Chapter Five

Presciption Drug Coverage in CKD Patients Medicare Part

As of September 2008, 26 million Medicare-enrolled elderly and disabled people, as well as individuals with ESRD, were enrolled in a Medicare Part D prescription drug plan (PDP). Before 2006, these patients obtained drug coverage through various insurance plans, state Medicaid programs, or pharmaceutical-assistance programs, received samples from physicians, or paid out-of-pocket. After 2006, however, the majority obtained Part D coverage. Fifty-six to 60 percent of general Medicare patients and patients with CKD, diabetes, or cardiovascular disease were enrolled in Part D in 2008, as were 67 percent of patients with ESRD.

The retiree drug subsidy, designed to encourage employers to supply prescription coverage to Medicare-covered retirees that is at least as valuable as the Medicare Part D standard plan, provides employers with a tax-free rebate for 28 percent of retirees' drug costs. Other patients are enrolled in employer group health plans or government/military plans ("creditable coverage"), which provide coverage that is equivalent to or better than Part D.

The proportion of patients with other creditable coverage is similar among CKD and Medicare patients, at about 12 percent, but a higher proportion of CKD patients have retiree drug subsidy coverage, at 21 compared to 15 percent. Nine percent of CKD patients have no known source of drug coverage — a level lower than in the Medicare population, and similar to that of patients with diabetes or cardiovascular disease.

Prior to the start of the Medicare Part D program in 2006, patients dually-enrolled in Medicare and Medicaid received prescription benefits under state Medicaid programs. The Part D program, however, offers a substantial low-income subsidy (LIS) benefit to enrollees with limited assets and income, including those who are dually-enrolled. The LIS provides full or partial waivers for many out-of-pocket cost-sharing requirements, including premiums, deductibles, and copayments, and provides full or partial coverage during the coverage gap ("donut hole"). Nearly three in four dialysis patients enrolled in Part D have LIS, compared to 51 and 38 percent of their counterparts with CKD and in the general Medicare population. In general, CKD patients thus pay proportionally lower out-of-pocket costs than Medicare patients for their Part D prescriptions.

Part D does not cover every medication prescribed to Medicare enrollees. Several drug categories — including over-the-counter medications, barbiturates, benzodiazepines, anorexia and weight loss or gain medications, prescription vitamins (except for prenatal vitamins), and cough and cold medications — are excluded from the Part D program by law. This means that some drugs commonly used in CKD patients (oral iron, ergocalciferol, cholecalciferol) are not currently covered through Medicare Part D. Oral calcitriol, doxercalciferol, and paricalcitol, however, are not considered prescription vitamins, and are thus covered.

The Medicare Part D program works in concert with Medicare Part B, which covers medications administered in physician offices (e.g. erythropoiesis stimulating agents in CKD patients). Access to certain medications is dependent on whether the drug is included on the PDP formulary. A PDP could, for example, decide to include calcitriol and doxercalciferol on its formulary, but not paricalcitol.

Part D benefits can be managed through a stand-alone PDP or through a Medicare Advantage (MA) plan, which provides medical as well as prescription benefits. CKD patients can chose to enroll in an MA plan; ESRD patients, in contrast, are precluded from entering an MA plan if they are not already enrolled in one when they reach ESRD. Most data presented in this chapter encompass both types of plans.

Medicare-enrolled CKD patients obtain outpatient medication benefits through Part B, Part D, retiree drug subsidy plans, or other creditable coverage, including employer group health plans, Veterans Administration benefits, Medicaid wrap-around programs, and state kidney programs. Some also pay out-of-pocket for plan expenses and copayments, over-the-counter medications, and low-cost generic agents at retailers.

Figure 5.1 Sources of prescription drug coverage (see page 126 for analytical methods. Point prevalent Medicare enrollees alive on January 1, 2008.)

Medicare Part D Enrollment Patterns in Patients with CKD 

Figure 5.2 Sources of prescription drug coverage in Medicare enrollees, by age, 2008 (see page 126 for analytical methods. Point prevalent Medicare enrollees alive on January 1, 2008.)
Figure 5.3 Sources of prescription drug coverage in Medicare enrollees, by race, 2008 (see page 126 for analytical methods. Point prevalent Medicare enrollees alive on January 1, 2008.)
Figure 5.4 Sources of prescription drug coverage in Medicare enrollees, by CKD stage, 2008 (see page 126 for analytical methods. Point prevalent Medicare enrollees alive on January 1, 2008.)

