11%, composite renal outcomes by 21%, and all-cause mortality in patients with T2DM by 11%-12%.⁹ Additionally, SGLT2is reduce all-cause mortality by 14%, major adverse cardiovascular events by 9%, hospitalizations for HF by 31%, and the risk of CKD progression by 37%.¹⁰

Before 2021, the American Diabetes Association (ADA) guidelines recommended metformin as the preferred first-line agent for the treatment of T2DM in addition to lifestyle modification. As early as 2017, SGLT2is and GLP1-RAs were recommended as second-line agents if therapeutic goals were not met, especially for patients with or at high risk for atherosclerotic CVD (ASCVD), CKD, or HF.⁷ Based on an increasing body of evidence on the benefits of GLP1-RAs and SGLT2is, the 2022 ADA recommendations changed to include SGLT2is or GLP1-RAs as first-line agents for patients with HF, CKD, or risk factors for ASCVD.¹¹ Although GLP1-RAs and SGLT2is are efficacious, many have been subject to formulary restrictions, including the use of prior authorization (PA), step therapy (ST) requirements, or both in part because of their cost.¹² PAs requires approval from third-party payers before a claim for the medication is able to be processed, whereas ST requires patients to try and fail less expensive therapies before more expensive therapies can be used.¹³ Formulary restrictions have been shown to help control costs for third-party payers, yet they may be associated with unintended consequences, including restricted access to medications and negative impacts on provider treatment decisions.¹³ In a 2021 survey of a sample of US office-based physicians, 82% of physicians polled decided against prescribing certain treatments more than 50% of the time in anticipation of drug utilization management–related barriers across insurance.¹⁴

Although the ADA guidelines have changed to include both SGLT2is and GLP1-RAs as first-line agents, it is unclear if coverage of these medications has become less restrictive. This study aims to evaluate trends in formulary restrictions of GLP1-RAs and SGLT2is under Medicare coverage following guideline changes in first-line therapy options.

Methods

STUDY DESIGN

We conducted a before and after study analyzing formulary restrictions for GLP1-RAs and SGLT2is on Quarter 1 Medicare formularies between 2019 and 2023 for the treatment of T2DM. We defined a medication-level restriction as GLP1-RAs or SGLT2is with a PA or ST requirement; each unique National Drug Code (NDC) formulary pair was classified as its own medication (eg, empagliflozin 10 mg and empagliflozin 25 mg would be 2 separate data points for each formulary on which they were included). We defined formulary level restrictions as any formularies that had at least 1 restriction on any medication (single or combination agent) within each class (GLP1-RA or SGLT2is) identified by NDC numbers.

DATA SOURCE

We obtained Medicare formularies from the Centers for Medicare & Medicaid Services Prescription Drug Plan formulary, pharmacy network, and pricing information files.¹⁵ The nonidentifiable files include “formulary, pharmacy network, and pricing data for Medicare Prescription Drug Plans and Medicare Advantage Prescription Drug Plans.” Within the dataset there was a dichotomous variable for PAs and for ST for each NDC in the dataset. The data in the data files are from information sent to the Centers for Medicare & Medicaid Services by Medicare insurance providers.

STUDY POPULATION

In our study, only GLP1-RAs and SGLT2is and their combination products identified by RxCUI codes from Quarter 1 Medicare formularies were used. RxCUI codes are part of the National Library of Medicine RxNorm system, which provides “normalized names and unique identifiers for medicines… to allow computer systems to communicate drug-related information efficiently and unambiguously.”¹⁶

We chose to only examine quarter 1 formularies because of data availability for 2023.

OUTCOMES

Our primary outcome was the change in the percentage of formulary restrictions by year for GLP1-RAs and SGLT2is on a medication and formulary level. Secondary outcomes included changes in the percentage of plans and/or medication with restrictions before and after the ADA guidelines changes (where 2019-2021 were considered “before” and 2022-2023 were considered “after”), proportion of restrictions on single or combination products, and proportion of restrictions on SGLT2is or GLP1-RAs with additional US Food and Drug Administration (FDA) indications (Supplementary Table 1 ^{(156.4KB, pdf)} , available in online article). For FDA indications, we treated combination products as having the same indications as their individual active ingredients.

