The launch of several expensive hepatitis C drugs over the past few years has ushered the topic of high cost prescription drugs back into the public’s and policymakers’ attention. With over 70 million beneficiaries, many of whom have complicated health needs, Medicaid is one of the largest providers of prescription drugs in the United States. State Medicaid agencies have a variety of mechanisms to help control outpatient drug spending that they have been using widely for the past decade. However, new prescription drugs are always coming to market, and the health needs of the Medicaid population change over time, especially as enrollment grows because of ACA Medicaid expansion. As a result, it is helpful for policy makers to understand which drugs used by the Medicaid program are most expensive. In this issue brief, we determine the 50 most costly drugs before rebates used by the Medicaid program from January 2014 through June 2015. We then examine reasons why these drugs are so costly, as well as exploring case studies on opioids, hepatitis C drugs, and the drug Abilify. Key findings include the following:

• Among the most commonly prescribed outpatient prescription drugs in Medicaid, the top five drugs are used for pain relief (hydrocodone-acetaminophen and ibuprofen), management of chronic illness (lisinopril and omeprazole), and antibiotics (amoxicillin). However, these drugs are not necessarily among the most costly used by Medicaid as many are inexpensive at the per prescription level.
• Several drugs that are very costly at the per prescription level reflect Medicaid’s role in caring for individuals with substantial health needs. These include drugs to treat hemophilia (NovoSeven RT, Koate-DVI, Feiba), multiple sclerosis (HP Acthar), and rare infant diseases (Adagen). However, most of these drugs are not commonly used among the Medicaid population.
• Aggregate drug costs to Medicaid reflect both frequency of use and per prescription cost. Among the most costly drugs in aggregate used by the Medicaid program are drugs used to treat costly illnesses for which Medicaid is a key source of coverage, including behavioral health conditions (Abilify and Vyvanse), hepatitis C (Sovaldi and Harvoni), and HIV (Truvada). Nearly three quarters of the 50 most costly drugs fall into five drug groups, the most prevalent of which is antivirals, which includes drugs used to treat HIV as well as hepatitis C drugs.
• 45 of the 50 most costly drugs fall into the high-cost category in part or primarily because they are frequently prescribed. Hydrocodone-acetaminophen and Suboxone, both opioids, a drug group which has garnered much public attention recently, fall into this category, as do several drugs used to treat ADHD.
• Many of the most costly drugs have some form of regulatory and consequently market exclusivity, thus enabling the manufacturers to charge a premium price for the drug at the prescription level. Twenty-two of the most costly drugs are particularly expensive at the prescription level, including the most costly drug before rebates used by Medicaid over this period, Abilify (an atypical antipsychotic).

As states continue to implement an array of measures to control Medicaid prescription drug costs, they are challenged to balance costs with access to drugs needed by beneficiaries. This challenge is not unique to Medicaid alone but is a common issue throughout the U.S. healthcare system, as reflected by the attention in general given to the high cost of prescription drugs. While this analysis may not reflect Medicaid’s net expenditures for a particular drug, as it is unable to account for drug rebates, it provides insights into which drugs are driving Medicaid prescription drug expenditures and why.

Introduction

With its list price of $84,000 per treatment, the launch of the hepatitis C drug Sovaldi in December 2013 garnered the public’s and policymakers’ attention and brought into the spotlight the issue of high-cost prescription drugs in the U.S. Most Americans now believe that prescription drugs are too expensive.¹  With over 70 million beneficiaries,²  the Medicaid program is larger than any other public or private insurer.³  Many Medicaid beneficiaries have poorer health than enrollees in private coverage⁴  and need prescription drugs to manage their medical conditions. As a result, Medicaid prescription drug spending is sizeable: in 2014, Medicaid spent $27.3 billion on outpatient drugs.⁵  Over the years, states have implemented an array of measures to control utilization and spending for prescription drugs.⁶ 

In this issue brief, we look at which outpatient prescription drugs were most expensive to Medicaid in 2014 and 2015 and explore why. As described in more detail in Appendix B, we merge state drug utilization data (available from the Centers for Medicare and Medicaid Services, or CMS) with Wolters Kluwer Clinical Drug Information (WKCDI) data to determine the 50 most costly outpatient drugs to the Medicaid program from January 2014 through June 2015 in terms of aggregate spending before rebates.⁷  Then, using WKCDI as well as data from the Food and Drug Administration (FDA), we focus on common attributes among them and discuss policy implications.

Background

The Medicaid outpatient prescription drug benefit is not a mandatory benefit, but all states provide this benefit in their Medicaid programs. Typically, a Medicaid beneficiary receives a prescription from their physician and fills it at a pharmacy. Medicaid either reimburses the pharmacy for the prescription or pays a capitation fee to a managed care company that reimburses the pharmacy for the prescription.⁸  States then collect rebates for the drugs prescribed to their beneficiaries from the manufacturers and share this with the federal government.⁹  As a result of the Medicaid Drug Rebate Program created by Congress in 1991, all drug manufacturers must enter into a rebate agreement with the Secretary of Health and Human Services in return for Medicaid reimbursement for their prescription drugs.¹⁰  The rebate program was put into place to control the rising cost of drugs within Medicaid.¹¹ 

In the early 2000s, as Medicaid spending for prescription drugs grew, states implemented a number of cost control measures in this area, such as supplemental rebates, preferred drug lists, and generic substitution. With the implementation of Medicare Part D (and the shift of drug costs for dually eligible beneficiaries from Medicaid to Medicare), Medicaid saw a large decline in prescription drug spending, and state activity in this area slowed.¹²  However, after years of slow growth, Medicaid spending on outpatient prescription drugs jumped 24% in 2014 to $27.3 billion, or about 6 percent of Medicaid spending.¹³  This increase in the growth rate is in large part due to increased enrollment resulting from Medicaid expansion and new high cost hepatitis C drugs.¹⁴ 

In January 2016, CMS released the Covered Outpatient Drugs final rule, which details drug reimbursement provisions outlined in the ACA, some of which were put into place to help states control drug spending. One such provision is the requirement that states must base drug ingredient cost reimbursement to pharmacies on the actual price of the drug, as opposed to using list prices.¹⁵  The previously used ingredient cost reimbursements, referred to as “Estimated Acquisition Costs,” often relied on “commercially published reference prices” which did not necessarily represent the actual cost of the drug to the pharmacy and were therefore susceptible to being manipulated.¹⁶  Many states had already made the change to using actual acquisition costs prior to the final rule.¹⁷  Despite this policy change, some drug manufacturers may still have the leverage to charge a high price for a drug due to lack of competition.

As states continue to grapple with controlling Medicaid prescription drug spending, many are focusing on high-cost “specialty drugs.” In FY 2015, over two-thirds of the states reported actions to refine and enhance their pharmacy programs in response to new and emerging specialty and high-cost drug therapies.¹⁸  However, it is unclear what role specialty drugs—or other types of drugs— play in Medicaid prescription drug spending. This analysis examines which drugs are the most costly to Medicaid and looks at common attributes of the most costly drugs. This information may be helpful as states undertake policy actions to balance high costs with high needs.

What Makes a Drug a High Cost to the Medicaid Program?

Aggregate drug costs to Medicaid reflect both frequency of use and per prescription costs. Among the most commonly prescribed outpatient prescription drugs in Medicaid, the top five drugs are used for pain relief (hydrocodone-acetaminophen and ibuprofen), management of chronic illness (lisinopril and omeprazole), and antibiotics (amoxicillin) (see Appendix Table A3). However, these drugs are not necessarily among the most costly to Medicaid as many are inexpensive at the per prescription level. Similarly, many drugs that are quite costly at the per prescription level are not commonly used by Medicaid enrollees. These drugs, which include drugs to treat hemophilia (NovoSeven RT, Koate-DVI, Feiba), multiple sclerosis (HP Acthar), and rare infant diseases (Adagen), reflect Medicaid’s role in caring for individuals with substantial health needs.

When we examined which drugs were high cost by assessing total Medicaid spending by drug name for each outpatient drug provided from January 2014 through June 2015,¹⁹  Abilify, Sovaldi, Vyvanse, Harvoni, and Truvada top the list, respectively (see Appendix Table A1). These are drugs to treat costly illnesses for which Medicaid is a key source of coverage, including behavioral health conditions (Abilify and Vyvanse), hepatitis C (Sovaldi and Harvoni), and HIV (Truvada).

In Figure 1, we classify the 50 most costly (in aggregate) drugs in terms of how frequently prescribed and how expensive they are at the prescription level. If they are in the top 10^th percentile of drugs by number of prescriptions, we identify them as “frequently prescribed.” If they were in the top 10^th percentile of drugs by spending per prescription, we refer to them as an “expensive at the prescription level.”

We found that of the 50 most costly drugs, 45 fall into the high-cost category at least in part because they are very frequently prescribed. Over half (28) are frequently prescribed but not expensive at the prescription level. Among others, these drugs include treatments for ADHD and hydrocodone-acetaminophen, the most frequently prescribed drug in Medicaid over this period (Appendix Table A3 and “Case Study A: Opioids”).

A smaller number of the 50 most costly drugs (17) are both frequently prescribed and expensive prescriptions. Seven antiretrovirals included among the 50 most costly drugs fall into this group (Truvada, Atripla, Prezista, Stribild, Complera, Reyataz, and Isentress). Among others, this category also includes Abilify, an atypical antipsychotic and the most costly drug to Medicaid (Appendix Table A1), and Humira Pen, a drug used to treat arthritis.

The five remaining drugs are expensive at the prescription level but not in the top tenth percentile by prescriptions, and thus we have not defined them as being frequently prescribed.²⁰  Among others, these drugs include the hepatitis C agents Sovaldi and Harvoni.

Which Drugs are High Cost to Medicaid?

As shown in Figure 2, 72% (36) of the 50 most costly drugs are in five drug groups. Antivirals are the most common drug group among the most costly drugs, accounting for 20% of the top 50 drugs. The antivirals comprise seven antiretrovirals (drugs that are used primarily in the treatment of HIV), two hepatitis C agents, and one other type of antiviral. Reflecting the Medicaid population’s serious health needs, antiretrovirals are widely used in the program.²⁵  Antiretrovirals are also costly on a per prescription basis. Hepatitis C agents are also costly on a per prescription basis but are not as frequently prescribed (Figure 1).

