Medicare Part D Insulin Tiering Change: Impact on Health Outcomes

, , , ,
The American Journal of Pharmacy Benefits, September/October 2016, Volume 8, Issue 5

This study explores the impact of a Medicare Part D formulary tier change, specifically with respect to branded insulin products, on insulin adherence, hemoglobin A1C, and diabetes-related events in elderly diabetic beneficiaries strategies for selection of appropriate and cost-effective therapy will also be discussed.

ABSTRACT

Objective: The study aimed to determine the impact of a Medicare Part D formulary tier change, specifically with respect to branded insulin products, on insulin adherence, hemoglobin A1C (A1C), and diabetes-related events in elderly diabetic beneficiaries.

Study Design: Retrospective database analysis based on pharmacy claims from a Medicare Advantage Prescription Drug (MAPD) plan in Texas.

Methods: Elderly (≥65 years) beneficiaries with diabetes who were continuously enrolled from January 1, 2013, to December 31, 2014, with at least 2 prescription fills of any insulin product each year, were included. Outcomes included changes in the following, before and after the tier change (January 1, 2014): adherence to insulin medications (proportion of days covered for insulin products [PDC-i]); A1C; and diabetes-related events (hyperglycemia [790.21, 790.22, and 790.29], hyperosmolar hyperglycemic state [250.2-0.23], or diabetic ketoacidosis [249.1, 249.11, and 250.1-0.13]). For change in A1C, a minimum of 2 records of A1C measurement, 1 each in 2013 and after May 2014, was required.

Results: The sample consisted of 3285 beneficiaries, with the majority female (57.17%), aged 65-74 years (61.10%), and receiving the low-income subsidy (51.29%). Decline in PDC-i (mean = 0.48; SD = 0.25; P <.001) was observed among 2333 beneficiaries (71%). Only 1516 beneficiaries met the criteria for calculating change in A1C, out of which 755 (49.80%) had an increase in A1C (mean = 0.92; SD = 0.98; P <.001) after the tier change. Diabetes-related events were observed in 19 beneficiaries before and 32 beneficiaries after the tier change.

Conclusions: Change in tier of insulin medications was associated with a statistically significant decline in insulin adherence and increase in A1C within the MAPD population.

Am J Pharm Benefits. 2016;8(5):-e0

Implemented in 2006, Medicare Part D was established to provide prescription drug coverage for Medicare beneficiaries. Medicare beneficiaries who receive Medicare Part D have had improved adherence to medications, increased rates of medication fills, decreased out-of-pocket expenses, decreased medical spending, and fewer avoidable hospitalizations.1-5 However, the design of the Part D benefit introduced a unique feature known as the coverage gap, or “donut hole.”

In 2013, Medicare beneficiaries paid an initial $325 deductible (or a deductible set by the health plan), followed by 25% of covered costs (or a co-payment set by the health plan) until the initial coverage limit of $2970 was met, as represented in Figure 1. The limit is reached by totaling the amounts paid by the beneficiary and the payer.

After the initial coverage limit was met, the beneficiary entered the coverage gap, during which they paid 47.5% of the cost on brand-name drugs and 79% of the cost on generic drugs until their total out-of-pocket spending reached a catastrophic limit of $4750. However, the amount of money that the payer pays does not contribute toward reaching the catastrophic limit. The catastrophic coverage period is when the beneficiary is paying the greater of 5% or $2.65 for generic/preferred drugs and $6.60 for other drugs.6

Guidance provided by CMS plays a significant role in Medicare Part D sponsor formulary decisions. Many Medicare Part D sponsors place branded medications that do not have generic therapeutic equivalences—ie, insulin products—on a generic tier to decrease the cost burden of these medications for their Medicare beneficiaries. Per CMS, brand medications offered on a generic tier leave the health plan open to tiering exceptions for any brand medication in a higher tier.