STATISTICAL ANALYSIS

We described our data using counts and proportions. We used a Cochran-Armitage trend test to determine if there was a statistically significant trend in the proportion of medications and formularies with restrictions. We performed chi-square tests to determine if there was a statistically significant difference in medications and formularies with restrictions between any 2 years. We also used a chi-square test to assess differences in formulary restrictions in the before and after periods because we could not match formularies between years. Finally, we used logistic regression to quantify the effect of year (ie, the independent variable) on the risk of medications restrictions (ie, the dependent variable) at the formulary level. Each test was run independently on GLP1-RA and SGLT2i restrictions by medication and by formulary.

Additionally, we examined restrictions placed on single active ingredient products compared with combination products using a chi-square test. We also examined individual GLP1-RAs with and without FDA approval for concomitant T2DM and ASCVD as well as SGLT2is with additional indications for CKD, HF, or ASCVD compared with those without using chi-square tests. We used logistic regression to see the effect of each additional FDA-approved indication (ie, independent variable) on the risk of formulary restriction (ie, dependent variable). All analyses were conducted using SAS Enterprise Guide version 7.1 (SAS Institute Inc.).

Results

We analyzed all 2,212 unique Medicare formularies across 5 years. The average number of formularies per year was 442 (Supplementary Table 2 ^{(156.4KB, pdf)} ).

GLP1-RAs

Medication Level. Between 2019 and 2023, we analyzed 24,857 GLP1-RA medication coverage policies across all Medicare formularies. On average, there were 11.2 unique GLP1-RA NDC numbers included in a formulary. From 2019 to 2023, the odds of a GLP1-RA medication having a restriction decreased by 6.7% with each increase in year (odds ratio [OR] = 0.93; 95% CI = 0.91-0.95; P < 0.001). There was a statistically significant medication-level downward trend in the proportion of GLP1-RA medications with restrictions over this period (P < 0.001). In 2019, formulary restrictions were placed on 29.2% (1,034/3,528) of medications containing a GLP1-RA, which decreased to 20.2% (1,207/5,971) in 2022. However, there was an increase in the medication-level restrictions on GLP1-RAs in 2023 when 24.7% (1,492/6,084) had restrictions (Figure 1). GLP1-Ras were 13% (relative risk [RR] = 0.87; 95% CI = 0.84-0.91; P < 0.001) less likely to have restrictions after the ADA guideline changes. Combination GLP1-RA products were more likely to have formulary restrictions (RR = 1.18; 95% CI = 1.09-1.27; P < 0.001) (Table 1). The GLP1-RAs least likely to have formulary restrictions were semaglutide, dulaglutide, liraglutide, and combination lixisenatide/insulin glargine. The restrictions for each agent from 2019 to 2023 are shown in Supplementary Table 3 ^{(156.4KB, pdf)} .

FIGURE 1.

Proportion of Medications with Restrictions on GLP1-RAs and SGLT2is

TABLE 1.

Medication Restrictions by Formulary, ST, and PA Before and After the American Diabetes Association Guideline Changes

Medication	Restriction	Frequency of restrictions before, n (%)	Frequency of restrictions after, n (%)	Relative risk (95% CI)	P value
GLP1-RA	Medication level	3,295/12,838 (25.7)	2,699/12,019 (22.5)	0.87 (0.85-0.91)	<0.001
	Formulary level	627/1,305 (48.0)	291/907 (32.1)	0.67 (0.60-0.74)	<0.001
	ST	2,302/12,838 (17.9)	1,597/12,019 (13.3)	0.74 (0.70-0.79)	<0.001
	PA	994/12,838 (7.7)	1,102/12,019 (9.2)	1.18 (1.09-1.29)	<0.001
SGLT2i	Medication level	4,979/26,461 (18.8)	478/20,899 (2.3)	0.12 (0.11-0.13)	<0.001
	Formulary level	52/1,305 (4.0)	17/907 (1.9)	0.47 (0.27-0.80)	0.005
	ST	4,792/26,461 (18.1)	398/20,899 (1.9)	0.11 (0.10-0.12)	<0.001
	PA	188/26,461 (0.7)	80/20,899 (0.4)	0.54 (0.42-0.70)	<0.001

Chi-square tests were used to complete the analyses.

