According to 2014 statistics from the CDC, diabetes affects 12.6% of the US adult population (CDC 2016), and of those individuals, fewer than half are deemed to be at their treatment goal (ODPHP 2016). Both patient-related and provider-related factors contribute to observed suboptimal outcomes. Patient-related factors include high out-of-pocket treatment costs that contribute to reduced adherence to medication regimens and outpatient monitoring of diabetes, particularly in the setting of high-deductible health plans (Fronstin 2013). Because individuals in the lowest socioeconomic category have the highest prevalence of diabetes, costs borne by patients are a major factor in adherence to diabetes medications and monitoring. Behavioral or cognitive factors, including health literacy and health care consumer engagement, are a particular concern for this subpopulation and may also influence treatment adherence in other subgroups as well.
From the perspective of a health care delivery system, other factors likely influence diabetes treatment outcomes, including timeliness of therapeutic intensification, effectiveness of instruction regarding diabetes self-care, and attention to social determinants of health. High out-of-pocket costs for patients may reduce the frequency of physician visits for routine diabetes care, thereby limiting diabetes monitoring and management capabilities. The resulting suboptimal disease control is associated with an increased risk of diabetes-related complications and their associated costs. Notably, these problems may be most prevalent among low-wage workers, where disease prevalence is highest and control may be poor (Sherman 2016).
Patient-focused (demand-side) initiatives have yielded limited benefit. Some value-based benefit designs that reduce enrollees’ out-of-pocket costs for treatments that improve clinical outcomes, lower cost, or both, have resulted in only single-digit improvements in medication adherence (Look 2015). Many employers have implemented first-dollar coverage of preventive medication lists for individuals in high-deductible health plans. However, these plans typically include only generics, so their relevance to diabetes is limited because branded products account for 47% of market volume (Express Scripts 2016). Condition-management programs, as well as incentives for compliance with evidence-based treatment, also have been utilized in an effort to improve diabetes control (Misra-Hebert 2016). Digital self-care programs have yielded favorable treatment outcomes (Lorig 2016). However, these approaches have not been widely implemented. Employers, managed care entities, and community groups have deployed them in specific circumstances or on a trial basis.
Health care system–based (supply-side) quality improvement programs have also been used to improve diabetes care management. These efforts have included policies and projects aimed at closing HEDIS gaps in care and improving patient engagement, as well as value-based contracting for diabetes medications (Barlas 2016). Yet, despite both patient- and health care system–focused initiatives, national survey data indicate that less than half of people with diabetes have the disease under control (ODPHP 2016), with recent deterioration in the proportion of those with controlled diabetes despite greater medication use (Lipska 2016). There is a clear need for alternative approaches to diabetes management to improve treatment outcomes.
Ongoing product-development efforts have addressed an array of considerations to improve diabetes management, including improved insulin efficacy and novel drug-delivery technologies and monitoring devices (Bancroft 2016). One focus for improved diabetes management involves novel medication-delivery devices that improve medication adherence for insulin and noninsulin drugs.
Insulin-delivery devices have been a focus of industry investment, with the goal of enhancing the ease and effectiveness of insulin delivery. Substantial research-and-development activity has resulted in a number of products receiving recent FDA approval, including a disposable insulin-delivery device called V-Go, made by Valeritas, and an insulin pump called MiniMed 670G, made by Medtronic.
Disposable, 24-hour subcutaneous insulin device V-Go, Valeritas
As a self-administered, disposable, 24-hour subcutaneous insulin device with preprandial bolus dosing capabilities, V-Go has been shown to improve diabetes treatment outcomes, as measured by hemoglobin A1c (HbA1c), in several clinical and real-world trials. For example, in a 27-week retrospective analysis of 204 poorly controlled, predominantly type 2 diabetes patients who were transitioned from multiple daily insulin injections (MDI), use of V-Go for insulin infusion resulted in a statistically significant HbA1c reduction of 1.8 percentage points (P<.001) from a baseline value of 9.6% (Lajara 2015).
Improvement in HbA1c outcomes with V-Go can be attributed in part to better patient adherence, given that the once-daily application of the device and ease of preprandial bolus administration reduces the burden and risk of suboptimal compliance with MDI during the day. The pharmacokinetics of continuous insulin infusion also may contribute to improved HbA1c outcomes.