Among both general Medicare beneficiaries and those with CKD, the percentage enrolled in Part D declines with age. In the CKD population, however, 91 and 75 percent of those age 20–44 and 45–64, respectively, are enrolled, compared to 81 and 65 percent of their general Medicare counterparts.

Nearly 73 percent of general Medicare patients age 20–44 receive the low income subsidy (LIS). It is important to note that most patients in these younger two age groups are disabled. In the two older age groups, similar proportions of general Medicare and CKD patients age 65 and older are enrolled in Part D, at 55–58 percent. The proportion of patients with LIS declines with age in both populations, but CKD patients in each age category are more likely to receive this subsidy.

Patterns of coverage by race are similar in the general Medicare and CKD populations, with both Part D enrollment overall and Part D coverage with LIS highest in Asian and African American patients, and lowest in whites. LIS coverage is higher across races for CKD patients than among their general Medicare counterparts.

There is less variation in Part D enrollment by CKD stage, with 54 and 60 percent of CKD Stage 3 and Stage 4–5 patients enrolled in Part D, respectively, compared to 59 percent of general Medicare patients. In the CKD population, LIS is least common among those with Stage 3 CKD, at 24 percent.

Figure 5.5 Medicare Part D enrollees, by low income subsidy (LIS) status & population, 2008 (see page 126 for analytical methods. Point prevalent Medicare enrollees alive on January 1, 2008.)

Fifty-one percent of Medicare patients with part D coverage had LIS benefits in 2008, compared to 73 percent of dialysis patients and 64 percent of those with a kidney transplant. CKD patients are thus more likely to experience the coverage gap and to have higher premiums, deductibles, and drug copayments, on average, than dialysis and transplant patients.

Table 5.a Medicare Part D enrollees with low income subsidy (LIS), by CKD stage & demographic characteristics, 2008 (percent) (see page 126 for analytical methods. Point prevalent Medicare enrollees alive on January 1, 2008.)

The proportion of patients with LIS varies more by age and race than by CKD stage. In each category, patients with known CKD are more likely to have LIS than their general Medicare counterparts. Among CKD patients, women are far more likely to have LIS than men; in the general Medicare population, in contrast, proportions are similar by gender. In both the general Medicare and CKD populations, Asians are the most likely by race to have LIS, and whites the least.

Figure 5.6 Geographic variations in the percent of Medicare Part D enrollees with low income subsidy (LIS), 2008 (see page 126 for analytical methods. Point prevalent Medicare enrollees alive on January 1, 2008.)

There is substantial state-to-state variability in the percentage of Medicare Part D enrollees with LIS, and the patterns seen in CKD and ESRD patients differ from those of the general Medicare population. In each state there is a monotonic increase in the percentage of patients receiving LIS as populations change from general Medicare to CKD and then to ESRD patients. In Delaware, Minnesota, and Utah, the difference between general Medicare and CKD patients is less than 4 percentage points; in California, in contrast, it is more than 25.

Medicare Part D coverage plans 

Figure 5.7 Medicare Part D benefit parameters, 2006–2008 (see page 126 for analytical methods. http://www.q1medicare.com/PartD-The-2008-Medicare-Part-D-Outlook.php.)

CMS provides prescription drug plans (PDPs) with guidance on structuring a ''standard'' Part D PDP. The upper portion of Figure 5.7 shows the standard benefit design for PDPs in 2006, 2007 and 2008. In 2008, for example, beneficiaries shared costs with the PDP (as coinsurance or copayments) until the combined total reached $2,510 during the initial coverage period. After reaching this level, beneficiaries went into the coverage gap or "donut hole," where they paid 100 percent of costs. Since 2010, the government has been providing those reaching the coverage gap with more assistance each year. In 2008, beneficiaries who obtained a yearly out-of pocket drug cost of $4,050 reached the catastrophic coverage phase, in which they paid only a small copayment for their drugs until the end of the year.

PDPs have the latitude to structure their plans differently from what is presented here; companies offering non-standard plans must show that their coverage is at least actuarially equivalent to the standard plan. Many have developed plans with no deductibles or with drug copayments instead of the 25 percent coinsurance, and some plans provide generic and/or brand name drug coverage during the coverage gap.

