Type 2 diabetes mellitus (T2DM) affects approximately 30.3 million individuals in the United States, a number that is expected to grow to almost 36 million by the year 2030.1,2 Diabetes is responsible for $176 billion in annual direct medical costs and for $69 billion annually in disability, lost work productivity, and premature mortality in the United States.3 The goals of diabetes treatment are to achieve glycemic control and to reduce the risk for long-term complications.4 To be able to achieve and maintain their glycemic control goals, and to avoid complications, patients should remain adherent to their prescribed treatment regimen.5
Treatments are currently available from 14 classes of antidiabetes medications, including, but not limited to, biguanides, sulfonylureas, thiazolidinediones, dipeptidyl peptidase (DPP)-4 inhibitors, glucagon-like peptide (GLP)-1 agonists, sodium-glucose cotransporter (SGLT)-2 inhibitors, meglitinides, alpha-glucosidase inhibitors, amylin analogs, and insulin, which provide patients, physicians, and health plans multiple therapy options at varying costs. Unless contraindicated, metformin, a member of the biguanide class, is the recommended initial therapy for T2DM.6 Glycemic control is evaluated in diabetic patients approximately every 3 months, and patients who do not achieve appropriate glycemic control can add agents to their treatment regimen.6
In an effort to control healthcare costs, health plans have responded by updating diabetes medication formularies and implementing new formulary restrictions, such as multitier formularies and step therapy programs.7 Multitier formularies, and the related use of preferred medication lists, are designed to encourage the use of lower-cost alternatives within a medication class through increased consumer cost-sharing.7 Step therapy programs encourage the use of lower-cost medications, by requiring participation in a trial of a lower-cost or preferred drug before a patient can receive a higher-cost or a nonpreferred drug.7
Previous studies have explored the effects of cost-sharing and step therapy programs for a variety of indications and medication classes, reporting that medication adherence and utilization decrease, and clinical outcomes worsen, after the introduction of such formulary restrictions.7-15 In a previous article, Huang and colleagues assessed the impact on medication use of a formulary restriction that moved sitagliptin—a DPP-4 inhibitor that is indicated by the US Food and Drug Administration for the treatment of T2DM—from the second to the third tier.16 They found that approximately 36% of patients discontinued sitagliptin after the restriction was implemented. Among those who discontinued sitagliptin therapy, approximately 44% switched to a preferred DPP-4 inhibitor. Patients who discontinued sitagliptin were more likely to switch to a GLP-1 agonist or to insulin after the sitagliptin tier change than those who continued using sitagliptin.16
Despite the relatively large body of literature showing that such formulary restrictions are associated with decreased medication utilization overall, reduced adherence, and unfavorable clinical outcomes,7-12,14,15 very little is known about the effects of step therapy programs, specifically for antidiabetes medications.14,17-21 Our study contributes to this literature, and addresses the dearth of studies focusing on antidiabetes medications by assessing the effects of a sitagliptin step therapy program that required a trial of a preferred DPP-4 inhibitor (ie, linagliptin or saxagliptin) before sitagliptin claims would be reimbursed. Specifically, patients who had not previously received DPP-4 inhibitor therapy were required to use a preferred DPP-4 inhibitor, and current users of sitagliptin were required to switch to one of the preferred DPP-4 inhibitors.
Study Design and Sample
This descriptive, retrospective cohort study used pharmacy claims data from the Symphony Health Solutions’ Integrated Dataverse (IDV), a platform that combines clinical and demographic information with hospital, medical, and pharmacy claims for 274 million active lives in the United States. We identified 3 health plans, each from neighboring Northeastern US states, whose patient data were available in the IDV: Plan A, covering 666,212 lives, had already placed sitagliptin into tier 3 when it implemented a sitagliptin step therapy program on July 1, 2013; Plan B, covering 228,747 lives, also placed sitagliptin in tier 3, but did not implement a sitagliptin step therapy program; and Plan C, covering 667,997 lives, had placed sitagliptin in tier 2 and also did not implement a sitagliptin step therapy program (Plans B and C cover geographic areas adjacent to Plan A’s coverage area). Open enrollment for the 3 plans started on October 1, 2012, and lasted 2 months. The annual coverage cycles began on January 1, 2013, for the 3 plans.