GLP1-RA = glucagon-like peptide 1 receptor agonist; PA = prior authorization; SGLT2i = sodium glucose cotransporter 2 inhibitor; ST = step therapy.

In total, 24% (5,994/24,857) of GLP1-RAs analyzed had some type of formulary restriction, with 1 medication requiring both PA and ST, for a total restriction count of 5,995. From 2019 to 2023, 15.7% (3,899/24,857) of GLP1-RAs were subject to ST and 8.4% (2,096/24,857) were subject to PA (Table 1). Of the 5,995 total restrictions, 65.0% (n = 3,899) were ST and 35.0% (n = 2,096) were PA. Before the ADA guideline changes, 17.9% (2,302/12,838) of GLP1-RAs required ST, which decreased to 13.3% (1,597/12,019), a statistically significant decrease (RR = 0.74; 95% CI = 0.70-0.79; P < 0.001).

Combination Products and FDA-Approved Indications. Only 8.0% (313/3,899) of GLP1-RAs requiring ST were for combination products. Alternatively, 7.7% (994/2,096) of GLP1-RAs required PA before the ADA guideline changes, which increased to 9.2% (n = 1,102) in the post-ADA guideline changes (RR = 1.18; 95% CI = 1.09-1.29; P < 0.001) (Table 1). Furthermore, on a medication level, GLP1-RAs with FDA indications for CVD benefits were 22% less likely to have formulary restrictions (RR = 0.78; 95% CI = 0.74-0.81; P < 0.001). Formulary restrictions were placed on 4,120 of 18,375 (22.4%) GLP1-RAs with established FDA indications for CVD benefits compared with 1,874 of 6,482 (28.9%) GLP1-RAs without indications for CVD benefits (Table 2).

TABLE 2.

Medicare Restrictions for GLP1-RAs and SGLT2is With Additional FDA Indications

Medications class and FDA indication	Frequency of formulary restrictions 2019-2023, n (%)	Relative risk (95% CI)	P value
GLP1-RA w/ CV indication	4,120/18,375 (22.4)	0.78 (0.74-0.81)	<0.001
GLP1-RA w/o CV indication	1,874/6,482 (28.9)
SGLT2is w/ CV indication	5,075/46,294 (11.0)	0.31 (0.28-0.33)	<0.001
SGLT2is w/o CV indication	382/1,066 (35.8)
SGLT2is w/ CKD indication	2,699/19,542 (13.8)	1.39 (1.33-1.46)	<0.001
SGLT2is w/o CKD indication	2,758/27,818 (9.9)
SGLT2is w/ HF indication	2,788/36,899 (7.6)	0.30 (0.28-0.31)	<0.001
SGLT2is w/o HF indication	2,699/10,461 (25.5)

Chi-square tests were used to complete the analyses.

CKD = chronic kidney disease; CV = cardiovascular; FDA = US Food and Drug Administration; GLP1-RA = glucagon-like peptide 1 receptor agonist; HF = heart failure; SGLT2i = sodium glucose cotransporter 2 inhibitor; w/ = with; w/o = without.

Formulary Level. A total of 918 of 2,212 formularies had restrictions on GLP1-RAs from 2019 to 2023 (41.5%). From 2019 to 2023, the percentage of formularies with GLP1-RA restrictions decreased from 58.4% (234/401) to 30.4% (137/450) (Figure 2). Formularies were 33% less likely to have GLP1-RA restrictions after the ADA guideline changes (before = 48% [627/1,305] vs after = 32.1% [291/907]; RR = 0.67, 95% CI = 0.60-0.74, P < 0.001).

FIGURE 2.