The lower portions of the table show drug copayment, coinsurance, and deductible amounts for beneficiaries with full and non-full dual eligibility and with full or partial subsidies.

Figure 5.8 Medicare Part D non-LIS enrollees with specified monthly premium, by CKD stage & year (see page 126 for analytical methods. Point prevalent Medicare enrollees alive on January 1, excluding those in Medicare Advantage Part D plans.)
Figure 5.9 Medicare Part D non-LIS enrollees with gap coverage or no deductible, by CKD stage, 2008 (see page 126 for analytical methods. Point prevalent Medicare enrollees alive on January 1, excluding those in Medicare Advantage Part D plans.)
Figure 5.10 Medicare Part D LIS enrollees with specified coinsurance/copayment, by CKD stage & year (see page 126 for analytical methods. Point prevalent Medicare enrollees alive on January 1, excluding those in Medicare Advantage Part D plans.)

Patients without LIS pay full monthly premiums. From 2007 to 2008, enrollment increased in plans with higher premiums. In 2008, 28–31 percent of CKD patients enrolled in plans with premiums greater than $35 per month, compared to 25 percent of Medicare patients.

The percentage of Part D non-LIS enrollees with no deductible is similar in the general Medicare and CKD populations, at 76–78 percent. Gap ("donut hole") coverage, in contrast, is more common in CKD patients, at 17 compared to 12 percent. Nineteen percent of CKD Stage 4–5 patients in 2008 were enrolled in plans with gap coverage.

Most Part D LIS enrollees (full-benefit dual-eligible patients) pay no monthly premium, but non-institutionalized LIS patients do pay drug copayments or coinsurance based on income and assets. Sixty-nine to 75 percent of CKD patients with LIS have low or no copayments for their Part D medications, compared to 68 percent of general Medicare patients. Only 3–4 percent pay 15 percent coinsurance for their medications. And even CKD patients with high copayments (23 percent, on average, in 2008) paid a maximum of just $2.25 per generic and $5.60 for branded medication.

Overall Costs of Part D Enrollment 

Figure 5.11 Net Part D payment for Medicare enrollees (see page 126 for analytical methods. Medicare Part D enrollees.)

In 2008, total net Part D payment for patients with identified kidney disease (CKD patients not on dialysis, and ESRD patients) was $5 billion — 10 percent of total Part D prescription drug costs. These costs do not include costs of drugs billed to Part B, including intradialytic medications (ESAs, IV vitamin D, iron) and immunosuppressants.

Figure 5.12 Per person per year Medicare & out-of-pocket Part D costs for enrollees, 2008 (see page 126 for analytical methods. Medicare patients surviving 2007 with Medicare as primary payor & enrolled in Part D. Medicare PPPY: sum of Medicare payment & low income subsidy. Out-of-pocket: sum of patient payment & other qualifying amounts.)

At $4,255, the per person per year (PPPY) total cost of medications covered by Medicare Part D is 1.74 times higher in CKD patients than in the general Medicare population. Proportional to total Part D costs, however, out-of-pocket costs are lower in CKD patients, representing 17 percent of their PPPY costs compared to 19 percent for the general Medicare population. There is little variation in total Medicare Part D or out-of-pocket costs across CKD stages.

Figure 5.13 Per person per year Part D costs for enrollees, by low income subsidy (LIS) status, 2008 (see page 126 for analytical methods. Medicare patients surviving 2007 with Medicare as primary payor & enrolled in Part D, & period prevalent dialysis patients, 2008, with Medicare as primary payor. Medicare PPPY: sum of Medicare payment & low income subsidy.)

PPPY total costs for Part D-covered medications in 2008 were 3.3–3.8 times greater for LIS patients than for those without LIS. Costs in LIS and non-LIS patients vary from $3,694 and $978 PPPY, respectively, in the general Medicare population to $5,451 and $1,672 among patients with CKD, and to $6,674 and $2,043 among those on dialysis.

Table 5.b Total per person per year (PPPY) Medicare part D costs (dollars) for enrollees, by CKD stage & low income subsidy (LIS) status, 2008 (see page 126 for analytical methods. Medicare patients surviving 2007 with Medicare as primary payor & enrolled in Part D. Medicare PPPY: sum of Medicare payment & low income subsidy.)