MiniMed 670G, Medtronic
Comparative cost-effectiveness analysis of V-Go use versus MDI insulin also has yielded favorable outcomes. In a 27-week study of 116 patients with type 2 diabetes suboptimally controlled on basal insulin and oral agents and who received intensification therapy with either V-Go or MDI, use of V-Go resulted in statistically significantly improved HbA1c levels along with reduced daily insulin dosing. Direct pharmacy-related costs were $55 less for V-Go users (Lajara 2016).
The MiniMed 670G device pairs an insulin pump with a continuous glucose monitoring sensor as a closed-loop system. This hybrid system adjusts basal insulin administration based on readings from the continuous glucose monitor every five minutes. If glucose levels fall below a critical threshold, insulin administration is suspended to allow glucose levels to increase prior to resumption of pumping. Similarly, the results of glucose monitoring will prompt increases in basal insulin infusion rates if blood glucose measurements are above a predetermined threshold. Individuals still must self-administer meal-related boluses of insulin.
Critical outcomes data were generated from a three-month, open-label study of device use by individuals previously using an insulin pump, using the run-in period of 14 days when the pump was operational but not in closed-loop mode as the comparator. Results showed improvement in HbA1c from 7.4% to 6.9%, 44% reduction in time spent with glucose <70 mg/dL, and 11% less time with blood glucose >180 mg/dL, though there were six episodes of hyperglycemia with blood glucose >300 mg/dL (Bergenstal 2016). These findings were sufficient for the FDA to approve the device, with product sales expected to begin in spring 2017.
Despite their perceived clinical value, cost-effectiveness of these and other closed-loop systems has been challenged (Riemsma 2016), particularly given the timeline for projected cost savings. Further research is necessary to quantify the relative value of these integrated devices in comparison with the effectiveness of their separate components.
Innovations in noninsulin diabetes medication-delivery devices have also received attention. ITCA 650 from Intarcia Therapeutics, a clinician-implanted drug-delivery device that uses an osmotic pump mechanism to deliver exenatide in steady-state concentrations, is one such offering. Exenatide, a glucagon-like peptide-1 (GLP-1) receptor agonist, is commercially available as twice-daily and once-weekly (extended release formulation) self-injection therapy for type 2 diabetes. Once the ITCA 650 device has been implanted under the patient’s abdominal skin, either a six- or 12-month supply of exenatide may be provided, depending on the amount of medication included in the device. Adherence rates are effectively 100%, given the subcutaneous device placement and drug-delivery method. The pivotal phase 3 clinical trial for ITCA 650 was completed in 2016, but study results have not yet been released. The impact of ITCA 650 on diabetes-related complications, including ocular, renal, neurologic, and cardiovascular, has not yet been evaluated. Similarly, data are not yet available regarding cost-effectiveness because the product is not yet commercially available, although a new drug application for ITCA 650 was submitted to the FDA in November 2016.
Because they have not been on the market very long, these products have had limited evaluation in real-world settings. For the insulin-delivery devices, patient training and compliance with appropriate device use is particularly important. Additionally, HbA1c outcomes depend upon patient compliance with bolus dosing prior to meals.
Cost may be a barrier to their use. The initial cost of the closed-loop insulin administration device and continuous glucose monitoring sensor may appear prohibitive for payers, who tend to have a shorter-term benefits–cost time horizon. Health plans, as well, will benefit from a better understanding of the health benefit cost-savings implications and a projected timeline that may potentially offset acquisition costs of these devices. Paradoxically, the people best suited for the closed-loop devices are likely those who have sufficient clinical knowledge to operate insulin pumps that are currently available. The potential savings in this subpopulation may be less than for patients whose diabetes is poorly controlled and who poorly adhere to treatment.
In comparison, limitations for ITCA 650 appear to be more straightforward, given that exenatide is a widely used medication with an established safety profile. The long-term impact of exenatide administration via ITCA 650, along with implantation-related complications, have yet to be evaluated and reported outside of the context of clinical trials.
Given patient-adherence concerns, as well as the clinical and cost implications of suboptimally managed diabetes, health plans will need to weigh coverage of these products carefully. Numerous questions will come up. Should these products be included in the medical or pharmacy benefit? How will the magnitude of the value proposition be affected by selective product placement in one but not the other benefit offering? What entity is best equipped to manage clinical appropriateness determinations for these offerings? How should value-based benefit design be applied, and who should do it—the medical carrier or pharmacy benefit manager (PBM)?