Antiasthmatics and bronchodilators are the second most common drug group in the 50 most costly drugs, accounting for 16% of the top 50 drugs. These are drugs used in the treatment of asthma and chronic obstructive pulmonary disease (COPD). None of these drugs are among the most expensive per prescription, but all are very frequently prescribed.²⁶ 

With 14% (7), ADHD/Anti-Narcolepsy/Anti-Obesity/Anorexients are the third most common drug group in the most costly drugs. Although the drug group implies that these drugs can be used for a variety of conditions, the seven drugs within this drug group all have a common indication of attention deficit hyperactivity disorder (ADHD). Like the antiasthmatics and bronchodilators, all of these drugs are frequently prescribed and none of them are as expensive at the prescription level (Appendix Table A1). This high prescription rate is in part attributable to the high incidence rate of ADHD/ADD among children in general, among children in Medicaid, and among foster children in Medicaid.²⁷ 

Twelve percent (6) of the most costly drugs are antipsychotics. Abilify, the most costly drug to Medicaid, is included within this group. Five out of the six antipsychotics in the 50 most costly drugs are both frequently prescribed and expensive at the prescription level (Appendix Table A1).  Antipsychotics are FDA-approved to treat schizophrenia and bipolar disorder; however, they are also frequently prescribed “off-label” for other conditions.²⁸ 

At ten percent, antidiabetics are the fifth most common drug group among the most costly drugs, with five drugs in this class falling into the top 50 most costly Medicaid drugs. Four of these five antidiabetics are insulins. None of the antidiabetics are in the top tenth percentile of drugs by spending per prescription, although the price of insulin has been increasing in recent years.²⁹  All antidiabetics are very frequently prescribed, reflecting the high number of people in Medicaid who have diabetes: prior to Medicaid expansion, about four million Medicaid beneficiaries were diagnosed with diabetes.³⁰ 

How Does Market Exclusivity Affect Price?

In the absence of competition, a manufacturer may be able to price a drug higher. Patents and regulatory exclusivity, put into place as an incentive for innovation, are ways that a manufacturer can protect their product against competition. Patents have a twenty year duration, but manufacturers generally obtain them while their product is in preclinical and clinical trials, well before the FDA approves their product and well before the product launches. As a result, the duration remaining on the manufacturer’s original patent once the drug launches is usually much shorter than 20 years.⁴⁵  Separate from the patent system, a manufacturer is also able to obtain regulatory exclusivity for their product from the FDA. Many of the 50 most costly drugs have some form of regulatory exclusivity.

Brand Versus Generic Drugs

A brand drug is generally considered to be a drug that has received FDA approval after the manufacturer has proven the drug’s safety and efficacy. The FDA awards a regulatory exclusivity period of 3 or 5 years to brand drugs.⁵³  Regulatory exclusivity provides the manufacturer with a degree of market exclusivity, enabling them to price the drug accordingly and providing incentive for them to market it as a non-commodity, which includes naming the drug with appealing brand name. Alternatively, a manufacturer can obtain FDA approval for their drug by proving that it is bioequivalent to a brand drug,⁵⁴  skipping the long and expensive process of proving a drug is safe and effective. The FDA identifies these drugs as generic.⁵⁵  They cannot enter the market while the corresponding brand still has exclusivity.⁵⁶  Once generic drugs enter the market, the price of the drug usually falls due to competition.

Compared to all drugs reimbursed through Medicaid from January 2014 through June 2015, we found a disproportionate number of the 50 most costly drugs are brands, as opposed to being generics (Figure 3).⁵⁷  Ninety-four percent of the 50 most costly drugs were available as brand-name drugs, compared to 45% of all drugs reimbursed by Medicaid.

Biologics

A biologic is a drug that is derived from an animal or microorganism. It is more complex than traditional small-molecule drugs synthesized in a lab.⁵⁸  Because biologics are structurally very different from small molecule drugs and are approved through a different process,⁵⁹  there was not automatically a structure in place for generic approvals resulting in an absence of a generic market to commoditize biologic drugs. However, as part of the ACA,⁶⁰  biologics now have 12 years of regulatory exclusivity,⁶¹  with an abbreviated pathway for the biosimilars, the biologic equivalent of a generic, now in place. Although biosimilars are expected to lower the price of the original biologic, they are not expected to lower it to that degree that generics lower the price of the original small-molecule brand drug.⁶²  In March 2015, the FDA approved its first biosimilar, Zarxio, and the drug launched the following September.⁶³ 

We found that 12% (6) of the 50 most costly drugs in Medicaid were biologics, compared with 3% of outpatient drugs reimbursed through Medicaid overall (Figure 3). The 6 biologics in the 50 most costly drugs are Humira Pen (used to treat psoriasis and some types of arthritis), Synagis (used to prevent serious lung disease), Advate (used in the treatment of hemophilia), Enbrel SureClick (used to treat psoriasis and some types of arthritis), Neulasta (used to treat possible side effects of chemotherapy), and NovoSeven RT (used in the treatment of hemophilia). All of these 6 biologics are in the top 5^th percentile in terms of Medicaid spending per prescription. Twenty-nine of the 50 most expensive drugs by spending per prescription are biologics, showing that although these drugs are often very expensive, they do not necessarily appear as large budget items for Medicaid because of lower utilization (Appendix Table A2).

Orphan Drugs

The FDA provides orphan drug designations to drugs that treat fewer than 200,000 people in the U.S. or those that treat a disease for which the manufacturer does not expect to recover the cost of the drug.⁶⁴  Having an orphan drug designation entitles the sponsor to many benefits,⁶⁵  including a seven-year period of regulatory exclusivity associated with the drug’s indication. The 1982 Orphan Drug Act has generated an increase in the number of drug designations targeting rare diseases.⁶⁶  However, some argue that it is being used to create blockbuster drugs, as manufacturers slice more common diseases into subtypes affecting fewer than 200,000 Americans and gain an orphan drug designation for a subtype, with the drug ultimately being used widely for other conditions.⁶⁷ 

We found that 22% (11) of the 50 most costly drugs had achieved an orphan drug designation at some point (Figure 3). This includes Abilify, the most costly drug to Medicaid. Nine of these 11 drugs were among Medicaid’s top 10^th percentile of most prescribed drugs.

Specialty Drugs

Although there is not one universally accepted definition, a specialty drug is generally considered to be a drug that requires difficult or unusual handling or is for a difficult-to-treat disease. Price is also often an indicator of a specialty drug.⁶⁸  We found that a disproportionate number of drugs in the 50 most costly drugs are considered specialty drugs based on formulary review, with 34% (17) of the most costly drugs being considered specialty drugs, compared to 6% of all Medicaid covered outpatient drugs (Figure 3). Given that price factors into whether a drug is considered a specialty drug, it is not entirely surprising that so many of the most costly drugs are specialty drugs, and that many have a type of regulatory exclusivity. Of the 17 specialty drugs in the most costly drugs, none are multi-source generics (Appendix Table A1), and all but one are only available as branded single-source products with one labeler (data not shown). Six of the 17 specialty drugs are biologics, meaning that all biologics among the most costly drugs are considered specialty drugs (Appendix Table A1).

Policy Implications

In this analysis, we found that although all of the most costly drugs to Medicaid are frequently prescribed, expensive at the prescription level, or both; a majority are frequently prescribed. Access to prescription drugs is crucial for the treatment of many conditions found in the Medicaid population, which is more likely to have health issues than the privately insured. Although the prescription drug benefit is not mandatory for states, all state Medicaid programs provide it to their Medicaid beneficiaries. However, it can be expensive, and states are forced to grapple with the costs of the benefit. Additionally, although state Medicaid programs may know a new drug is coming to market, they may not have a sense for how expensive it will be until it reaches the market. States have taken a number of actions to control the budgetary effects of high cost drugs, including implementing new prior authorization requirements and negotiating higher supplemental rebates and lower prices for certain drugs.⁶⁹ 

Balancing Cost with Access

Often, drugs that are high-priced are “orphan drugs” that treat rare diseases, which are allowed longer exclusivity so that the manufacturer can earn a profit on the drug. Sovaldi and Harvoni attracted attention because they were priced like orphan drugs,⁷⁰  but they are for diseases that are prevalent in the Medicaid population. As these drugs were coming to market, nearly all states expressed concern about how the cost of this treatment would affect their Medicaid spending.⁷¹  However, while high cost, these drugs are cures for most patients; they are more effective than the previous standard drug treatment for the disease;⁷²  and a full treatment of Sovaldi or Harvoni is less costly than a liver transplant, ⁷³  for which hepatitis C is the leading cause.⁷⁴  It is important to take a broad view when considering prescription drug costs, as many costly drugs prevent expensive emergency department visits and hospital stays.  Regardless, states felt that it was not feasible to provide this drug to every beneficiary with hepatitis C immediately.⁷⁵  In response, CMS published guidance reminding state Medicaid programs that certain utilization controls are permissible, but when doing so, states must ensure that they are in compliance with statute.⁷⁶ 

In general, states can help control the overall costs of drugs by monitoring utilization and aiming to ensure that drugs are not overprescribed to patients. For example, states can require prior authorization or use a preferred drug list to control access. However, states generally must cover FDA-approved drugs for their medically accepted indications.⁷⁷  States may not replace unreasonable limits on access to medically necessary drugs, such as those that are contrary to the professional standard of care. Many states have attempted to balance the public health need for hepatitis C drugs with their high cost through the implementation of prior authorization and other requirements.⁷⁸   Spending and utilization for hepatitis C agents is lower than it would have been without these restrictions, but some beneficiaries who need these drugs are unable to access them, and a number of lawsuits have been filed against different state Medicaid agencies challenging state policies that limit access contrary to generally accepted standard of care among medical professionals.