Total per person per year (PPPY) Medicare Part D costs vary widely between those with and without LIS. Overall, there is less variation by CKD stage and gender (except for patients with Stage 1–2 CKD and LIS, in whom costs are lower for women) and more by age and race. Because they are Medicare-enrolled due to disability, patients younger than 65 have higher Part D costs than older patients of the same CKD stage. By race, and regardless of LIS status, PPPY costs in the Medicare population are highest for African Americans, and in the CKD population are highest for whites.

Figure 5.14 Geographic variations in Parts A & B & Part D PPPY costs (dollars) in general Medicare & CKD patients, by HSA, 2008 (see page 126 for analytical methods. Medicare patients surviving 2007 with Medicare as primary payor & enrolled in Part D. Medicare PPPY: sum of Medicare payment & low income subsidy.)

General Medicare and CKD Part A and Part B per person per year costs show similar geographic patterns and are highest in areas of Texas, Mississippi, California, and Florida. CKD Part A and B costs are twice those found in general Medicare patients, averaging $24,925 and $10,334, respectively, in the upper quintile. Similar geographic cost patterns exist as well for Part D patients. CKD Part D costs in the upper quintile average $4,416, compared to $2,640 for general Medicare Part D; costs for both populations tend to be highest in Alaska, parts of California and Appalachia, and areas in New England.

Coverage Phase Analyses for Part D Enrollees 

Figure 5.15 Part D non-LIS enrollees who reach each coverage phase, by CKD stage, 2008 (see page 126 for analytical methods. Point prevalent Medicare enrollees alive on January 1, excluding those in employer-sponsored & national PACE Part D plans.)
Figure 5.16 Cumulative percent of Part D non-LIS enrollees who reach the coverage gap, by CKD stage, 2008 (see page 126 for analytical methods. Point prevalent Medicare enrollees alive on January 1, excluding those in employer-sponsored & national PACE Part D plans.)
Figure 5.17 Cumulative percent of Part D non-LIS enrollees who reach catastrophic coverage after reaching the coverage gap, by CKD stage, 2008 (see page 126 for analytical methods. Point prevalent Medicare enrollees alive on January 1, excluding those in employer-sponsored & national PACE Part D plans.)

Part D enrollees without the low income subsidy (LIS) may encounter three coverage phases, depending on total and out-of-pocket costs per year. In 2008, patients with total Part D drug costs up to $2,510 fell into the initial coverage phase, while those with costs over that amount entered the coverage gap ("donut hole"), in which they were responsible for 100 percent of drug costs. Patients whose out-of-pocket total reached $4,050 then entered the catastrophic coverage phase, in which they paid only a fraction of overall drug costs.

In 2008, 42 percent of all CKD patients (those not on dialysis) reached the coverage gap, compared to 23 percent in the general Medicare population; this varied little by CKD stage. Eight percent reached catastrophic coverage, compared to 3 percent of general Medicare patients.

Patients with Stage 3 CKD reach the coverage gap slightly sooner, on average, than those with CKD of other stages, while general Medicare patients take the longest. And 18–21 percent of CKD patients who reach the coverage gap subsequently attain catastrophic coverage, compared to 12.5 percent of general Medicare patients. Patients with Stage 4–5 CKD reach catastrophic coverage slightly faster than do patients in the earlier stages of CKD, and patients with CKD of any stage reach this coverage considerably faster than general Medicare patients.

Table 5.c Twelve-month probability of reaching the coverage gap in Part D non-LIS enrollees, by CKD stage, 2008 (see page 126 for analytical methods. Point prevalent Medicare enrollees alive on January 1, excluding those in employer-sponsored & national PACE Part D plans.)

Across CKD stages, 40–46 percent of non-LIS Part D enrollees reach the coverage gap within 12 months; this varies little by age or gender. In the general Medicare population, white patients are more likely to reach the gap. This is true in the overall CKD population as well, but for Stage 3–5 CKD patients the probability is highest among Asians, and lowest in African Americans. By comorbidity, patients with diabetes reach the gap at a higher rate than do those with other diagnoses. Not surprisingly, the likelihood of reaching the gap rises with the number of prescription fills per month in the previous year.