CMS defines a tiering exception as a drug plan’s decision to charge a lower amount for a drug that is on its nonpreferred drug tier, based on specific criteria.7,8 CMS further states, within Chapter 18 of its Prescription Drug Benefit Manual, that if a beneficiary requests that a branded medication be provided at a generic cost, these medications do not meet the tiering exception criteria unless the health plan covers other brands at the generic tier and/or they meet specific criteria.9

This means that all Medicare Part D sponsors who were placing branded medications on a generic tier to provide coverage through the coverage gap for their Medicare beneficiaries were susceptible to having any branded medication that has alternatives on generic tiers to be potentially approved for generic cost sharing. On January 1, 2014, many Medicare Part D sponsors moved brand name products from a generic drug tier to a brand drug tier.

As a result of this change, co-payments increased significantly for diabetic beneficiaries prior to entering the coverage gap. In addition, insulin products became more susceptible to the coverage gap and Medicare beneficiaries who required insulin therapy had to pay 47.5% out-of-pocket for these maintenance medications. Our study is the first known study to assess the effect of this change on patient outcomes.

In 2012, diabetes affected 25.9% of US residents 65 years or older, or 11.2 million residents. As a result, the estimated total cost of diagnosed diabetes increased by 41%: from $174 billion in 2007 to $245 billion in 2012.10,11 In addition, the majority (62.4%) of the cost for diabetes care in the United States is provided by government insurance (including Medicare and Medicaid).11

Further, between 2010 and 2012, 6 million adults 18 years or older with diagnosed diabetes reported using insulin and pills, or insulin only, for the treatment of diabetes.12 Currently, there are no available generic therapeutic alternatives for insulin products in the United States. Although the cost of insulin can be expensive, anti-diabetic agents and diabetes supplies make up only 12% of medical expenditures for diagnosed diabetes.11

The 2 largest components that contribute to diabetes medical expenditures are prescription medications to treat complications of diabetes (18% of total medical cost) and hospital in-patient care (43%).11 The National Diabetes Statistic Report of 2014 found that “about 175,000 emergency room visits had a hyperglycemic crisis (eg, diabetic ketoacidosis and hyperglycemic hyperosmolar state) as the first-listed diagnosis.”12

In addition, diabetes was listed 44% of the time as the primary cause of all new cases of kidney failure in the United States in 2011.12 The American Diabetes Association estimated the cost of diagnosed diabetes, in 2012, to be $245 billion, including $176 billion in direct medical costs and $69 billion in reduced productivity.11 As for hospital in-patient care, in 2009 in the United States, the CDC found that the average length of stay for a patient with diabetes as any-listed diagnosis was 4.8 days.13 These reasons, and more, are why controlling the cost of anti-diabetes medications, especially insulin therapy, is important to help control the cost of healthcare overall.

For Medicare beneficiaries who live with diabetes, poor adherence to medications can lead to a worsening of glycemic control, increased comorbidities, and diabetes-related hospitalizations. Zhang et al reported that entering the coverage gap reduced medication use (mainly brand name drugs) among Medicare beneficiaries who had heart failure and/or diabetes.14

In addition, Gu et al found that the coverage gap significantly decreased adherence to diabetes medications by 39% and 30% for those beneficiaries in the coverage gap who had no coverage or generic coverage only, respectively.15 These reductions in medication use and adherence can negatively impact medical costs (ie, disease progression can lead to comorbidities and more hospitalizations).

In this study, we examined the effects of changing the tier status of insulin on adherence to insulin medications (aspart, glulisine, lispro, glargine, detemir, and regular insulin), hemoglobin A1C (A1C), and diabetes-related events. In the light of existing literature, we predicted that Medicare beneficiaries who have higher co-payments but no coverage through the coverage gap would have decreased adherence to insulin therapy, an increase in A1C, and an increase in diabetes-related events compared with Medicare beneficiaries who had lower co-payments and coverage through the coverage gap. Evidence of any significant negative impact on health outcomes due to the policy changes may help in future investigations for creating better access to essential insulin medications.

METHODS

Data

The study was a retrospective database analysis conducted using data obtained from the pharmacy and medical claims of a Medicare Advantage Prescription Drug (MAPD) plan in Texas. The study population consisted of diabetes beneficiaries enrolled in an MAPD 1 year before (January 1, 2013-December 31, 2013) and 1 year after (January 1, 2014-December 31, 2014) the insulin tier status change, which was implemented on January 1, 2014.