Monitoring High Utilization

Many of the most costly drugs to Medicaid are so costly because they are frequently prescribed, including hydrocodone-acetaminophen, an opioid. While there are many medically necessary reasons to prescribe this drug, there is also a great deal of evidence to suggest overutilization of opioids. There is much that states can do to address the misuse of opioids, such as undertaking provider education; removing methadone⁷⁹  from the preferred drug lists; establishing clinical criteria for obtaining a methadone prescription; requiring step therapy, prior authorization, or prescription quantity limits; using drug utilization review⁸⁰  measures to identify potential misuse of opioids; increasing access to and use of prescription drug monitoring program data, and implementing patient review and restriction programs.⁸¹  States have acknowledged the severity of this public health crisis, and nearly all have prescription monitoring programs in place.⁸²  There are hundreds of proposals in legislatures to regulate clinics and prescription behavior.⁸³  The federal government has awarded money to health centers to focus on opioid abuse,⁸⁴  and in March the Centers for Disease Control and Prevention released opioid prescription guidelines.⁸⁵  Also as part of its collection of Medicaid quality measures, CMS is beginning to collect information on the use of opioids from multiple providers among non-cancer patients.⁸⁶ 

Opioids are not the only drug at risk of being overprescribed. Three of the seven ADHD/Anti-Narcolepsy/Anti-Obesity/Anorexients are amphetamines, which are stimulants and have a black box warning of “high potential for abuse.”⁸⁷  There is also general concern over the use of antipsychotics among children enrolled in Medicaid. As part of its collection of Medicaid and CHIP quality measures, CMS is beginning to collect information on children prescribed more than one antipsychotic at a time.⁸⁸  As states monitor opioid and other drug groups as a way to address public health needs, they may realize budget savings due to lower utilization as well.

Promoting Innovation

Although policy makers are concerned with the cost of the prescription drug benefit, they also are concerned with beneficiaries having access to cures and treatments, and part of enabling this is incentivizing the pharmaceutical industry to research and bring to market new and needed drugs. In the U.S., innovation in the pharmaceutical industry is incentivized through regulatory exclusivity. This analysis has shown that many high cost drugs have some form of regulatory exclusivity. As Congress searches for ways to encourage innovation, they also remain aware of the varying ability for different payers to sustain these incentives. Nonetheless, although different payers face different parameters than Medicaid when paying for prescription drugs, this struggle to balance incentivizing innovation with actually being able to pay for the prescribed drugs is not unique to Medicaid alone, but a common challenge across the entire U.S. health care system.

For our analysis of the 50 most costly drugs in the Medicaid program, we used 2014 and quarters one and two of 2015 State Drug Utilization Data available from CMS merged with data from Wolters Kluwer Clinical Drug Information, Inc (“WKCDI”).⁸⁹  The State Drug Utilization Data is publicly available data used as part of the Medicaid Drug Rebate Program. This data provides information on the number of prescriptions, Medicaid spending, and cost-sharing for rebate-eligible Medicaid outpatient drugs at the National Drug Code (NDC) level. The WKCDI data provides product information for drug products.  We accessed State Drug Utilization Data in February 2016 and used the most recent data available for all states. The WKCDI data is also from February 2016. The use of WKCDI data does not represent and should not be characterized as a WKCDI endorsement of any data, findings, or other content presented in this report.

We merged the State Drug Utilization Data and the WKCDI data at the NDC-level to consistently identify the drug name, as well as to incorporate brand versus generic status and the WKCDI Therapeutic Classifications System’s drug group. We classified single-source and multi-source, originator drugs as brand drugs. If a drug was available as both a brand and a generic, we categorized it as a brand when summarizing how many of the most costly drugs were brands and how many were generics. Using the Center for Drug Evaluation and Research List of Licensed Biological Products and the Center for Biologic Evaluation and Research List of Licensed Biological Products⁹⁰  as of February 2016, we identified all biologics in the State Drug Utilization Data based on drug name. We looked up the orphan drug status of the 50 most costly drugs in Medicaid in the Orphan Drug Product designation database.⁹¹  To identify specialty drugs in the State Drug Utilization Data, we compiled a list using formularies from the top Medicaid pharmacy benefit managers.⁹²  We identified specialty drugs using the drug name.

To determine the 50 most costly drugs to Medicaid, we summed total Medicaid spending before rebates by drug name over the 2014 quarter one through 2015 quarter two period. Due to data reliability, we were unable to include 2014 quarter two data from Kansas or Virginia. We also calculated average Medicaid spending per prescription and summed total prescriptions by drug name over this period. We ranked the drugs by spending per prescription and total prescriptions, calculated their percentiles, and identified a drug as “frequently prescribed” or “expensive at the prescription level” if it was in the top 10^th percentile of either. We included the 50 most expensive drugs by spending per prescription and the 50 most prescribed drugs over the period in the Appendix tables. When reporting the former, we only included drugs with ten or more prescriptions to avoid any outliers in the data.

Limitations

This analysis does not include rebates, because this data is unavailable to the public at the NDC level. Rebates have a considerable effect on Medicaid drugs spending overall, but lower spending at the drug level at different rates.

Additionally, although Medicaid beneficiaries largely self-administer drugs that are prescribed in an outpatient setting, medical practitioners must administer some drugs. Although states are to collect drug rebates on all reimbursed outpatient drugs, regardless of whether they are physician- or self-administered, research has shown that not all states are collecting rebates on physician-administered drugs.⁹³  Because biologics and other specialty drugs are often physician-administered, it is possible that the data reflects lower Medicaid spending and utilization of certain drugs of this kind.