Table 5.d Rate of Part D-covered prescription fills per person per month in Part D non-LIS enrollees, by CKD stage, 2008 (see page 126 for analytical methods. Point prevalent Medicare enrollees alive on January 1, excluding those in employer-sponsored & national PACE Part D plans.)

Number, fill rate, and prescription cost influence whether patients stay in the initial coverage phase or progress to the coverage gap and then to catastrophic coverage. Among CKD patients who reach the initial coverage phase or coverage gap, the fill rate rises monotonically from Stages 1–2 to Stages 4–5. In patients reaching catastrophic coverage, however, the rate decreases monotonically with CKD stage.

Among patients who reach the coverage gap, the fill rate consistently declines from that of the initial coverage period. This could be due either to a reduction in medication adherence or to a decision to obtain medications outside the Part D plan, and it is a pattern not seen in patients who reach catastrophic coverage. In these patients, the fill rate rises as patients move from initial coverage to the gap, and then again as they reach catastrophic coverage. Patients with a higher number of Part D medications could be incentivized to fill prescriptions in order to reach this phase more quickly, as their out-of-pocket expenses then decrease dramatically.

Table 5.e Top 15 drugs used in general Medicare Part D enrollees, by frequency & net cost, 2008 (see page 126 for analytical methods. Part D claims for all patients in the Medicare 5 percent sample (claims & costs scaled up by a factor of 20 to estimate totals). Costs are the sum of Medicare payment & low income subsidy. All patients in the Medicare 5 percent sample (5.e). CKD Stage 1–2 (5.f) & CKD Stage 3–5 (5.g) Medicare patients, with Medicare as primary payor for calendar year 2007; all Part D claims for calendar year 2008 are included.)
Table 5.f Top 15 drugs used in general Medicare Part D enrollees with Stage 1-2 CKD, by frequency & net cost, 2008 (see page 126 for analytical methods. Part D claims for all patients in the Medicare 5 percent sample (claims & costs scaled up by a factor of 20 to estimate totals). Costs are the sum of Medicare payment & low income subsidy. All patients in the Medicare 5 percent sample (5.e). CKD Stage 1–2 (5.f) & CKD Stage 3–5 (5.g) Medicare patients, with Medicare as primary payor for calendar year 2007; all Part D claims for calendar year 2008 are included.)
Table 5.g Top 15 drugs used in general Medicare Part D enrollees with Stage 3-5 CKD, by frequency & net cost, 2008 (see page 126 for analytical methods. Part D claims for all patients in the Medicare 5 percent sample (claims & costs scaled up by a factor of 20 to estimate totals). Costs are the sum of Medicare payment & low income subsidy. All patients in the Medicare 5 percent sample (5.e). CKD Stage 1–2 (5.f) & CKD Stage 3–5 (5.g) Medicare patients, with Medicare as primary payor for calendar year 2007; all Part D claims for calendar year 2008 are included.)

In terms of frequency of use, the top 15 drugs covered by Medicare Part D are similar in the general Medicare and CKD populations. Furosemide, for example, is the most frequently used drug in the CKD population, and fifth on the list for general Medicare patients. Two drugs — hydrochlorothiazide and metformin — appear in the top 15 for general Medicare patients, but not for CKD patients, in whom furosemide (a loop diuretic) has a more potent diuretic effect, and metformin is contraindicated secondary to the increased risk of lactic acidosis. Carvedilol, allopurinol, and atenolol, in contrast, make the list only for CKD patients. Interestingly, potassium chloride is one of the most frequently used medications in the CKD population, which may indicate a more aggressive use of diuretics in these patients.

When ranked by net cost, the list of medications used in the general Medicare population contains more psychiatric drugs than do the lists for CKD patients. Epoetin alfa, in contrast, appears only in the CKD lists. Regardless of CKD stage, the highest net costs in the CKD population are for insulin, reflecting both the high prevalence of diabetes in these patients and the fact that many new insulin therapies are still under patent and not available as generics.