Inclusion criteria were as follows: (a) aged ≥65; (b) continuously enrolled in an MAPD plan for 2 years, 1 year before and 1 year after the tiering change, in the southeast Texas region (South Texas and greater Houston areas); (c) diabetes diagnosis (International Classification of Diseases, Ninth Revision, Clinical Modification [ICD-9-CM] codes 249.XX and 250.XX); and (d) at least 2 claims, in each year, for 1 or more of the following insulin medications: aspart, glulisine, lispro, glargine, detemir, or regular insulin (identified using Generic Product Identifiers, 14-digit codes that each contain 7 pairs of digits).

The calculation of change in A1C required additional criteria of having 2 A1C measurements, 1 before and 1 after the tiering change. “Before the tier change measurement” was defined as an A1C measurement any time in 2013 and “after the tier change measurement” was defined as an A1C measurement after May 2014, to account for the effect of the tier change. Exclusion criteria were: (a) beneficiaries receiving prescription drugs from a Veterans Affairs health provider, (b) hospice members, (c) those with cancer diagnoses (ICD-9-CM codes 140.XX-239.XX), (d) those with a change of low-income subsidy (LIS) status during the study period, and (e) those having Medicare without prescription coverage within the same health plan.

Outcome Variables

Insulin can be used, and is sometimes the sole therapy, to maintain glycemic control and slow the progression of diabetes. Poor adherence to insulin medications results in insufficient glycemic control, micro- and macrovascular complications leading to higher probabilities of hospitalization, higher medical cost, and increased chance of mortality.1-3,11,12 Hence, a change in medication adherence was evaluated only for insulin medications.

Adherence to insulin medication was calculated as the proportion of days covered for insulin products (PDC-i). Very little literature exists about the use of PDC as a measure of adherence for injectable products, due to its complex dosing based on a sliding scale: the quantity of insulin dispensed to a patient is calculated according to the days of supply.

Because a patient using the maximum dose daily will use the dispensed quantity within the calculated days of supply, any deviations in the refill time will be considered nonadherence, just as in the case of oral medications. Therefore, PDC-i was defined as the number of days the insulin medication was possessed by the beneficiary divided by the total number of days between the first insulin fill and the end of study period.

Change in PDC-i was calculated, for each beneficiary, as the difference between PDC-i values for the year before and the year after the tier change. PDC-i has been used in many studies as a standard measure for medication adherence.15,16 We used a PDC-i cutoff value of 0.8, with beneficiaries having PDC-i ≥0.8 considered as being adherent to insulin medications.17

A1C is a clinical test widely used to indicate the mean blood glucose concentration over the lifespan of a red blood cell. The measure correlates best with mean blood glucose over the previous 8 to 12 weeks.18 A1C was calculated as at least 1 value during the indicated study period. If a beneficiary had more than 1 measurement for either of the 2 time periods, an average value of the measurements was taken and used for further calculations. Change in A1C was calculated by a difference between the 2 measurements of A1C for each beneficiary, 1 before and 1 after the tiering change.

Diabetes-related events were defined as having a diagnosis, reported within a healthcare setting, for any of the following events (designated by ICD-9-CM codes): hyperglycemia (790.21, 790.22, and 790.29), hyperosmolar hyperglycemic state (250.2-0.23), or diabetic ketoacidosis (249.1, 249.11, and 250.1-0.13). These complications were included because they can be directly correlated with nonadherence to insulin therapies. Percent change in the average number of diabetes-related events after the tier status change was calculated. Any increase in the number of events by more than 5% was classified as significant.

Participants’ age, gender, and LIS status were also obtained. The subjects were divided into 2 groups: those aged 65-74 years and those 75 years or more. Beneficiaries were divided into those who were non-LIS and received no “extra help” outside of their purchased insurance through the MAPD plan and those who had full LIS status, receiving “extra help” from a government program.