1. Bianca DiJulio, Jamie Firth, and Mollyann Brodie, Kaiser Health Tracking Poll: August 2015 (Washington DC: Kaiser Family Foundation, August 2015), https://modern.kff.org/health-costs/poll-finding/kaiser-health-tracking-poll-august-2015/. ↩︎
2. This total also includes CHIP enrollment. See “Total Monthly Medicaid and CHIP Enrollment,” Kaiser Family Foundation, accessed July 6, 2016, https://modern.kff.org/health-reform/state-indicator/total-monthly-medicaid-and-chip-enrollment/. ↩︎
3. U.S. Congress, Senate, Committee on Finance, The Price of Sovaldi and Its Impact on the U.S. Health Care System, 114th Congress, 1st session, 2015, 80, http://www.finance.senate.gov/imo/media/doc/1%20The%20Price%20of %20Sovaldi%20and%20Its%20Impact%20on%20the%20U.S.%20Health%20Care%20System%20(Full%20Report).pdf. ↩︎
4. Julia Paradise and Rachel Garfield, What is Medicaid’s Impact on Access to Care, Health Outcomes, and Quality of Care? Setting the Record Straight on the Evidence (Washington DC: Kaiser Commission on Medicaid and the Uninsured, August 2013), https://modern.kff.org/medicaid/issue-brief/what-is-medicaids-impact-on-access-to-care-health-outcomes-and-quality-of-care-setting-the-record-straight-on-the-evidence/. ↩︎
5. This includes rebates and spending through managed care. See “National Health Expenditure Accounts,” CMS, accessed July 6, 2016, https://www.cms.gov/Research-Statistics-Data-and-Systems/Statistics-Trends-and-Reports/NationalHealthExpendData/NationalHealthAccountsHistorical.html. ↩︎
6. Laura Snyder and Robin Rudowitz, Trends in State Medicaid Programs: Looking Back and Looking Ahead, (Washington DC: Kaiser Commission on Medicaid and the Uninsured, June 2016), https://modern.kff.org/medicaid/issue-brief/trends-in-state-medicaid-programs-looking-back-and-looking-ahead/. ↩︎
7. In aggregate, the 50 most costly drugs compose about 40% of Medicaid drug spending before rebates over this period. ↩︎
8. This is the case for beneficiaries not also dually enrolled in Medicare. Since 2006, Medicare Part D has paid for outpatient prescriptions of beneficiaries dually enrolled in Medicaid and Medicare. ↩︎
9. 42 USC  § 1396r-8(b)(1)(A). ↩︎
10. 42 USC  § 1396r-8(a)(1). ↩︎
11. Brian Bruen and Katherine Young, What Drives Spending and Utilization on Medicaid Drug Benefits in States? (Washington DC: Kaiser Commission on Medicaid and the Uninsured, December 2014), http://files.kff.org/attachment/brief-what-drives-spending-and-utilization-on-medicaid-drug-benefits. ↩︎
12. Vernon Smith, Kathleen Gifford, Eileen Ellis, Robin Rudowitz, Laura Snyder, and Elizabeth Hinton, Medicaid Reforms to Expand Coverage, Control Costs and Improve Care: Results from a 50-State Medicaid Budget Survey for State Fiscal Years 2015 and 2016, (Washington DC, Kaiser Commission on Medicaid and the Uninsured, October 2015), https://modern.kff.org/medicaid/report/medicaid-reforms-to-expand-coverage-control-costs-and-improve-care-results-from-a-50-state-medicaid-budget-survey-for-state-fiscal-years-2015-and-2016/. ↩︎
13. This includes outpatient prescription drug spending for both managed care and fee-for-service, as well as incorporating rebates.  Share of Medicaid spending does not include administrative or other non-personal care Medicaid spending. See National Health Expenditure Accounts, op. cit. ↩︎
14. Medicaid and CHIP Payment and Access Commission, “June 2016 Report to Congress on Medicaid and CHIP,” (Washington DC, June 2016), https://www.macpac.gov/wp-content/uploads/2016/06/June-2016-Report-to-Congress-on-Medicaid-and-CHIP.pdf. ↩︎
15. 42 CFR Part 447. ↩︎
16. 77 Fed. Reg. 5318 (Feb. 2, 2012). ↩︎
17. Brian Bruen and Katherine Young, Paying for Prescribed Drugs in Medicaid: Current Policy and Upcoming Changes, (Washington DC, Kaiser Commission on Medicaid and the Uninsured, May 2014), https://modern.kff.org/medicaid/issue-brief/paying-for-prescribed-drugs-in-medicaid-current-policy-and-upcoming-changes/ and Smith, Gifford, Ellis, Rudowitz, Snyder, and Hinton, op. cit. ↩︎
18. Smith, Gifford, Ellis, Rudowitz, Snyder, and Hinton, op. cit. ↩︎
19. Spending does not include rebates. For further details see the methodology section in the appendix. ↩︎
20. Four out of five of these drugs are in the top quartile by prescriptions.  The fifth is in the top 50th percentile by prescriptions. Although these drugs do not meet our definition of being “frequently prescribed,” they also are not rarely prescribed. ↩︎
21. “Drug overdose deaths in the United States hit record numbers in 2014,” CDC, accessed July 6, 2016, http://www.cdc.gov/drugoverdose/epidemic/index.html. ↩︎
22. Drew Altman, “Why Painkiller Addiction and Abuse are Rising Health-Care Priorities,” The Wall Street Journal, November 24, 2015, http://blogs.wsj.com/washwire/2015/11/24/why-painkiller-addiction-and-abuse-are-rising-health-care-priorities/. ↩︎
23. Vikki Wachino, “Best Practices for Addressing Prescription Opioid Overdoses, Misuse and Addiction,” CMCS Informational Bulletin, CMS, January 28, 2016, https://www.medicaid.gov/federal-policy-guidance/downloads/cib-02-02-16.pdf. ↩︎
24. The FDA has approved three medications for the treatment of opioid use disorder: buprenorphine, methadone, and naltrexone. Ibid. ↩︎
25. Prior to Medicaid expansion, in FY 2011, over 200,000 Medicaid enrollees were diagnosed with HIV or AIDS. See “Medicaid Enrollment and Spending on HIV/AIDS”, Kaiser Family Foundation, accessed July 6, 2016, https://modern.kff.org/hivaids/state-indicator/enrollment-spending-on-hiv/. Medicaid plays an important role in the care of people with HIV or AIDS and Medicaid expansion has increased the number of HIV patients in care. See Jennifer Kates, Lindsey Dawson, Tresa Undem, and Kathleen Perry, Health Insurance Coverage for People with HIV Under the Affordable Care Act: Experiences in Five States, (Washington DC, Kaiser Family Foundation, December 2014), https://modern.kff.org/hivaids/issue-brief/health-insurance-coverage-for-people-with-hiv-under-the-affordable-care-act-experiences-in-five-states/. ↩︎
26. Twenty-four percent of non-institutionalized Medicaid and CHIP enrollees had respiratory diseases in 2013 that incurred medical expenses. KFF analysis of 2013 Medical Expenditure Panel Survey (MEPS) household component, meps.ahrq.gov/mepsweb/. ↩︎
27. Medicaid and CHIP Payment and Access Commission, “Behavioral Health in the Medicaid Program – People, Use, and Expenditures,” June 2015 Report to Congress on Medicaid and CHIP, (Washington DC, June 2015), https://www.macpac.gov/wp-content/uploads/2015/06/June-2015-Report-to-Congress-on-Medicaid-and-CHIP.pdf. ↩︎
28. Louise Carton, Olivier Cottencin, Maryse Lapeyre-Mestre, Pierre Geoffroy, Jonathan Favre, Nicolas Simon, Regis Bordet, and Benjamin Rolland, “Off-Label Prescribing of Antipsychotics in Adults, Children, and Elderly Individuals: A Systematic Review of Recent Prescription Trends,” Current Pharmaceutical Design, 21, 23 (2015): 3280-3297, http://www.eurekaselect.com/132298/article. This practice of prescribing medications for uses not approved by the FDA is called “off-label prescribing” and is legal for non-controlled substances such as opioids.  However, it is illegal for a pharmaceutical company to market the drug for off-label uses. See “Off-label Drug Use,” American Cancer Society, accessed July 6, 2016, http://www.cancer.org/treatment/treatmentsandsideeffects/treatmenttypes/chemotherapy/off-label-drug-use. ↩︎
29. Ed Silverman, “Insulin prices have skyrocketed, putting drug makers on the defensive,” Stat News, April 5, 2016, https://www.statnews.com/pharmalot/2016/04/05/insulin-prices-skyrocketed-putting-drug-makers-defensive/. ↩︎
30. KCMU and Urban Institute estimates based on data from FY 2011 MSIS. ↩︎
31. Committee on Finance, op. cit., 1. ↩︎
32. Ibid. ↩︎
33. Paradise and Garfield, op. cit. ↩︎
34. FDA, FDA Approves Sovaldi for Chronic Hepatitis C, December 6, 2013, http://www.fda.gov/NewsEvents/Newsroom/PressAnnouncements/ucm377888.htm ↩︎
35. “What is Sovaldi?” Drugs.com, accessed July 6, 2016, http://www.drugs.com/sovaldi.html. ↩︎
36. Gilead Sciences, U.S. Food and Drug Administration Approves Gilead’s Harvoni ® (Ledipasvir/Sofosbuvir), the First Once-Daily Single Tablet Regimen for the Treatment of Genotype 1 Chronic Hepatitis C, October 10, 2014, http://www.gilead.com/news/press-releases/2014/10/us-food-and-drug-administration-approves-gileads-harvoni-ledipasvirsofosbuvir-the-first-oncedaily-single-tablet-regimen-for-the-treatment-of-genotype-1-chronic-hepatitis-c ↩︎
37. B.E. and A.R. v. Teeter, No. C16-227-JCC (W.D. Wa. May 27, 2016), https://today.law.harvard.edu/wp-content/uploads/2016/06/40-5-27-16-Order-Granting-Preliminary-Injunction.pdf. ↩︎
38. Committee on Finance, op. cit., 87-88. ↩︎
39. Ibid, 82. ↩︎
40. 42 USC  § 1396r-8. ↩︎
41. “Assuring Medicaid Beneficiaries Access to Hepatitis C (HCV) Drugs,” Medicaid Drug Rebate Program Notice, CMS, November 5, 2015, https://www.medicaid.gov/Medicaid-CHIP-Program-Information/By-Topics/Benefits/Prescription-Drugs/Downloads/Rx-Releases/State-Releases/state-rel-172.pdf. ↩︎
42. Micael Ollove, “Are States Obligated to Provide Expensive Hepatitis C Drugs?” Kaiser Health News, February 10, 2016, http://kffhealthnews.org/news/are-states-obligated-to-provide-expensive-hepatitis-c-drugs/. ↩︎
43. Ed Silverman, “Washington State Told to Lift Restrictions on Hepatitis C Medicines,” Stat News, May 27, 2016, https://www.statnews.com/pharmalot/2016/05/27/washington-state-hepatitis-drug-prices/. ↩︎
44. Associated Press, “Florida Changes Hep C Drug Policy For Medicaid After Lawsuit,” June 1, 2016, Tampa Bay Times, http://health.wusf.usf.edu/post/florida-changes-hep-c-drug-policy-medicaid-after-lawsuit#stream/0. Don Sapatkin, “Delaware will treat all Medicaid patients with hepatitis C,” June 9, 2016, http://articles.philly.com/2016-06-09/news/73647247_1_medicaid-patients-hepatitis-c-state-medicaid-program. ↩︎
45. Robert Field, “Regulation of Drugs and Health Care Products” in Health Care Regulation in America: Complexity, Confrontation, and Compromise, (Oxford, England: Oxford University Press, 2007), 115-140. ↩︎
46. Atypical antipsychotics are the second generation of antipsychotic drugs. They came onto the market in 1993, had fewer neurological side-effects than first generation antipsychotics, and were thought to be more effective in some schizophrenia symptoms. See Richard A. Friedman, “A Call for Caution on Antipsychotic Drugs,” New York Times, September 24, 2012, http://www.nytimes.com/2012/09/25/health/a-call-for-caution-in-the-use-of-antipsychotic-drugs.html. ↩︎
47. “Orange Book: Approved Products with Therapeutic Equivalence Evaluations,” FDA, accessed July 6, 2016, http://www.accessdata.fda.gov/scripts/cder/ob/docs/querytn.cfm. ↩︎
48. A boxed warning, commonly known as a “black box warning,” is a FDA warning added to the drug label describing serious or life-threatening risks. See “A Guide to Drug Safety Terms at the FDA,” FDA, accessed July 6, 2016, http://www.fda.gov/downloads/ForConsumers/ConsumerUpdates/UCM107976.pdf. It warns prescribing physicians to seriously consider the benefits of the drug given the known serious risks. Black box warnings can have an effect on the sales of the drug. See Lara Maggs and Aaron Kesselheim, “The Role of Black Box Warnings in Safe Prescribing Practices,” Health Affairs Blog, August 20 2014, http://healthaffairs.org/blog/2014/08/20/the-role-of-black-box-warnings-in-safe-prescribing-practices/. ↩︎
49. Gardiner Harris, “Popular Drugs for Dementia Tied to Deaths,” New York Times, April 12, 2005, http://www.nytimes.com/2005/04/12/health/popular-drugs-for-dementia-tied-to-deaths.html?login=email. ↩︎
50. Friedman, op. cit. ↩︎