Proportion of Formularies With Restrictions on GLP1-RAs and SGLT2is

SGLT2is

Medication Level. Between 2019 and 2023, we analyzed 47,360 SGLT2i medication coverage policies across all Medicare formularies. On average, there were 21.4 unique NDC numbers for SGLT2is included per formulary. From 2019 to 2023, the odds of an SGLT2i medication having a restriction decreased by 52% for each additional year (OR = 0.48; 95% CI = 0.46-0.49; P < 0.001). There was a statistically significant downward trend in the proportion of SGLT2i medications with restrictions for formularies during this period (P < 0.001). In 2019, 23.8% (1,859/7,816) of medications containing SGLT2is had restrictions, which increased to 25.6% (2,208/8,711) in 2020. The percentage of restrictions of SGLT2is decreased from 2020 through 2023, when 2.0% (204/10,453) had restrictions (Figure 1). SGLT2is were 88% less likely to have restrictions after the ADA guideline changes (RR = 0.12; 95% CI = 0.11-0.13; P < 0.001) (Table 1). There was no significant difference in the proportion of single agent products compared with combination products that had restrictions (RR = 1.04; 95% CI = 0.98-1.11; P = 0.21). The SGLT2is that were least likely to be subject to formulary restrictions were dapagliflozin, dapagliflozin/metformin, and empagliflozin/linagliptin/metformin. A full list of SGLT2is with restrictions from 2019 to 2023 is provided in Supplementary Table 3 ^{(156.4KB, pdf)} .

On a medication level, 11.5% (5,457/47,360) of SGLT2is had a formulary restriction with 1 medication requiring both PA and ST, making the total number of restrictions 5,458 for this class. From 2019 to 2023, 11.0% (5,190/47,360) of SGLT2is were subject to ST whereas 0.6% (268/47,360) of SGLT2is were subject to PA. In that time 95.1% (5,190/5,458) of the total restrictions were ST and 4.9% (268/5,458) were PA. Before the guideline changes, 18.1% (4,792/26,461) of SGLT2is required ST. After the guideline changes, the number of SGLT2is requiring ST decreased to 1.9% (398/20,899) (RR = 0.11; 95% CI = 0.10-0.12; P < 0.0001).

Combination Products and FDA Indications. Out of 5,190 SGLT2is with ST requirements, 4,137 (79.7%) were combination products. A total of 0.7% (188/26,461) SGLT2is required PA before the guideline change, which decreased to 0.4% (80/20,899) after the guidelines (RR = 0.54; 95% CI = 0.42-0.70; P < 0.001) (Table 1).

Restrictions for SGLT2is decreased by 57% for each additional non-T2DM FDA indication (OR = 0.43; 95% CI = 0.41-0.45; P < .001). Of the SGLT2is with 3 non-T2DM FDA-approved indications, 4.1% (412/10,147) had restrictions and, of those with 2 non-T2DM indications, 12.9% (4,663/36,147) had restrictions. Of SGLT2is with no additional FDA indications 35.8% (385/1,066) had restrictions (Table 2). SGLT2is with additional FDA indications were 69% less likely to have formulary restrictions (RR = 0.31; 95% CI = 0.28-0.33; P < .001). SGLT2is with FDA indications for CKD were 39% more likely to have formulary restrictions placed on them (RR = 1.39; 95% CI = 1.33-1.46; P < 0.001). SGLT2is with FDA indications for patients with HF were 70% less likely to have formulary restrictions (RR = 0.30; 95% CI = 0.28-0.31; P < 0.0001) (Table 2).

Formulary Level. Of the 2,212 unique Medicare formularies analyzed, a total of 69 formularies had restrictions on SGLT2is from 2019 to 2023. In 2019, 1.8% (7/401) of formularies had restrictions on SGLT2is, which increased to 8.5% (38/447) in 2020. After the ADA guideline change in 2021, the percentage declined from 1.5% (7/457) in 2021 to 0.9% (4/450) in 2023 (Figure 2). After the ADA guideline change, formularies were 55% less likely to have restrictions on SGLTis (before = 4% [52/1,350] vs after = 1.9% [7/907]; RR = 0.47, 95% CI = 0.27-0.81, P < 0.001).