Statistical Analyses

Descriptive analyses were conducted by calculating the mean for continuous variables and frequencies for categorical variables. To evaluate changes in adherence to insulin medications, in A1C, and in the number of diabetes-related events before and after the tier change, t tests were conducted. Logistic regression analyses were conducted to test the associations between: (a) change in adherence to insulin medications and age, gender, and LIS status; and (b) change in A1C and adherence, age, gender, and LIS status. All analyses were conducted using SAS version 9.3. The study was approved by the University of Houston’s Institutional Review Board.

RESULTS

The final sample consisted of beneficiaries continuously enrolled during the 2 time periods: before the tier-status change (2013) and after the tier-status change (2014). After applying inclusion and exclusion criteria, 3285 beneficiaries were identified and included in the final analyses. Figure 2 represents the flow diagram for selection of study participants.

A total of 5006 beneficiaries were initially selected who met the inclusion criteria and exclusion criteria of being enrolled throughout the study period, were located within the study region, had a diagnosis of diabetes and 2 prescription claims for insulin within each year; were not hospice members or receiving prescription drugs from Veterans Affairs, and did not have a change in full LIS status during the study period. From the 5006 beneficiaries, 1130 and 591 were then excluded due to age (being <65 years) or a cancer diagnosis, respectively.

The final cohort of 3285 beneficiaries was included for analyses. Table 1 displays the sample baseline characteristics for beneficiaries. The mean age of the sample was 73.73 years (± 6.24), with a majority being female (57.17%) and within the younger group (aged 65-74 years).

Table 2 represents the PDC-i, A1C, and diabetes-related event values for before and after the tier change. As indicated in Table 2, 1784 (54.31%) of the beneficiaries had PDC-i ≥0.8 before the tiering change compared with only 761 (23.17%) after the tiering change. Of the total of 3285 beneficiaries, 2333 (71.02%) had a decline in PDC-i after the tier change, with an average decline of 0.48 (P <.001; 95% CI, 0.47-0.49).

Logistic regression analyses (Table 3) indicated that beneficiaries with LIS were 17 times less likely to show a decline in PDC-i compared with those without LIS (odds ratio [OR], 0.828; 95% CI, 0.710-0.965). Beneficiaries 75 years or older were 17 times less likely to show a decline in PDC-i compared with those aged 65-75 years (OR, 0.829; 95% CI, 0.710-0.968).

Of the total 3285 beneficiaries, 1516 met the criteria for inclusion in the analyses for change in A1C (Table 2). The average values for A1C were 8.19% (standard deviation [SD] = 1.46%) before the tier change and 8.17% (SD = 1.54%) after the tier change. Although the overall average A1C remained constant, 755 (49.80%) beneficiaries had a statistically significant increase in A1C after the tier change, with an average increase of 0.91% (95% CI, 0.85%-0.99%; P <.001) (Figure 3). This change in A1C, however, was not associated with gender, age, LIS status, or change in adherence (Table 3).

A total of 22 beneficiaries had diabetes-related events before the tier change: 21 beneficiaries had 1 event and 1 beneficiary had 2 events. After the tier change, 36 beneficiaries had diabetes-related events: 32 beneficiaries had 1 event, 3 had 2 events, and 1 had 3 events. It was observed that among these beneficiaries, 35 did not have any such event before the tier change, indicating an almost 100% increase in the number of diabetes-related events for those patients after the tier change. A t test indicated that the increase in the number of diabetes-related events was significant (P <.001).

DISCUSSION

Medicare beneficiaries’ insulin adherence decreased after the tier status change, when all branded insulin products were switched from a generic drug tier with gap coverage and lower co-payments to a brand drug tier with no gap coverage and higher co-payments. The study followed the same beneficiaries before and after the insulin tier change. After the tier change, these beneficiaries had an increase in initial co-payment and were susceptible to the coverage gap.