Study observations were divided into the following 3 periods: (1) January 1, 2013, through June 30, 2013, which corresponded to the 6 months before Plan A implemented its step therapy program (ie, the pre–step therapy period); (2) July 1, 2013, through September 30, 2013, which corresponded to the 3-month period during which we assumed the new policy was fully implemented and previous sitagliptin prescriptions were exhausted; and (3) October 1, 2013, through March 31, 2014, which corresponded to the 6 months after Plan A implemented its sitagliptin step therapy program (ie, the post–step therapy period).
Patients were included in our analyses if they had been enrolled in any of the 3 health plans with continuous medical and prescription coverage between January 1, 2013, and March 31, 2014; were not enrolled in more than 1 plan; were aged ≥18 years; filled at least 1 sitagliptin prescription during the pre–step therapy period; and did not use any DPP-4 inhibitors other than sitagliptin during the pre–step therapy period (Figure 1).
Study Measures and Analyses
We evaluated the utilization of sitagliptin during the post–step therapy period and the utilization of other antidiabetes medications during the pre– and post–step therapy periods. The key end point, sitagliptin use during the post–step therapy period, was defined by the presence of sitagliptin prescription claims. Patients who received prescriptions for sitagliptin during the post–step therapy period were considered to have continued using sitagliptin. Patients without any evidence of prescriptions for sitagliptin were considered to have discontinued taking sitagliptin. Other antidiabetes medication use was assessed for patients who discontinued sitagliptin during the post–step therapy period. Those patients could have switched to a preferred DPP-4 inhibitor (saxagliptin or linagliptin); discontinued all other antidiabetes medications (no prescriptions for any type of antidiabetes medication); continued other antidiabetes medications without replacing sitagliptin (did not change prescriptions of antidiabetes medications other than discontinued sitagliptin); started an SGLT-2 inhibitor, GLP-1 agonist, and/or insulin; and started other antidiabetes medications.
We also evaluated the total number of treatment classes (ie, alpha-glucosidase inhibitors, amylin analogs, biguanides, meglitinides, sulfonylureas, thiazolidinediones, DPP-4 inhibitors, GLP-1 agonists, SGLT-2 inhibitors, and insulin) during the pre– and post–step therapy periods. Pre–step therapy antidiabetes medication use in addition to sitagliptin was also assessed (ie, sitagliptin monotherapy or combination therapy).
Patient age, sex, and several comorbid conditions based on International Classification of Diseases, Ninth Revision, Clinical Modification codes (ie, presence of benign essential hypertension, hyperlipidemia, pure hypercholesterolemia, coronary atherosclerosis of native coronary artery, unspecified obesity, shortness of breath, and unspecified chest pain) were assessed during the pre–step therapy period. Patient characteristics and pre– and post–step therapy period use of sitagliptin and other antidiabetes medications were summarized separately according to the health plan.
We identified 2995 patients enrolled in Plan A, 751 patients in Plan B, and 394 patients in Plan C who used sitagliptin in the pre–step therapy period and met the study eligibility criteria. The patient characteristics were similar across health plans during the pre–step therapy period (Table 1).
The patients were, on average, aged 56 to 57 years, and 64% to approximately 70% were male across the 3 health plans. Although approximately 33% of patients had hypertension and approximately 25% had hyperlipidemia, slightly more than 50% of patients had none of the 7 comorbid conditions assessed (Table 1). Concomitant insulin use was similar across the 3 plans (13.5% in Plan C to 14.7% in Plan A).
Approximately 10% of patients overall received sitagliptin monotherapy during the pre–step therapy period; approximately 30% (Plan C) to 38% (Plans A and B) received sitagliptin plus metformin (including fixed-dose combinations); 6% to 8% used sitagliptin and 1 other oral antidiabetes drug; and 44% to 52% of patients used other sitagliptin combination therapies (Table 2). Regardless of their plan, patients received antidiabetes medications from similar numbers of treatment classes (2.5-2.6) during the pre–step therapy period.