51. The FDA approved Abilify for treating children with Tourette syndrome in late 2014 and provided it with an orphan drug designation. Then, despite the sponsor not having applied for it, the FDA broadened Abilify’s designation to also include adults with Tourette syndrome, meaning that the population with the disease was too large for an orphan drug designation. Although Abilify’s manufacturer contested this, saying that the FDA could not provide a drug with a designation for which the sponsor had not applied, they were not successful in their bid. See Ed Silverman, “FDA Approves Generic Abilify After Unusual Legal Battle with Otsuka,” The Wall Street Journal, April 29, 2015, http://blogs.wsj.com/pharmalot/2015/04/29/fda-approves-generic-abilify-after-unusual-legal-battle-with-otsuka-2/. ↩︎
52. The 2014 Drug Trend Report, (Express Scripts, March 2015). ↩︎
53. If the New Drug Application (NDA) is for a drug with an active ingredient that the FDA has previously approved, but this drug is a new dosage form, new indication, or is now available as over-the-counter, the FDA approval grants a 3-year regulatory exclusivity period. If the drug contains chemical entities that have not previously been submitted to the FDA, the FDA generally grants a 5-year regulatory exclusivity period. The FDA provides an additional six months of exclusivity to sponsors who run pediatric studies on the submitted drug. See John Thomas, The Role of Patents and Regulatory Exclusivities in Pharmaceutical Innovation, (Washington DC: Congressional Research Service, January 7, 2013), http://www.ipmall.info/hosted_resources/crs/R42890_130107.pdf. ↩︎
54. To obtain FDA approval for a generic drug, a manufacturer must show that the drug is identical to the brand in “dosage form, safety, strength, route of administration, quality, performance characteristics and intended use.” “What Are Generic Drugs?” FDA, accessed July 6, 2016, http://www.fda.gov/drugs/resourcesforyou/consumers/buyingusingmedicinesafely/understandinggenericdrugs/ucm144456.htm. ↩︎
55. Ibid. ↩︎
56. The Drug Price Competition and Patent Term Restoration Act of 1984, usually referred to as the Hatch-Waxman Act, created this framework of bringing generic drugs to market by allowing sponsors to file Abbreviated New Drug Applications (ANDAs) where they proved bioequivalence. At the same time, the Hatch-Waxman Act incentivized innovation by rewarding sponsors of new molecular entities and other brand drugs with varying lengths of regulatory exclusivity. ↩︎
57. We considered a drug a brand-name drug if it was single-source, or if it was the originator product, but is now available from multiple sources. ↩︎
58. “Frequently Asked Questions About Therapeutic Biological Products,” FDA, accessed July 6, 2016, http://www.fda.gov/Drugs/DevelopmentApprovalProcess/HowDrugsareDevelopedandApproved/ ApprovalApplications/TherapeuticBiologicApplications/ucm113522.htm. ↩︎
59. A manufacturer obtains FDA approval for a biologic through a Biologic License Approval (BLA) as opposed to an NDA or an ANDA. ↩︎
60. Specifically, the part of the ACA that creates a framework for the approval of biosimilars and regulatory exclusivity to the original biologic is referred to as the Biologics Price Competition and Innovation Act. Biosimilars are not bioequivalent to the previously approved FDA biologic, in contrast to generic drugs which are bioequivalent to previously approved small-molecule drugs. Instead, biosimilars must be highly similar and interchangeable with the previously approved biologics. ↩︎
61. It is possible that this may change in the future if the Trans-Pacific Partnership passes, as one point of negotiations between countries was the period of regulatory exclusivity for biologics. ↩︎
62. Andrew Mulcahy, Zachary Predmore, and Soeren Mattke, “The Cost Savings Potential of Biosimilar Drugs in the United States,” (Rand Corporation, 2014), https://www.rand.org/content/dam/rand/pubs/perspectives/PE100/PE127/RAND_PE127.pdf. ↩︎
63. Zarxio is a biosimilar to Neupogen, a biologic used to stimulate white blood cell growth in the treatment of cancer, bone marrow transplants, and chemotherapy. (“Neupogen,” Drugs.com, accessed July 6, 2016, http://www.drugs.com/neupogen.html.) However, Neupogen was not one of the 50 most costly drugs, nor was it one of the 50 most expensive drugs per prescription (Appendix Table A2). ↩︎
64. Congress passed the Orphan Drug Act in 1982 to provide incentive for manufacturers to produce drugs for diseases rare enough that there had not previously been financial incentive to do so. Congress amended the law three times in the 80s, eventually arriving at these parameters for an orphan drug designation. See Gary Pulsinelli, “The Orphan Drug Act: What’s Right with It,” Santa Clara High Technology Law Journal, 5, 2. (January 1999): http://digitalcommons.law.scu.edu/cgi/viewcontent.cgi?article=1247&context=chtlj and “Developing Products for Rare Diseases & Conditions,” FDA, accessed July 6, 2016, http://www.fda.gov/ForIndustry/DevelopingProductsforRareDiseasesConditions/ucm2005525.htm. ↩︎
65. These benefits also include federal funding for clinical trials and tax breaks, as well as other advantages. See Aaron Kesselheim, “Innovation and the Orphan Drug Act, 1983-2009: Regulatory and Clinical Characteristics of Approved Orphan Drugs” in Rare Diseases and Orphan Products: Accelerating Research and Development, (Washington DC: Institute of Medicine, 2010), 291-308, http://www.ncbi.nlm.nih.gov/books/NBK56187/. ↩︎
66. Since 1983, the FDA has provided orphan drug designations to over 400 drugs that have gone to market. Between 1973 and 1983, there were fewer than 10. See “Developing Products for Rare Diseases & Conditions,” FDA, accessed July 6, 2016, http://www.fda.gov/ForIndustry/DevelopingProductsforRareDiseasesConditions/ucm2005525.htm. ↩︎
67. Marty Makary, “One Pharma Fix: Limit the ‘Orphan Drug’ Incentives,” Wall Street Journal, Dec. 20, 2015, http://www.wsj.com/articles/one-pharma-fix-limit-the-orphan-drug-incentives-1450645511. ↩︎
68. Bruen and Young, December 2014, op. cit. ↩︎
69. Smith, Gifford, Ellis, Rudowitz, Snyder, and Hinton, op. cit. ↩︎
70. Steve Miller, “Harvoni: Orphan-Drug Pricing for a Non-Orphan Drug,” (Express Scripts, October 2014), http://lab.express-scripts.com/lab/insights/specialty-medications/harvoni-orphan-drug-pricing-for-a-nonorphan-drug. ↩︎
71. Vernon K. Smith, Kathleen Gifford, Eileen Ellis, Robin Rudowitz, and Laura Snyder, Medicaid in an Era of Health & Delivery System Reform: Results from a 50-State Medicaid Budget Survey for State Fiscal Years 2014 and 2015, (Washington DC, Kaiser Commission on Medicaid and the Uninsured, October 2014), https://modern.kff.org/medicaid/report/medicaid-in-an-era-of-health-delivery-system-reform-results-from-a-50-state-medicaid-budget-survey-for-state-fiscal-years-2014-and-2015/. ↩︎
72. B.E. and A.R. v. Teeter, op. cit. ↩︎
73. “Getting a Liver Transplant,” National Kidney Foundation, accessed July 6, 2016, https://www.kidney.org/atoz/content/livertx. ↩︎
74. Committee on Finance, op. cit., 1. ↩︎
75. Matt Salo, executive director of the National Association of State Medicaid Directors, conveyed the dilemma saying “We would be spending more on this one drug than all other drugs combined […] There isn’t the capacity to do that.” See Michael Ollove, “Are States Obligated to Provide Expensive Hepatitis C Drugs?” (Pew Trusts, February 2016), http://www.pewtrusts.org/en/research-and-analysis/blogs/stateline/2016/02/09/are-states-obligated-to-provide-expensive-hepatitis-c-drugs. ↩︎
76. “Assuring Medicaid Beneficiaries Access to Hepatitis C (HCV) Drugs,” op. cit. ↩︎
77. 42 USC § 1396r-8(d). States may restrict certain drugs categorically, but hepatitis C agents are not on this list. Statute allows states to restrict weight loss or weight gain drugs; fertility drugs; cosmetic or hair growth drugs; cold or cough relief drugs; smoking cessation drugs; vitamins and minerals, except for prenatal use; non-prescription drugs, except for pregnant women; drugs where along with them, the manufacturer requires certain tests or services be provided by a designee; barbituates; benzodiazepines; and drugs for sexual or erectile dysfunction, unless these drugs are used for another approved condition. ↩︎
78. B.E. and A.R. v. Teeter, op. cit. and CMS, “Assuring Medicaid Beneficiaries Access to Hepatitis C (HCV) Drugs,” op. cit. ↩︎
79. Methadone is a type of opioid. It is used for the treatment of opioid use disorder, and more recently, for the treatment of pain. However, it is a “complex medication to prescribe for pain relief,” and as a treatment for pain is associated with a disproportionate share of opioid-related deaths. See Wachino, 2016, op. cit. ↩︎
80. Drug Utilization Reviews are a mechanism for state Medicaid programs to monitor beneficiaries’ prescriptions to prevent potential oversights that might harm beneficiaries, such as drug-disease contraindications or incorrect dosages. They are also a way for state Medicaid programs to identify overuse, abuse, or fraud within the prescription drug benefit. See “Drug Utilization Review,” CMS, accessed July 6, 2016, https://www.medicaid.gov/medicaid-chip-program-information/by-topics/benefits/prescription-drugs/drug-utilization-review.html. ↩︎
81. Wachino, 2016, op. cit. CMCS also outlines ways to increase the availability and administration of naloxone, a drug used to counteract opioid overdoses, as well as ways to expand treatments for opioid use disorder. ↩︎
82. Barry Meier and Sabrina Tavernise, “States Move to Control How Painkillers Are Prescribed,” New York Times, March 11, 2016, http://www.nytimes.com/2016/03/12/business/states-move-to-control-how-painkillers-are-prescribed.html ↩︎
83. Ibid. ↩︎
84. HHS, HHS awards $94 million to health centers to help treat the prescription opioid abuse and heroin epidemic in America, March 11, 2016, http://www.hhs.gov/about/news/2016/03/11/hhs-awards-94-million-to-health-centers.html. ↩︎
85. “CDC Guideline for Prescribing Opioids for Chronic Pain.” CDC, accessed July 6, 2016, http://www.cdc.gov/drugoverdose/prescribing/guideline.html. ↩︎
86. Vikki Wachino, “2016 Updates to the Child and Adult Core Health Care Quality Measurement Sets,” CMCS Informational Bulletin, CMS, December 11, 2015. https://www.medicaid.gov/federal-policy-guidance/downloads/cib-12-11-15.pdf. ↩︎
87. “Drug Advertising: A Glossary of Terms,” FDA, accessed July 6, 2016, http://www.fda.gov/Drugs/ResourcesForYou/Consumers/PrescriptionDrugAdvertising/ucm072025.htm. ↩︎
88. Wachino, 2015, op. cit. ↩︎
89. Copyright 2016, Wolters Kluwer Clinical Drug Information, Inc. ↩︎
90. “CDER List of Licensed Biological Products,” FDA, Center for Drug Evaluation and Research, accessed July 6, 2016, http://www.fda.gov/Drugs/DevelopmentApprovalProcess/ HowDrugsareDevelopedandApproved/ApprovalApplications/TherapeuticBiologicApplications/Biosimilars/ucm411418.htm “CBER List of Licensed Biological Products” FDA, Center for Biologic Evaluation and Research,  accessed July 6, 2016, http://www.fda.gov/Drugs/DevelopmentApprovalProcess/ HowDrugsareDevelopedandApproved/ApprovalApplications/TherapeuticBiologicApplications/Biosimilars/ucm411418.htm. ↩︎
91. Searchable version of the Orphan Drug Product designation database, FDA, http://www.accessdata.fda.gov/scripts/opdlisting/oopd/. ↩︎
92. We pulled formulary information from CVS CareMark, Express Scripts, and Optum Rx. ↩︎
93. “States Collection of Medicaid Rebates for Physician Administered Drugs,” (Washington DC: U.S. Department of Health and Human Services- Office of Inspector General, June 2011), http://oig.hhs.gov/oei/reports/oei-03-09-00410.pdf. For further discussion, see Bruen and Young, December 2014, op cit.     ↩︎