Medicare Part D Prescription Drug Use & Costs 

Figure 5.18 Top 15 drugs in general Medicare Part D enrollees, by frequency, 2008 (see page 126 for analytical methods. Part D claims for all patients in the Medicare 5 percent sample (5.18–19). Medicare CKD patients with Medicare as primary payor for calendar year 2007; includes Part D claims for calendar year 2008 (5.20–21). Cumulative percentage is the percentage of all Part D, & costs are the sum of Medicare payment & low income subsidy. Claims & costs scaled up by a factor of 20 to estimate totals.)
A Levothyroxine
B Metoprolol
C Simvastatin
D Lisinopril
E Furosemide
F Amlodipine
G Atorvastatin
H Atenolol
I Hydrochlorothiazide
J Metformin
K Potassium chloride
L Omeprazole M Warfarin
N Clopidogrel bisulfate
O Insulin

Figure 5.19 Top 15 drugs in general Medicare Part D enrollees, by cost, 2008 (see page 126 for analytical methods. Part D claims for all patients in the Medicare 5 percent sample (5.18–19). Medicare CKD patients with Medicare as primary payor for calendar year 2007; includes Part D claims for calendar year 2008 (5.20–21). Cumulative percentage is the percentage of all Part D, & costs are the sum of Medicare payment & low income subsidy. Claims & costs scaled up by a factor of 20 to estimate totals.)
A Atorvastatin
B Clopidogrel bisulfate
C Insulin
D Quetiapine
E Olanzapine
F Esomeprazole
G Donepezil
H Fluticasone/salmeterol
I Risperidone
J Pioglitazone
K Aripiprazole
L Lansoprazole
M Oxycodone
N Pantoprazole
O Divalproex

Figure 5.20 Top 15 drugs in Medicare Part D enrollees with CKD, by frequency, 2008 (see page 126 for analytical methods. Part D claims for all patients in the Medicare 5 percent sample (5.18–19). Medicare CKD patients with Medicare as primary payor for calendar year 2007; includes Part D claims for calendar year 2008 (5.20–21). Cumulative percentage is the percentage of all Part D, & costs are the sum of Medicare payment & low income subsidy. Claims & costs scaled up by a factor of 20 to estimate totals.)
A Furosemide
B Metoprolol
C Levothyroxine
D Amlodipine
E Lisinopril
F Insulin
G Simvastatin
H Potassium chloride
I Clopidogrel bisulfate
J Warfarin
K Atorvastatin
L Omeprazole
M Carvedilol
N Allopurinol
O Atenolol

Figure 5.21 Top 15 drugs in Medicare Part D enrollees with CKD, by cost, 2008 (see page 126 for analytical methods. Part D claims for all patients in the Medicare 5 percent sample (5.18–19). Medicare CKD patients with Medicare as primary payor for calendar year 2007; includes Part D claims for calendar year 2008 (5.20–21). Cumulative percentage is the percentage of all Part D, & costs are the sum of Medicare payment & low income subsidy. Claims & costs scaled up by a factor of 20 to estimate totals.)
A Insulin
B Clopidogrel bisulfate
C Atorvastatin
D Esomeprazole
E Donepezil
F Epoetin Alfa
G Pioglitazone
H Quetiapine
I Fluticasone/salmeterol
J Pantoprazole
K Olanzapine
L Lansoprazole
M Tamsulosin
N Omeprazole
O Oxycodone

As measured by total days supply, insulin therapies represented 2.8 percent of Part D drug use among CKD patients in 2008, but 6.2 percent of their Part D costs — the same relative proportion seen in general Medicare patients, where insulin therapies represented 1.3 percent of Part D drug use and 2.9 percent of costs. This suggests that CKD patients are being prescribed branded insulin therapies at about the same rate as their general Medicare counterparts. The same is true of clopidogrel, accounting for 2.1 percent of Part D use in the CKD population, and 1.5 percent among general Medicare patients.

Epoetin alfa, used for anemia treatment in CKD patients, accounted for 2.1 percent of their total Part D costs in 2008. This may represent CKD patients who are self-administering epoetin alfa at home, rather than having it administered in a doctor's office, where billing would be covered under Medicare Part B rather than Part D.

Furosemide, metoprolol, and levothyroxine were the top three drugs, as measured by total days supply, in the identified CKD population in 2008; they accounted for nearly 12 percent of total Part D drug use in this population. The top 15 drugs represented 37 and 34 percent of Part D use and costs in these patients, similar to the 35 and 30 percent seen in general Medicare patients.