The use of sitagliptin and other antidiabetes medications during the post–step therapy period differed, sometimes drastically, across the health plans (Table 3). The patients enrolled in Plans A and B used antidiabetes medications from an average of 1.9 drug classes, and patients enrolled in Plan C used medications from an average of 2.2 drug classes, including DPP-4 inhibitors, during the post–step therapy period. In all, approximately 70% of the patients in Plan A (with the step therapy program), stopped using sitagliptin during the post–step therapy period compared with approximately 51% and 25% of the patients in Plans B and C (without the step therapy program), respectively. Approximately 30% of the patients in Plan A switched to linagliptin or to saxagliptin therapy compared with approximately 15% and 2% of the patients in Plans B and C, respectively.
Patients enrolled in Plan A were more likely to stop using sitagliptin therapy without replacement (17%) than patients enrolled in Plans B (approximately 13%) and C (approximately 8%); patients enrolled in Plan A (5%) were slightly more likely to have stopped sitagliptin treatment and started SGLT-2 inhibitors, GLP-1 agonists, or insulin than those enrolled in Plan B (approximately 4%) and Plan C (approximately 3%); differences that were not clinically significant. Although significant proportions of the patients continued to use sitagliptin treatment after the step therapy period; continued to use their other previous antidiabetes medications without sitagliptin; or started to use new medications, approximately 14% of patients in Plans A and B and 8% of patients in Plan C discontinued sitagliptin treatment, as well as all other antidiabetes medications. Approximately 3% to 4% of patients in all 3 plans discontinued sitagliptin treatment and received no other antidiabetes medication during that period. Figure 2 summarizes the changes in antidiabetes medication use.
The results of our analyses contribute to the literature on the impact of formulary restrictions by demonstrating that a step therapy program requiring a trial of a preferred DPP-4 inhibitor changed the overall use of DPP-4 inhibitors and other antidiabetes medications in patients with T2DM. We found that approximately 70% of patients enrolled in Plan A—the health plan that implemented the sitagliptin step therapy program and placed sitagliptin in the third tier—who received sitagliptin during the first half of 2013 were no longer using sitagliptin during the 6-month post–step therapy follow-up period between October 1, 2013, and March 31, 2014.
Similar patients enrolled in plans without the step therapy program were much less likely to have discontinued sitagliptin treatment during the same time (approximately 51% in Plan B, which placed sitagliptin in the third tier, and 25% in Plan C, which placed sitagliptin in the second tier). As expected, these results show that more restrictive formulary policies resulted in lower utilization of sitagliptin treatment.
Although the step therapy program implemented in Plan A appeared to have reduced sitagliptin use and redirected patients to preferred DPP-4 inhibitors, it was also associated with other outcomes, such as less intensive therapy (ie, the discontinuation of sitagliptin treatment without replacement) for some patients and the discontinuation of all medications for other patients. Starting insulin treatment or taking other branded, and potentially costly, antidiabetes medications (such as GLP-1 agonists, SGLT-2 inhibitors, or insulin) was another outcome associated with the sitagliptin step therapy requirement. These findings are consistent with the expected impact of the step therapy program in Plan A and with the tiers in which sitagliptin was placed in all 3 plans.
Although patients in Plan A were more likely to discontinue sitagliptin treatment, with or without switching to another DPP-4 inhibitor, than patients in the other plans, the differences between patients in Plans A and B are much smaller than the differences between patients in Plans A and C. This is likely a result of sitagliptin’s placement in tier 3 in Plans A and B and in tier 2 in Plan C. Furthermore, these results are also consistent with how restrictive the plans were. Plan A was the most restrictive (step therapy program and sitagliptin was in tier 3), and Plan C was the least restrictive (no step therapy program and sitagliptin was in tier 2).