KEY FINDINGS:

• In a campaign that so far has not been about policy issues, equal shares of voters name terrorism and national security, along with a candidate’s personal characteristics like leadership ability, character, values, and experience, as top voting considerations for the 2016 presidential election, with health care ranking further down the list with a cluster of other issues.
• While overall attitudes of the ACA lean unfavorable, a larger share of voters say they trust the Democratic Party than say they trust the Republican Party to do a better job dealing with health care issues including determining the future of the ACA, Medicare, lowering prescription drug costs, and dealing with the opioid epidemic, with the widest issue advantage for either political party being women’s reproductive health.
• Overall, the survey finds more voters say the Democratic Party (47 percent) than the Republican Party (35 percent) best represents their health care views. When asked which of the major party’s presumptive presidential nominees best represents their health care views, more voters choose Hillary Clinton (46 percent) than Donald Trump (32 percent).

Election Watch: The Role of Health Care Issues in the 2016 Presidential Election

The presidential campaign is in full swing with candidates and political parties negotiating party platforms, frequent murmurings about possible vice presidential candidates, and party conventions starting in days. As part of Kaiser’s ongoing analysis of the role of health care issues in the political debate, the July Kaiser Health Tracking Poll examines the role that health care may play in the 2016 presidential election, how important health care is to voters, what health care issues voters would most like to hear the candidates discuss, and which party and candidates voters feel most closely aligned with on health care issues.

Terrorism is Top Issue For Voters, Large Shares Also Say Candidate’s Personal Characteristics Are Important

Terrorism and national security is the top issue for registered voters in the 2016 presidential election, according to the most recent tracking survey.¹  Nearly half (46 percent) of all voters say terrorism and national security is “extremely” important to their vote for president. About four in ten voters (37 percent) say health care is “extremely” important, which is similar to the share of voters who say the same about gun policy (38 percent), education (35 percent), and the economy and jobs (37 percent). Smaller shares of voters in the presidential election say foreign policy, the federal budget deficit, immigration, and taxes are “extremely” important to their vote.

While a large share of voters say specific issues will be important to their vote for president, previous research demonstrates that voters also make their decision to support candidates based on other factors including their perceptions of the personal characteristics of the candidate – including their leadership ability, character, values, and experience. Slightly less than half (46 percent) of voters say a candidate’s personal characteristic are “extremely” important to their vote for president, which is the same as the share who say terrorism and national security is important. Unlike policy issues cited by voters, the share naming personal characteristics as important to their vote does not vary by party affiliation.

Across the factors that are important to voters, there are noticeable party differences. Nearly two-thirds (63 percent) of voters who identify as Republicans say terrorism and national security is “extremely” important to their vote compared to 40 percent of voters who identify as Democrats and 43 percent of independents. The top issues for Democratic voters are education (46 percent) and health care (42 percent). Among independent voters, terrorism and national security is the top issue (43 percent) but a larger share of independents say a candidate’s personal characteristics is important (46 percent) than who say the same about any other substantive policy issues.

Partisanship and Health Care

The July Tracking Survey also examines how voters view each political party and their prospective presidential candidates on health care issues.

Overall Party Differences on Health Care Issues

When it comes to perceptions of whether the two political parties differ on health care issues, 65 percent of voters say there are major differences between the parties with an additional 21 percent saying there are minor differences. Overall, majorities across political parties say there are major differences between the Republican Party and the Democratic Party on health care issues.

Voters Trust Their Own Political Party on Health Care Issues

In general, 85 percent of Republican voters and 93 percent of Democratic voters say their own political party best represents their views on health care. Independents are much more closely divided with four in ten (39 percent) saying the Democratic Party best represents their health care views and one-third (32 percent) saying the Republican Party represents their views. In addition, one in five independents (21 percent) say that neither political party represents their own views on health care.

When asked which political party (the Republican Party or the Democratic Party) they trust to do a better job dealing with a series of health care issues, a larger share of voters say they trust the Democratic Party to do a better job than the Republican Party.

The widest issue advantage for either political party was women’s reproductive health, with voters saying at a margin of 2 to 1 that they trust the Democratic Party to do a better job of making decisions about women’s reproductive health choices and services than the Republican Party (58 percent compared to 30 percent). The Democratic Party also holds a significant advantage over the Republican Party when it comes to lowering the cost of prescription drugs (50 percent and 32 percent). Larger shares also say they trust the Democratic Party rather than the Republican Party on Zika funding (45 percent vs. 31 percent) and the opioid epidemic in the U.S. (41 percent vs. 34 percent).

Higher shares of female voters and younger voters say they trust the Democratic Party than say they trust the Republican Party on all of these health issues, with the largest differences found on women’s health issues and lowering prescription drug costs. On these two issues, more than half of female voters and voters between the ages of 18 and 29 say they trust the Democratic Party compared to less than one-third of each group who say they trust the Republican Party. Among voters 65 and older, similar shares say they trust both parties on most health issues. For example, when it comes to determining the future of the Medicare program, similar shares of senior voters say they trust the Democratic Party (45 percent) and the Republican Party (41 percent).

Across all issues, vast majorities of both Republican voters and Democratic voters say they trust their own political party to do a better job handling each health care issue and very few say they trust the opposing party. For example, nine out of ten Democratic voters (91 percent) say they trust the Democratic Party to do a better job determining the future role of government in health care, while a similar share of Republican voters (87 percent) say they trust the Republican Party to do a better job.

With both Republican voters and Democratic voters saying they trust their own political party to do a better job handling each of the health care issues, independent voters’ views play an important role. Overall, larger shares of independent voters trust the Democratic Party than the Republican Party to do a better job of making decisions about women’s reproductive health (60 percent compared to 25 percent), lowering prescription drug costs (46 percent vs. 28 percent), and funding efforts to limit the effects of the Zika virus outbreak (42 percent vs. 24 percent). Independents are more split on whether they trust the Democratic Party or the Republican Party on the other issues, with similar shares saying they trust the Democratic Party and the Republican Party on dealing with the future of the 2010 health care law, lowering health care costs for people like them, and dealing with the opioid epidemic.

Views of the 2016 Presidential Candidates On Health Care

Majorities of Voters Say Their Party’s Candidate Represents Their Views on Health Care

When asked which presidential candidate best represents their own views on health care, nearly half of voters (46 percent) choose Hillary Clinton while about one-third (32 percent) say Donald Trump best represents their views. As expected, a vast majority of Democratic voters (87 percent) say Hillary Clinton best represents their own views on health care, while only 3 percent say Donald Trump. On the other side of the political spectrum, three-fourths of Republican voters (74 percent) say Donald Trump best represents their own views on health care, while one in ten (11 percent) say Hillary Clinton and 13 percent say neither candidate represents their views. More independent voters say Hillary Clinton best presents their health care views than Donald Trump — 39 percent compared to 30 percent – yet one in four (23 percent) say neither candidate represents theirs health care views.

Large Shares of Voters Say Candidates and Political Parties Are Paying too Little Attention to Health Care Issues

With less than one week before the start of the 2016 presidential conventions,²  the majority of American voters say they are interested in what happens during the Republican National Convention (71 percent) and the Democratic National Convention (65 percent), and more than half of individuals from both parties say they care “a lot” about what is included in their party’s platform.

And while the political parties and their presumptive candidates, Donald Trump and Hillary Clinton, are working on defining their party’s 2016 platforms, slightly more than half of voters say the Republican Party and Donald Trump are paying too little attention to health issues. More than half of voters (56 percent) say Donald Trump is paying too little attention and 55 percent say the Republican Party is paying too little attention to health care issues. In contrast, about one-third of voters say Hillary Clinton (35 percent) and the Democratic Party (33 percent) is paying too little attention to health care issues, while about four in ten say the Democratic Party and Hillary Clinton are paying the right amount of attention.

There are distinct partisan differences in these findings: larger shares of Republican voters say the Democratic Party and Hillary Clinton are paying too little attention, 43 percent and 44 percent, while large majorities of Democratic voters say the Republican Party and Donald Trump are paying too little attention, 71 percent and 75 percent. About half of independent voters say Donald Trump and the Republican Party are paying too little attention to health care issues, 54 percent each, compared to about four in ten who say the same about Hillary Clinton (40 percent) and the Democratic Party (37 percent).

Republicans Want to Hear Candidates Discuss ACA, Democrats Want to Hear About Health Care Costs

When asked specifically what health issues they want to hear the presidential candidates talk about, health care costs are mentioned more often than any other health issue (38 percent), followed by increasing access for uninsured (28 percent), the ACA (27 percent), and Medicare (10 percent).

Four in ten Republican voters (41 percent) say they want to hear the 2016 presidential candidates talk about the ACA (of which 26 percent say they want to hear the candidates talk about repealing or opposing the law), compared to one-fourth of independent voters (26 percent) who want to hear the candidates discuss the ACA and only 15 percent of Democratic voters. By comparison, four in ten Democratic voters say they want the candidates to talk about health care costs (41 percent) and expanding health insurance coverage for the currently uninsured (40 percent). A larger share of independent voters (39 percent) say they want candidates to discuss health care costs than say they want the candidates to talk about expanding access to the uninsured (27 percent) or the ACA (26 percent).

While health care is currently not the top issue for voters, majorities of Republicans, Democrats, and independents say it is “extremely” or “very” important to their presidential vote choice (71 percent, 86 percent, and 75 percent, respectively). Among the individuals, regardless of party, who say health care is important to their vote for president this year, 38 percent say health care costs is the health issue they most want to hear the presidential candidates discuss, followed by increasing access to the uninsured (30 percent) and the ACA (27 percent). Similar to the partisan differences found among all voters, a higher share of Democratic voters who say health care is “extremely” or “very” important say they want to hear the candidates discuss health care costs more than any other issue; Republican health care voters want to hear the candidates discuss the ACA.