Discussion

To our knowledge, this is the first study to examine Medicare formulary restrictions on GLP1-RAs and SGLT2is over time and in relation to ADA guideline changes. Between 2019 and 2023, the proportion of medications and formularies with restrictions for GLP1-RAs and SGLT2is decreased overall, with a significant drop in restrictions following the ADA guideline changes. We also found significant differences in restrictions for medications with and without other FDA-approved indications as well as for combination and single active ingredient products. ST restrictions were more prevalent than PA restrictions throughout the time frame.

Formulary restrictions have been cited as a reason for the underuse of GLP1-RAs and SGLT2is.¹⁷ We found that restrictions on GLP1-RAs and SGLT2is decreased over time on both the medication level and the formulary level, which suggests that prescribing levels should be increasing. Studies on population-level GLP1-RA and SGLT2i use have increased but remain low in subsets of populations that would most benefit from their use, such as Medicare patients who have comorbid ASCVD, CKD, or HF.^18-21 A 2022 study demonstrated that Medicare beneficiaries were 18% more likely to initiate either an GLP1-RA or SGLT2i when there were at least 2 of the desired medications in formulary tiers 1-3.¹⁷ This suggests that a preferred status for GLP1-RAs and SGLT2is will improve medication initiation. Additionally, a survey of physicians found that they are less likely to prescribe medications for which they anticipate formulary restrictions.¹⁴ Another systematic review looking at formulary restrictions (ST, PA, or both) and their impact on medication adherence revealed that such restrictions resulted in negative medication adherence outcomes in 70.6% of cases.¹³ This may suggest that, although formulary restrictions are a reason for suboptimal prescribing, other factors, such as patient out-of-pocket costs and drug shortages because of off-label prescribing may be limiting the use of these medications. Alternatively, prescribers may be unaware of insurance-specific formulary restrictions given the large number of insurances that cover their patients and the yearly changes in formularies that can occur.²²

Not only did GLP1-RA and SGLT2i restrictions decrease over time but they were also significantly lower after the ADA guideline changes at both the medication and formulary level. Restrictions were especially low for SGLT2is after the guideline change, with only 2.3% of SGLT2is having a formulary restriction after the guideline change. This may suggest that Medicare formularies have already started to reflect the recommendations set by the ADA, thereby improving patient access to GLP1-RAs and SGLT2is. It should be noted, however, that ST restrictions outnumber PA restrictions overall. Additionally, the majority of ST requirements for GLP1-RA products were on single agent products, suggesting that some Medicare formularies are still requiring patients to try a less expensive medication before progressing to a GLP1-RA. Based on the ADA guideline change, we would have expected fewer ST requirements than PA as ST places these agents as a second-line therapy and likely requires the use of metformin at first-line therapy. Although PA restrictions may still limit access to GLP1-RAs and SGLT2is, it may be reasonable for insurance companies to require PA for these medications to limit their use for non-FDA–approved indications, especially considering that Medicare does not cover any weight loss medication for which GLP1s are being used extensively. The ADA guidelines recommend these agents as first-line treatments specifically for individuals with ASCVD, CKD, and HF; therefore, the PA requirements may be to ensure the individuals who are prescribed GLP1-RAs and SGLT2is as first-line therapy have the comorbidities mentioned by the FDA. GLP1-RAs and SGLT2is are substantially more expensive than metformin, and it is reasonable for payers to prefer metformin to GLP1-RAs and SGLT2is in individuals without ASCVD, HF, or CKD. Based on this, PA could be deemed appropriate, especially considering GLP1-RAs and SGLT2is have not yet proven to be cost-effective.²³