Our findings are also consistent with the recent literature on the effects of cost-sharing and step therapy programs. Reviews of more than 170 articles have shown that medication adherence and utilization decrease after the introduction of formulary restrictions, including the complete discontinuation of all medications.8,11,15 Despite the size of the evidence base on the effects of formulary restrictions, very few studies have focused on antidiabetes medications; however, those studies’ results are consistent with those for other conditions and other types of medications: lower copayments and medication tier reductions are related to lower healthcare costs, improved adherence, and enhanced clinical outcomes.13,19,20 Our study did not examine clinical outcomes; however, as previous studies suggest, formulary restrictions, including step therapy programs, are associated with poorer glycemic control (ie, worsening glycated hemoglobin [HbA1c] levels)18 and a lower chance of patients achieving their HbA1c goals than patients who are not managed with such programs.14
This study has several limitations that should be considered when interpreting these results. First, like other analyses of administrative data, we were unable to determine if patients consumed their filled prescriptions or if other therapies were optimized when sitagliptin was discontinued, nor could we assess clinical outcomes, such as changes in HbA1c levels.
Second, we were unable to account for other plan-related factors, such as the formulary tiers assigned to antidiabetes medications other than sitagliptin; patients’ out-of-pocket healthcare costs or their socioeconomic status, which may be related to medication adherence; coupons or patient assistance programs; or other formulary restrictions that may have been placed on antidiabetes drugs other than sitagliptin.
Third, these results are based on unadjusted analyses; however, the pre–step therapy characteristics we assessed were similar across the health plans. Furthermore, our results do account for secular trends in antidiabetes medication utilization from our use of 2 contemporaneous comparison groups.
Fourth, these analyses focused only on existing users of sitagliptin, rather than on new users of sitagliptin, during the pre–step therapy period.
Fifth, these results may not be generalizable to other health plans, including Medicare and Medicaid, or to other geographic regions.
Finally, the length of the post–step therapy period (ie, 6 months) may not have been long enough to capture the long-term effects of formulary restrictions or to rule out the impact of other factors.
T2DM is a chronic disease that is well-controlled in some patients, but not for many others. Depending on their initial HbA1c level, approximately 30% to 72% of patients with T2DM will need to add another drug within 6 months of starting antidiabetes therapy to attain their glycemic control goals.22 Policies that result in less intensive therapy for a substantial proportion of patients may result in worse glycemic control and related clinical outcomes for high-risk patients than less restrictive policies.14
Further research linking the attributes of formulary restriction policies, patient out-of-pocket costs, changes in medication use patterns, and changes in clinical outcomes is needed to understand more fully the impact of such policies. Although the step therapy program implemented in Plan A was associated with reduced sitagliptin use and redirected patients to preferred DPP-4 inhibitors, it was also associated with less intensive therapy, and with the total discontinuation of antidiabetes medications for some patients.
Source of Funding
This study and this manuscript were funded by Merck & Co, Inc.
Author Disclosure Statement
Dr Tang, Dr Shankar, and Dr Rajpathak are employees of and own stocks in Merck & Co, Inc.; Dr Huang is an employee of and owns stocks in Mallinckrodt Pharmaceuticals; Dr Liu is an employee of Merck & Co, Inc. Dr Ganz is an employee of Evidera, which provides consulting services to various healthcare companies; he is therefore precluded from receiving direct payments or honoraria from these organizations.
Dr Tang is Associate Director, MRL, Merck & Co, Inc., Kenilworth, NJ; Dr Huang was Director, MRL, Merck & Co, Inc., Kenilworth, NJ, during this study; Dr Liu is Associate Director, MRL, Merck & Co, Inc., Kenilworth, NJ; Dr Shankar is Senior Principal Scientist, MRL, Merck & Co, Inc., Kenilworth, NJ; Dr Ganz is Senior Research Scientist, Evidera, Waltham, MA; Dr Rajpathak is Executive Director, MRL, Merck & Co, Inc., Kenilworth, NJ.
1. American Diabetes Association. Statistics about diabetes. July 19, 2017. www.diabetes.org/diabetes-basics/statistics. Accessed November 6, 2017.