Americans’ Opinions of the Affordable Care Act

While the ACA remains a top health care issue for Republican voters, overall attitudes toward the law remained largely unchanged over the past month. The most recent survey finds 40 percent of Americans reporting favorable opinions of the health care law and 46 percent reporting unfavorable opinions.

Kaiser Health Policy News Index: July 2016

This month’s Health Policy News Index finds a large majority (83 percent) of the American public reports closely following news about the 2016 presidential campaigns, up slightly from 79 percent in June. This is similar to the share who say they have closely followed news this month about recent fatal police shootings of black men like Alton Sterling and Philando Castile³  (81 percent) and the share who report closely following news about terrorist attacks in Istanbul and Baghdad and other conflicts involving ISIS or other militant groups (78 percent). Congressional debate about gun control legislation also garnered significant attention in July, with nearly three-fourths of Americans (72 percent) saying they have followed news about it closely, while Britain’s vote to leave the European Union – Brexit – was closely followed by a smaller share (53 percent). Fewer Americans report following this month’s top health stories, with about half closely following the Zika virus outbreak and news about the ongoing opioid epidemic in the nation (52 percent each) and 46 percent closely following the Supreme Court decision regarding Texas abortion clinics.

The Zika Virus Outbreak

The Kaiser Tracking Poll has been tracking the public’s awareness of the Zika virus since February 2016. The most recent survey finds that the vast majority of Americans (89 percent) have heard or read at least a little about the Zika virus, up slightly from June 2016.⁴  While the majority of those who have heard or read about the virus are not worried that they or someone in their family will be affected by the virus, six in ten are worried the U.S. will see a large number of cases of the virus (59 percent) and the upcoming summer Olympics will cause an increase in the number of cases (59 percent).

The Ongoing Opioid Epidemic

On July 8^th, the U.S. House of Representatives passed the Comprehensive Addiction and Recovery Act (CARA). This bill, which is also supported by the Senate and President Obama, is part of the government’s effort to combat the ongoing opioid epidemic in the U.S. Similar to results from the April,⁵  the June Tracking Poll finds that slightly more than four in ten Americans (43 percent) say they know someone who has been addicted to prescription painkillers, including one in five (19 percent) who say they have a family member who has been addicted.

This Kaiser Health Tracking Poll was designed and analyzed by public opinion researchers at the Kaiser Family Foundation (KFF). The survey was conducted July 5-11, 2016, among a nationally representative random digit dial telephone sample of 1,212 adults ages 18 and older, living in the United States, including Alaska and Hawaii (note: persons without a telephone could not be included in the random selection process). Computer-assisted telephone interviews conducted by landline (426) and cell phone (786, including 477 who had no landline telephone) were carried out in English and Spanish by Princeton Data Source under the direction of Princeton Survey Research Associates International (PSRAI). Both the random digit dial landline and cell phone samples were provided by Survey Sampling International, LLC. For the landline sample, respondents were selected by asking for the youngest adult male or female currently at home based on a random rotation. If no one of that gender was available, interviewers asked to speak with the youngest adult of the opposite gender. For the cell phone sample, interviews were conducted with the adult who answered the phone. KFF paid for all costs associated with the survey.

The combined landline and cell phone sample was weighted to balance the sample demographics to match estimates for the national population using data from the Census Bureau’s 2014 American Community Survey (ACS) on sex, age, education, race, Hispanic origin, and region along with data from the 2010 Census on population density. The sample was also weighted to match current patterns of telephone use using data from the January-June 2015 National Health Interview Survey. The weight takes into account the fact that respondents with both a landline and cell phone have a higher probability of selection in the combined sample and also adjusts for the household size for the landline sample. All statistical tests of significance account for the effect of weighting.

The margin of sampling error including the design effect for the full sample is plus or minus 3 percentage points. Numbers of respondents and margins of sampling error for key subgroups are shown in the table below. For results based on other subgroups, the margin of sampling error may be higher. Sample sizes and margins of sampling error for other subgroups are available by request. Note that sampling error is only one of many potential sources of error in this or any other public opinion poll. Kaiser Family Foundation public opinion and survey research is a charter member of the Transparency Initiative of the American Association for Public Opinion Research.

1. The July Kaiser Tracking Poll was in the field from July 5th to July 11th, one week after the terrorist attacks at Istanbul’s airport and two days following the attacks in Baghdad which killed more than 200 civilians. During this week, there were also several domestic incidents, including police-involved shootings in Baton Rouge and Minnesota and the attack targeting police officers in Dallas on July 7th. ↩︎
2. The Republican National Convention starts on July 18th, 2016 and the Democratic National Convention starts on July 25th, 2016. ↩︎
3. The police-involved shooting of Alton Sterling in Baton Rouge, La. occurred on July 5th and the police-involved shooting of Philando Castille in Minnesota occurred on July 6th. The items were included in interviews beginning on July 7th. ↩︎
4. A Kirzinger, E Sugarman, M Brodie, Kaiser Health Tracking Poll: June 2016.  https://modern.kff.org/global-health-policy/poll-finding/kaiser-health-tracking-poll-june-2016/ ↩︎
5. J Firth, A Kirzinger, M Brodie, Kaiser Health Tracking Poll: April 2016. https://modern.kff.org/health-reform/report/kaiser-health-tracking-poll-april-2016/ ↩︎

On May 18, 2016, the Department of Health and Human Services (HHS) published a final rule to implement Section 1557 of the Affordable Care Act (ACA), which prohibits discrimination in health coverage and care based on race, color, national origin, age, disability, and sex. These provisions incorporate existing federal non-discrimination law and policy and also contain some new protections.  Key provisions include:

• Extending protections against sex discrimination to health coverage and care for the first time and including gender identity discrimination within the definition of sex discrimination;
• Codifying long-standing guidance regarding meaningful access for individuals with Limited English Proficiency, including the provision of free, accurate, and timely language assistance services;  
• Incorporating existing law that requires reasonable modifications, effective communication, and readily accessible buildings and information technology to avoid disability-based discrimination; and
• Prohibiting discriminatory health insurance benefit designs and including specific coverage protections for transgender individuals.

The final rule does not:

• Resolve whether Section 1557’s prohibition on sex discrimination extends to discrimination based on sexual orientation alone, although HHS notes that such discrimination is prohibited under Section 1557 when it is based on gender stereotypes; or
• Set specific accessibility standards for medical equipment for people with disabilities, pending action by the U.S. Access Board.

HHS emphasizes Section 1557’s importance in achieving the ACA’s goals of expanding access to health care and insurance, noting that discrimination within these areas can contribute to poor health outcomes or coverage, increase health disparities among underserved communities, and negatively impact the distribution of health care resources. The final rule is effective on July 18, 2016, except for provisions that require changes to health insurance plan benefit design, which will take effect in the first plan year beginning on or after January 1, 2017.

Introduction

On May 18, 2016, the Department of Health and Human Services (HHS) finalized regulations to implement Section 1557 of the Affordable Care Act (ACA).¹  Section 1557 prohibits certain entities that administer health programs and activities from excluding an individual from participation, denying program benefits, or discriminating based on race, color, national origin, sex, age or disability.²  While Section 1557 has been in effect since the ACA’s enactment, these regulations mark the first time HHS has issued implementing guidance. In the preamble to the final rule, HHS emphasizes Section 1557’s importance in achieving the ACA’s overarching goal of expanding access to health care and insurance for all individuals, noting that discrimination within health programs can contribute to poor and inadequate health outcomes or coverage, exacerbate existing health disparities in underserved communities, and lead to insufficient and ineffective distribution of health care resources.³ 

Section 1557 seeks to coordinate existing federal non-discrimination laws, regulations, and policy as they apply to health coverage and care and to extend sex discrimination protections to health programs and activities. Specifically, the final rule incorporates the existing anti-discrimination tenants of Title VI (race, color and national origin), the Age Discrimination Act, and Section 504 (disability) as they apply to health care. It also extends the sex discrimination protections of Title IX, which only apply to educational programs, to health care. As such, it is the first federal civil rights law to prohibit discrimination on the basis of sex in health care. Moreover, the final rule defines sex discrimination to include discrimination on the basis of gender identity. In addition, the final rule codifies long-standing HHS policy guidance about the provision of language assistance services. The final rule is effective on July 18, 2016, except for provisions that require changes to health insurance benefit design, which will take effect in the first plan year beginning on or after January 1, 2017.⁴   This issue brief summarizes key provisions of the final rule.

Key Provisions of the Section 1557 Final Rule

Entities Subject to Section 1557’s Non-Discrimination Provisions

The final rule applies to health programs and activities that receive Federal financial assistance from HHS; health programs and activities administered by HHS, including the Federally-facilitated Marketplace; and State-based Marketplaces established under the ACA.⁵  Box 1 provides examples of the types of entities subject to Section 1557. While Section 1557 applies more broadly to all health programs and activities that receive federal funding through any agency, HHS limits its final rule to the programs that it has authority to enforce.⁶  HHS encourages other federal agencies to adopt the standards from its rule when applying Section 1557 to the health programs and activities that they administer.⁷ 

The final rule defines health programs and activities to include all operations of an entity that is principally engaged in the provision or administration of health-related services or health-related insurance coverage.⁸  Health programs and activities also include providing assistance obtaining health-related services or health-related insurance coverage.⁹  HHS interprets health programs and activities to include health education and research programs and reiterates that federal law already prohibits discrimination in federally funded research as well as in research at universities. The final rule extends non-discrimination protections to any research conducted within HHS as well as non-educational institutions. In the preamble, HHS recognizes that research projects are often limited in scope, and therefore, research protocol criteria that target or exclude certain groups are warranted when justified for subjects’ health or safety, the scientific study design, or the research purpose.¹⁰  The rule also applies to employee health benefit programs in limited circumstances.¹¹ 

The final rule clarifies that Federal financial assistance includes not only funding received directly by covered entities but also premium and cost-sharing subsidies provided to individuals for coverage through the Federally-facilitated or State-based Marketplaces.¹²  As such, the rule applies to insurance carriers that offer plans through these Marketplaces. Moreover, the preamble clarifies that Section 1557 applies to all plans offered by these carriers, not just those available through the Marketplaces.¹³ 