When these beneficiaries entered the coverage gap, their co-payments substantially increased, to paying 47.5% of the cost for insulin. This noticeably large increase in cost for these insulin therapies creates barriers for beneficiaries in obtaining their insulin therapy, which can transpire to nonadherence and a progression of their diabetes. These findings are similar to those of other studies that noted a drop in adherence for beneficiaries who had an increase in the co-payment for their diabetes medications.15,16,19-21

However, unlike most of the previous studies, which involved oral medications, the current study assessed adherence and associated outcomes for injectable insulin products. It may be argued that measuring adherence for injectable insulin is difficult using a claims database due to the variable sliding scale regimen. However, in practice, a quantity appropriate to cover patients’ needs for the amount of insulin as per the dosing regimen is calculated and dispensed by the pharmacist. Thus, when dosed on a schedule, insulin has the same adherence measurement parameters as oral medications (ie, number of days’ supply [numerator] over days covered [denominator]).

The study results indicated that beneficiaries with full LIS status were less likely to be nonadherent. LIS is a government program that assists Medicare beneficiaries who have limited resources and income; they qualify to get “extra help” to pay for the costs—monthly premiums, annual deductibles, and prescription co-payments—related to a Medicare prescription drug plan.22

Eligibility for LIS status includes: being dual-eligible (beneficiaries who have both Medicare and Medicaid), beneficiaries who receive Social Security income (even in states in which they do not qualify for Medicaid), beneficiaries eligible for 1 of the Medicare Saving Programs, or beneficiaries who meet the financial eligibility criteria to become eligible.22

Although beneficiaries with LIS status have the additional challenge of having a low income, they did not see changes in cost for their prescription medications ($2.65 for generic and $6.60 for brand name medication),23 and they continued to have coverage through the gap after the insulin tier change. The correlation we found between having LIS status and becoming less likely to decline in adherence to insulin therapies further shows that beneficiaries who had an increase in their co-payment and no coverage through the coverage gap were more likely to be nonadherent to insulin therapies.

It is important to note: for beneficiaries whose A1C increased, that increase was significant, at an average 0.9%. These beneficiaries’ average A1C after the insulin tier change was 8.7%, which is trending toward poorer compliance with the CMS Part C A1C measure goal of <9.0% and is thus a significant finding. The use of this clinical test, as a measure for the control of diabetes, has been proposed by the National Quality Measures Clearinghouse (NQMC) and implemented as a Part C measure.24 It was included as a Part C measure because the NQMC noted that diabetes is an increasingly common, complex, serious disease that not only imposes economic burdens, but frequently leads to human suffering from the disease’s complications.

Moreover, the burden of diabetes and its complications largely affect minorities and the elderly, and it will likely increase as minority populations grow and as the US population ages.11 As of 2015, the National Committee for Quality Assurance’s Healthcare Effectiveness Data and Information Set measures have posited that an A1C >9.0% is poorly controlled. A Medicare Part D sponsor would find an increasing trend of A1C concerning because poor compliance to this measure can affect its Star ratings and reimbursements. Furthermore, beneficiaries with uncontrolled diabetes are at an increased risk of macro- and microvascular complications, which lead to increased cost due to worsening comorbidities and increased risk of mortality.

The analysis also revealed an increase in the number of diabetes-related events when comparing the months before and after the change in insulin tier. We anticipated a larger increase in diabetes-related events; however, we also hypothesized some potential reasons for lower numbers.

These included: beneficiaries could be using noninsulin anti-diabetes medications in addition to insulin therapy or beneficiaries could have failed to report diabetes-related events, possibly due to the cost of a hospitalization. Due to small sample size, no claims could be made regarding an association between an increase in the number of events and the lessened adherence to insulin.

Limitations

There were several limitations to this study. Insulin adherence was determined using pharmacy claims data to obtain the days’ supply for our PDC-i calculations. Some insulin regimens are not represented correctly by days’ supplies, such as sliding scale regimens. We were unable to obtain individual regimens, and thus used the days’ supply provided within the pharmacy claims data to determine insulin adherence. In addition, the use of PDC-i may overestimate the actual adherence. We did not note if the beneficiaries were using multiple insulins simultaneously for therapy. If a beneficiary is prescribed to take multiple insulins but only takes 1, the PDC-i calculation will note the beneficiary as adherent.