2. Shaw JE, Sicree RA, Zimmet PZ. Global estimates of the prevalence of diabetes for 2010 and 2030. Diabetes Res Clin Pract. 2010;87:4-14.
3. American Diabetes Association. Economic costs of diabetes in the U.S. in 2012. Diabetes Care. 2013;36:1033-1046. Erratum in: Diabetes Care. 2013;36:1797.
4. American Diabetes Association. 6. Glycemic targets. Diabetes Care. 2015;38(suppl 1):S33-S40.
5. García-Pérez LE, Alvarez M, Dilla T, et al. Adherence to therapies in patients with type 2 diabetes. Diabetes Ther. 2013;4:175-194.
6. American Diabetes Association. Standards of Medical Care in Diabetes—2017 abridged for primary care providers. Clin Diabetes. 2017;35:5-26.
7. Carlton RI, Bramley TJ, Nightengale B, et al. Review of outcomes associated with formulary restrictions: focus on step therapy. Am J Pharm Benefits. 2010;2:50-58.
8. Cox ER, Henderson R, Motheral BR. Health plan member experience with point-of-service prescription step therapy. J Manag Care Pharm. 2004;10:291-298.
9. Eaddy MT, Cook CL, O’Day K, et al. How patient cost-sharing trends affect adherence and outcomes: a literature review. P T. 2012;37:45-55.
10. Happe LE, Clark D, Holliday E, Young T. A systematic literature review assessing the directional impact of managed care formulary restrictions on medication adherence, clinical outcomes, economic outcomes, and health care resource utilization. J Manag Care Spec Pharm. 2014;20:677-684.
11. Motheral BR, Henderson R, Cox ER. Plan-sponsor savings and member experience with point-of-service prescription step therapy. Am J Manag Care. 2004;10(pt 1):457-464.
12. Motheral BR. Pharmaceutical step-therapy interventions: a critical review of the literature. J Manag Care Pharm. 2011;17:143-155.
13. Park Y, Raza S, George A, et al. The effect of formulary restrictions on patient and payer outcomes: a systematic literature review. J Manag Care Spec Pharm. 2017;23:893-901.
14. Williams SA, Buysman EK, Hulbert EM, et al. Hemoglobin A1c outcomes and health care resource use in type 2 diabetes mellitus patients treated with combination oral antidiabetes drugs through step therapy and loose-dose and fixed-dose combinations. Manag Care. 2012;21:40-48.
15. Yokoyama K, Yang W, Preblick R, Frech-Tamas F. Effects of a step-therapy program for angiotensin receptor blockers on antihypertensive medication utilization patterns and cost of drug therapy. J Manag Care Pharm. 2007;13:235-244.
16. Huang X, Liu Z, Shankar RR, Rajpathak S. Description of anti-diabetic drug utilization pre- and post-formulary restriction of sitagliptin: findings from a national health plan. Curr Med Res Opin. 2015;31:1495-1500.
17. Colombi AM, Yu-Isenberg K, Priest J. The effects of health plan copayments on adherence to oral diabetes medication and health resource utilization. J Occup Environ Med. 2008;50:535-541.
18. Cranor CW, Bunting BA, Christensen DB. The Asheville Project: long-term clinical and economic outcomes of a community pharmacy diabetes care program. J Am Pharm Assoc (Wash). 2003;43:173-184.
19. Mahoney JJ. Reducing patient drug acquisition costs can lower diabetes health claims. Am J Manag Care. 2005;11(suppl):S170-S176.
20. Roblin DW, Platt R, Goodman MJ, et al. Effect of increased cost-sharing on oral hypoglycemic use in five managed care organizations: how much is too much? Med Care. 2005;43:951-959.
21. Starner CI, Fenrick B, Coleman J, et al. Rosiglitazone prior authorization safety policy: a cohort study. J Manag Care Pharm. 2012;18:225-233.
22. Pantalone KM, Wells BJ, Chagin KM, et al. Intensification of diabetes therapy and time until A1C goal attainment among patients with newly diagnosed type 2 diabetes who fail metformin monotherapy within a large integrated health system. Diabetes Care. 2016;39:1527-1534.