In general, GLP1-RAs and SGLT2is with FDA indications for other disease states were less likely to have restrictions. Although this could be because of insurance companies recognizing the benefit of these agents for T2DM, for SGLT2is specifically the decreased restrictions could be driven by additional FDA-approved indications even in patients without T2DM. SGLT2is are considered first-line agents for HF regardless of ejection fraction or CKD with an estimated glomerular filtration rate above 20 mL/min/1.73 m².^24,25 Specific drugs approved for these indications can be found in Supplementary Table 1 ^{(156.4KB, pdf)} . Because of the expanding indications for patients without T2DM, we would expect formularies to decrease restrictions on these medications as they are first-line agents for other disease states. This is supported by individual medications with fewer restrictions, including dapagliflozin, canagliflozin, and empagliflozin, all of which have additional FDA-approved indications, compared with ertugliflozin, which has no additional FDA-approved indications. A similar trend can be seen with GLP1-RAs in that those with FDA indications for CVD benefits have fewer restrictions than those that do not. Alternatively, GLP1-RAs can cause weight loss and, thus, are being more extensively studied in obesity.²⁶ Although FDA-approved GLP1-RA medications for weight loss are marketed under a different brand name, off-label use of T2DM medications for weight loss has resulted in drug shortages.²⁷ For Medicare beneficiaries, weight loss medications are not covered under their prescription benefit, which may make it more likely for a medication marketed as a diabetes agent to be prescribed for weight loss making restrictions to monitor for this off-label use important. This may have contributed to the increase in formulary restrictions placed on GLP1-RAs from 2022 to 2023 and would be appropriate to prevent the use of GLP1-RAs for non-FDA–approved indications. This is further supported by combination GLP1-RA/insulin products having fewer restrictions, as these medications do not have a place in the treatment of obesity alone. Another reason why GLP1-RA/insulin products may have fewer restrictions is that combination products decrease regimen complexity for patients, especially those with multiple comorbidities. This could increase patient adherence to complex medication regimens and incentivize fewer restrictions. However, if this were the case, we would expect to see a similar trend with combination products involving SGLT2is, which was not seen. Alternatively, when initiating a patient on T2DM therapy, prescribers may prefer to add a single medication at a time to avoid overprescribing and hypoglycemia risk from T2DM medications.²⁸

We only looked at the requirement for formulary restrictions and were unable to categorize the criteria within the PA or ST requirement, making it difficult to draw definite conclusions about the reasons these restrictions exist and how well they align with FDA-approved indications and guidelines. Although having formulary restrictions restricts access to medications, in many situations they are necessary and appropriate to reduce overall pharmacy costs and ensure appropriate use of medications. It should also be noted that formulary restrictions are not the only limiting factor when it comes to the accessibility of these medications. Specifically, patient out-of-pocket cost may also limit whether a provider is willing to prescribe and the patient is willing to take a medication.^29,30 As formulary restrictions are typically applied to more expensive medications, patients who are prescribed GLP1-RAs and SGLT2is may still have substantial copays, especially in the deductible or coverage gap (donut hole) phases, which can be a limiting factor in their actual use. It is unknown how the Inflation Reduction Act (IRA) may alter the practice of third-party entities, meaning that the trends we observed may not continue or be accurate in years to come. By 2025, the IRA will cap Medicare beneficiary out-of-pocket costs at $2,000 per year and eliminate the coverage gap phase.³¹ This decision has raised speculation on how Medicare plans may account for this decrease in cost-sharing for patients. Although some of this deficit may be offset by reductions in drug prices because of a different clause in the IRA, third parties increasing formulary restrictions on expensive medications is a possibility.³²

LIMITATIONS

As only Medicare data were analyzed, the results are not applicable to individuals insured through Medicaid or employer-sponsored private insurance given the unique structure and features, such as not covering any medications indicated for weight loss. Another limitation is that we did not have the information necessary for PA requirement approval and were unable to characterize the PAs regarding whether the prescribing patterns aligned with FDA indications. Furthermore, our study is only able to evaluate restrictions on medications that are on the Medicare formularies. If a medication does not appear on a formulary, it was not included in our analysis and so we may overestimate the accessibility of GLP1-RAs and SGLT2is, especially given our exclusion of GLP1-RAs approved for weight loss.

Conclusions

There was a statistically significant decreasing trend for medications and formularies with restrictions, including ST and PA, for GLP1-RAs and SGLT2is following the 2022 ADA guideline changes, which recommended these medication classes as potential first-line options for patients with compelling indications. This indicated a possible improvement in the access of these medications. It is unclear as to whether this change is solely reflective of Medicare formularies recognizing changes in guideline recommendations or because of other factors, such as additional health outcomes data, new FDA-approved indications for patients with or without T2DM and other comorbidities, or the prevention the use of non-FDA–approved indications.