Also, beneficiaries could have received samples from providers or through patient assistance programs; pharmacy claims data do not account for those doses. We also did not collect data about concurrent use of noninsulin anti-diabetes medications. Many beneficiaries who live with diabetes use both insulin therapies and noninsulin anti-diabetes medications for glycemic control. Additionally, we did not randomize the beneficiaries based on their use of noninsulin anti-diabetes medications. The concurrent use of these medications can affect A1C levels and diabetes-related events.

Next, the research is based on a MAPD plan covering the south Texas and greater Houston areas. There can be variances between the population profiles of these regions and that of the national population. Lastly, the study period was 1 year prior and 1 year after the tiering change of insulin products. A longer study period would allow us to ascertain a better understanding of the beneficiaries’ behavior in reaction to an increased cost of their insulin therapies.

CONCLUSIONS

A change in the tier of insulin medications was associated with a significant decline in medication adherence and an increase in A1C within the MAPD population. The number of diabetes-related events also increased from 2013 to 2014. This analysis exposes a relative disparity of accessibility to insulin products for diabetics, which was never described previously. The proven clinical benefits afforded by the use of insulin medications, and the negative outcomes revealed within this short study, both point toward further investigation of the effects of a change in the insulin tiering. MAPD plans should seek to create ways to make insulin products consistently accessible to MAPD beneficiaries.

Author Affiliations: Cigna-HealthSpring (ATN, OS, TE), Houston, TX; Department of Pharmaceutical Health Outcomes and Policy, College of Pharmacy, University of Houston (RVS, SSS), TX.

Source of Funding: None.

Author Disclosures: The authors report no relationship or financial interest with any entity that would pose a conflict of interest with the subject matter of this article.

Authorship Information: Concept and design (AN, OS, TE, SS); acquisition of data (AN, OS); analysis and interpretation of data (AN, RS, SS); drafting of the manuscript (AN, OS); critical revision of the manuscript for important intellectual content (AN, RS, TE, SS); statistical analysis (); provision of study materials or patients (AN); administrative, technical, or logistic support (AN, OS); supervision (OS, TE, SS)

Address correspondence to: Sujit S Sansgiry, PhD, Professor, Department of Pharmaceutical Health Outcomes and Policy, College of Pharmacy, University of Houston, Texas Medical Center, 1441 Moursund St, Houston TX 77030. E-mail: sansgiry@central.uh.edu