On the surface, the answers to these questions aren’t clear, particularly for drug-delivery devices that could reasonably fall under either benefit offering. For example, standard insulin pumps typically are covered as durable medical equipment in the medical benefit, and that may continue to be true for closed-loop insulin administration devices. Yet the disposable insulin devices may warrant coverage under the pharmacy benefit, particularly for those PBMs already contracted to sell the offering. Of note, V-Go, because it is disposable and not durable, is a Medicare Part D benefit (as mandated by CMS), with an established pharmacy benefit path for payers. Given this designation, it may make little sense to include it under the medical benefit for non-Medicare patients. The ITCA 650 device could be categorized as either a medical device or a physician-administered medication, potentially amenable to either “brown-bagging” (where patients bring in the medication) or “white-bagging” (where the medication is shipped directly to the doctor). This blurring of the old-fashioned distinction between medical, pharmacy, and durable medical equipment domains can be expected to continue as other novel medication administration devices are brought to market.
Lessons learned from specialty pharmacy management in the medical and pharmacy plans would suggest that the PBM may well be the preferred option. PBMs have administrative capabilities to ensure appropriate drug use, including prior authorization, step therapy, utilization management, formulary management, and site of care management, in addition to case management. PBMs also have experience with minimizing potential “buy-and-bill” physician behaviors, to eliminate potential inappropriate medication use. Additionally, for plan sponsors who carve out separate pharmacy and medical benefits, consolidation of medication-related information into one data set will be important to improve treatment outcomes. PBM volume-based purchasing also may yield lower pricing that may further benefit the plan sponsor. Finally, PBM value-based contracting with product manufacturers may also help to align mutual financial objectives with treatment outcomes.
Benefit-design considerations also may support preferential product placement in the pharmacy benefit. Instead of dealing with additional outpatient copayments incurred for a clinician visit accompanying medication dispensing, individuals incur only the medication-related co-insurance or copayment in accordance with their PBM benefit design. This is clearly the case for V-Go, but it may not be appropriate for ITCA 650, which requires clinician placement of the infusion device. Value-based benefits design considerations may also be administratively easier to administer in the context of an existing tiered pharmacy benefit. With the current out-of-pocket cost pressures prevalent among health plan enrollees, the PBM option may therefore represent the most reasonable option.
With medication adherence representing a substantial—and continuing—barrier to achieving treatment goals, newer approaches to medication delivery that increase adherence merit serious consideration. For managed care, new diabetes treatment modalities that improve adherence and clinical outcomes while lowering treatment costs should warrant thoughtful consideration for inclusion in health plans. The likely reductions in risk of diabetes-related complications and associated costs may further expand the value proposition.
Managing these devices in the pharmacy benefit may well help to optimize stakeholder cost savings by consolidating volume-based purchasing and capitalizing on PBM strategies for improving adherence. Grouping medications and these new products in the pharmacy benefit will enable more consistent management and make direct product comparisons possible. Additionally, medication coding is more specific in the pharmacy benefit, so health plans are better able to analyze treatment outcomes. Inclusion in the pharmacy benefit may provide greater assurance for plan sponsors that reasonable drug cost containment and utilization management capabilities can be appropriately implemented, such as prior authorization criteria.
Management through the pharmacy benefit placement could have some disadvantages. It may limit the ability to integrate information with medical claims and potentially reduce the effectiveness of health plan–provided diabetes care management. In that scenario, two-way data sharing between medical and pharmacy plans should minimize that concern.
New diabetes medication-delivery devices can help to improve treatment outcomes while lowering near-term, disease-specific pharmacy costs and long-term overall health care costs. Plan sponsors will benefit from consideration of these and other novel drug-delivery devices as part of an expanded approach to effective diabetes management. Inclusion of these offerings as part of the pharmacy benefit may improve adherence because PBMs traditionally focus on medication adherence and may also lower acquisition costs for both patients and plan sponsors. Positioning of these products in the pharmacy benefit may also leverage existing PBM contracting and benefit design capabilities to benefit manufacturers, purchasers, and patients. Novel approaches deserve consideration because of their potential for bringing about drug-cost savings in the near term, cost savings from better management of diabetes in the long term, or both.
Bruce W. Sherman, MD
3175 Belvoir Blvd.
Shaker Heights, OH 44122
Disclosures: The author reports receiving an unrestricted educational grant from Valeritas in 2016 for preparation of this manuscript and an honorarium from Intarcia in 2015 for participation in an employer advisory board. The companies are manufacturers of two devices discussed in this article.