References

  • Yin W, Basu A, Zhang JX, Rabbani A, Meltzer DO, Alexander GC. The effect of the Medicare Part D prescription benefit on drug utilization and expenditures. Ann Intern Med. 2008;148(3):169-177.
  • Zhang Y, Lave JR, Donohue JM, Fischer MA, Chernew ME, Newhouse JP. The impact of Medicare Part D on medication adherence among older adults enrolled in Medicare-Advantage products. Med Care. 2010;48(5):409-417. doi: 10.1097/MLR.0b013e3181d68978.
  • Zhang Y, Donohue JM, Lave JR, O’Donnell G, Newhouse JP. The effect of Medicare Part D on drug and medical spending. N Engl J Med. 2009;361(1):52-61. doi: 10.1056/NEJMsa0807998.
  • McWilliams JM, Zaslavsky AM, Huskamp HA. Implementation of Medicare Part D and nondrug medical spending for elderly adults with limited prior drug coverage. JAMA. 2011;306(4):402-409. doi: 10.1001/jama.2011.1026.
  • Afendulis CC, He Y, Zaslavsky AM, Chernew ME. The impact of Medicare Part D on hospitalization rates. Health Serv Res. 2011;46(4):1022-1038. doi: 10.1111/j.1475-6773.2011.01244.x.
  • Closing the coverage gap: Medicare prescription drugs are becoming more affordable. CMS website. https://www.medicare.gov/Pubs/pdf/11493.pdf. Revised January 2015. Accessed May 26, 2015.
  • Medicare prescription drug appeals and grievances: exceptions. CMS website. https://www.cms.gov/Medicare/Appeals-and-Grievances/MedPrescriptDrugApplGriev/Exceptions.html. Updated November 22, 2013. Accessed January 7, 2016.
  • Glossary — E. Medicare website. http://www.medicare.gov/Glossary/e.html. Accessed May 26, 2015.
  • Prescription Drug Benefit Manual. Chapter 18 — Part D Enrollee Grievances, Coverage Determinations, and Appeals. CMS. 2006;18:1-101.
  • Statistics about diabetes. American Diabetes Association website. http://www.diabetes.org/diabetes-basics/statistics/?referrer=https://www.google.com/. Updated December 9, 2015. Accessed January 7, 2016.
  • American Diabetes Association. Economic costs of diabetes in the US in 2012. Diabetes Care. 2013;36(4):1033-1046. doi: 10.2337/dc12-2625.
  • National diabetes fact sheet, 2011. CDC website. http://www.cdc.gov/diabetes/pubs/pdf/ndfs_2011.pdf. Published 2011. Accessed October 28, 2014.
  • Diabetes public health resource: average length of stay (LOS) in days of hospital discharges with diabetes as any-listed diagnosis, United States, 1988-2009. CDC website. http://www.cdc.gov/diabetes/statistics/dmany/fig2.htm. Updated October 1, 2014. Accessed May 31, 2015.
  • Zhang Y, Baik SH, Lave JR. Effects of Medicare Part D coverage gap on medication adherence. Am J Manag Care. 2013;19(6):e214-e224.
  • Gu Q, Zeng F, Patel BV, Tripoli LC. Part D coverage gap and adherence to diabetes medications. Am J Manag Care. 2010;16(12):911-918.
  • Ho PM, Rumsfeld JS, Masoudi FA, et al. Effect of medication nonadherence on hospitalization and mortality among patients with diabetes mellitus. Arch Intern Med. 2006;166(17):1836-1841.
  • PQA measures used by CMS in the Star Ratings. Pharmacy Quality Alliance website. http://pqaalliance.org/measures/cms.asp. Published 2013. Accessed June 9, 2015.
  • Diabetes mellitus: average HbA1c value for diabetic patients in the clinical information system. Agency for Healthcare Research and Quality website. http://www.qualitymeasures.ahrq.gov/content.aspx?id=27617. Updated 2014. Accessed May 26, 2015.
  • Grant R, Adams AS, Trinacty CM, et al. Relationship between patient medication adherence and subsequent clinical inertia in type 2 diabetes glycemic management. Diabetes Care. 2007;30(4):807-812.
  • Pladevall M, Williams LK, Potts LA, Divine G, Xi H, Lafata JE. Clinical outcomes and adherence to medications measured by claims data in patients with diabetes. Diabetes Care. 2004;27(12):2800-2805.
  • Fung V, Mangione CM, Huang J, et al. Falling into the coverage gap: Part D drug costs and adherence for Medicare Advantage prescription drug plan beneficiaries with diabetes. Health Serv Res. 2010;45(2):355-375. doi: 10.1111/j.1475-6773.2009.01071.x.
  • Part D Low-Income Subsidy (LIS)/extra help: eligibility and coverage. National Council On Aging Center for Benefits Access website. http://www.ncoa.org/assets/files/pdf/center-for-benefits/part-d-lis-eligibility-and-coverage.pdf. Updated 2015. Accessed April 10, 2015.
  • Low Income Subsidy information for Part D of Medicare. New Hampshire’s Work Incentives Resource Center website. http://nhwirc.org/Files/LIS%20resource%20limit%20and%20cost-share%20for%20-%202013.pdf. Updated 2013. Accessed April 24, 2015.
  • Medicare 2016 Part C & D Star Rating technical notes. CMS website. https://www.cms.gov/Medicare/Prescription-Drug-Coverage/PrescriptionDrugCovGenIn/Downloads/2016-Technical-Notes-Preview-1-v2015_08_05.pdf. Updated August 5, 2015. Accessed January 7, 2016.