Protections Against Discrimination Based on Sex

Section 1557 and HHS’s final regulations, for the first time, extend protections against discrimination based on sex to health coverage and care. Covered entities must provide individuals equal access to health programs and activities without discrimination based on sex, including pregnancy, false pregnancy, termination of pregnancy, recovery from childbirth or related medical conditions. Sex discrimination also includes sex stereotypes, including expectations about gender roles, which could be an important protection for lesbian, gay, bisexual, and transgender individuals, among others.¹⁴ 

Moreover, the final rule extends the definition of sex discrimination to include discrimination based on gender identity for the first time in the health care context.¹⁵  HHS clarifies that its inclusion of gender identity as part of sex discrimination includes gender expression and “transgender status”¹⁶  and encompasses individuals with non-binary gender identities.¹⁷ 

In the preamble to the final rule, HHS specifies that Section 1557’s prohibition on sex discrimination includes discrimination related to sexual orientation where the evidence establishes that the discrimination is based on gender stereotypes.¹⁸  The final rule does not resolve whether Section 1557’s prohibition on sex discrimination extends to discrimination based on an individual’s “sexual orientation status” alone.¹⁹  HHS notes that the law in this area is continuing to evolve and will consider issuing further guidance.²⁰ 

Sex-specific health programs and activities are prohibited unless a covered entity can show that the sex-based classification is substantially related to the achievement of an important health-related or scientific objective.²¹  HHS expects that this standard will enable most health researchers to justify sex-specific clinical trials, such as those that test treatments for sex-specific conditions or that evaluate differences in responses to treatment regimens among the sexes.²²  However, the rule does not offer a blanket allowance for all sex-specific research. If there is no scientific or clinical reason for a study to be restricted to one sex researchers would need to justify the reasons for this limitation.

The final rule does not include a blanket religious exemption related to Section 1557’s general prohibition against sex discrimination. Existing laws protecting religious freedom and belief, including provider conscience laws, the Religious Freedom Restoration Act, the ACA’s provisions regarding abortion services, and the ACA’s preventive health services regulations, continue to apply.²³  The final rule clarifies that compliance with Section 1557 is not required if it would violate these laws.²⁴ 

Prohibition Against Discrimination Based on National Origin

Section 1557 prohibits discrimination based on an individual’s place of origin, his or her ancestor’s place of origin, and his or her manifestation of the physical, cultural, or linguistic characteristics of a national origin group.²⁵  The rule adopts the Equal Employment Opportunity Commission’s definition of national origin. While Section 1557 does not explicitly ban discrimination based on citizenship or immigration status, as opposed to national origin, HHS notes that Section 1557 does prohibit policies or practices related to citizenship or immigration status that disproportionately affect  individuals of a particular national origin group even if those policies or practices are not explicitly discriminatory.²⁶ 

Meaningful Access for Individuals with Limited English Proficiency (LEP)

Covered entities must take reasonable steps to provide meaningful access to each individual with LEP who is eligible to be served or likely to be encountered in their health programs and activities.²⁷  The final rule incorporates existing Title VI regulations and HHS’s LEP policy guidance and outlines specific requirements related to provision of language assistance services. When assessing an entity’s compliance with this provision, the HHS Office of Civil Rights (OCR) will take into account and give substantial weight to the nature and importance of the health program or activity and the communication at issue to the person with LEP.²⁸  OCR also will consider other relevant factors, including whether the entity has developed and implemented an effective written language access plan.²⁹  HHS did not establish a minimum threshold for non-English languages in which entities must provide language assistance, and it did not require entities to have a language access plan in the final rule. However, as indicated, the creation of a written language access plan is one factor OCR will consider in evaluating compliance.

Covered entities must provide free, accurate, and timely language assistance services and protect the privacy and independence of an individual with LEP. Covered entities must offer a qualified interpreter when reasonable for oral communication and use a qualified translator for written content.³⁰  Moreover, the entity can only rely on qualified bilingual/multilingual staff to communicate directly with individuals with LEP.³¹  The rule defines the knowledge, skills, and abilities of a qualified interpreter and a qualified translator who may provide these services as well as for qualified bilingual and multilingual staff.³²  The rule also adds performance standards for video remote interpreting services.³³  The rule specifies that individuals with LEP are not required to accept language assistance services,³⁴  although the preamble clarifies that a provider may use a qualified interpreter to assist the provider in communicating with an individual with LEP who has refused language assistance services.³⁵ 

An entity may not require an individual with LEP to provide his or her own interpreter.³⁶  Moreover, the entity cannot rely on an adult accompanying the individual with LEP to interpret, except in emergency situations or when the individual specifically requests the adult to provide those services, the adult agrees to provide the assistance, and reliance on the adult is appropriate for the circumstances.³⁷  Similarly, the entity may not rely on a minor child to interpret except in an emergency when no qualified interpreter is immediately available.³⁸ 

Protections Against Discrimination Based on Disability

Covered entities must make reasonable modifications in policies, practices, and procedures to avoid disability-based discrimination, unless doing so would fundamentally alter the nature of the health program or activity.³⁹  The final rule incorporates existing federal protections against disability-based discrimination from Section 504 of the Rehabilitation Act and the Americans with Disabilities Act (ADA).⁴⁰ 

Covered entities must take appropriate steps to ensure effective communication with people with disabilities.⁴¹  To the extent that the ADA effective communication rules differ between Title II (which applies to state and local government) and Title III (which applies to places of public accommodation), HHS adopted the Title II rules for all entities subject to Section 1557. This requires Section 1557 covered entities to give “primary consideration” to the person with a disability’s choice of auxiliary aid or service.⁴²  Auxiliary aids and services can include, as appropriate, qualified interpreters, a variety of assistive technology devices, and the provision of materials in alternative formats.⁴³  The final rule clarifies that all covered entities, regardless of the number of employees, must provide appropriate auxiliary aids and services to people with impaired sensory, manual, or speaking skills, where necessary to afford those individuals an equal opportunity to benefit from the health program or activity.⁴⁴ 

Buildings and facilities where health programs and activities are conducted must be constructed and altered in a way that is readily accessible to and useable by people with disabilities. Specifically, HHS generally adopts the 2010 ADA Standards for Accessible Design for new construction or alternations of facilities of covered entities that receive Federal funding and State-based Marketplaces.⁴⁵ 

While Section 1557’s prohibition against disability-based discrimination in health programs and activities extends to medical equipment, HHS does not set specific accessibility standards in this area. The U.S. Access Board is still developing its standards for accessible medical equipment. In the meantime, HHS notes that providers must ensure that health programs and activities that involve the use of medical equipment are accessible to people with disabilities.⁴⁶ 

Covered entities must make electronic and information technology in health programs and activities accessible to people with disabilities, unless doing so would create an undue financial or administrative burden or fundamentally alter the nature of the health program or activity.⁴⁷  When providing an accessible electronic format creates an undue burden or fundamental alteration, covered entities still must provide information in another format that does not create an undue burden or fundamental alternation but that does ensure, to the maximum extent possible, that people with disabilities receive the benefits or services of the health program or activity provided through electronic or information technology.⁴⁸ 

Prohibitions Against Discrimination in Health Coverage

The final rule outlines several protections against discrimination in health insurance coverage determinations and administration. Under Section 1557, covered entities may not take the following actions on the basis of race, color, national origin, sex, age or disability:

• Deny, cancel, limit, or refuse to issue or renew a health insurance policy;⁴⁹ 
• Deny or limit coverage of a health insurance claim;⁵⁰ 
• Impose additional cost sharing or other limitations or restrictions on coverage;⁵¹  or
• Use discriminatory marketing practices or insurance benefit designs.⁵² 

The final rule maintains that while health insurers cannot have coverage policies that operate in a discriminatory manner, they still may apply medical necessity rules when determining covered benefits.⁵³  HHS also notes in the preamble that Section 1557 does not require insurers to cover any particular services.⁵⁴  Provisions of the final rule that require changes in plan benefit design take effect in the first plan year on or after January 1, 2017. Such changes include covered benefits, benefits limitations or restrictions, and cost-sharing mechanisms such as coinsurance, copayments, and deductibles.⁵⁵ 

While the final rule states that covered entities may not employ discriminatory benefit designs, it is silent on whether certain practices are discriminatory.  For example, placing all drugs to treat a single medical condition on the plan’s highest cost-sharing tier has been challenged in an administrative complaint filed with the OCR (Box 2). In separate guidance, HHS has identified this practice as an example of potential discrimination, along with applying age limits to services that have been found clinically effective at all ages, and requiring prior authorization and/or step therapy for all or most medications in drug classes such as anti-HIV protease inhibitors and/or immune suppressants regardless of medical evidence.⁵⁶ 

The final rule includes specific coverage protections for transgender individuals. While the rule generally requires that covered entities treat individuals consistent with their gender identity, as described above, it specifies a limited exception. The exception prohibits entities from denying or limiting health services (or imposing additional cost-sharing on services) that are ordinarily or exclusively available to individuals of one sex or gender based on the fact that the individual’s sex assigned at birth, gender identity, or gender in a medical or health insurance plan record differs from the one to which such health services are ordinarily or exclusively available.⁵⁸  For example, a covered entity may not deny an individual treatment for ovarian cancer where the individual could benefit medically from the treatment, based on the individual’s identification as a transgender male.⁵⁹ 

The final rule also specifies requirements related to access to gender transition services. Specifically, an insurance plan may not categorically or automatically exclude or limit coverage for all health services related to gender transition or otherwise deny or limit coverage or deny a claim for specific health services related to gender transition, if such a policy results in discrimination against the individual seeking services.⁶⁰  HHS notes that blanket exclusions of all gender transition services, which historically have been used by some Medicaid programs and health insurers, are now recognized as outdated and not based on current standards of care.⁶¹  However, the final rule does not affirmatively require covered entities to cover any particular procedure or treatment for transition-related care.⁶² 

Compliance with and Enforcement of Section 1557’s Non-Discrimination Provisions

Each entity applying for federal financial assistance must assure its compliance with Section 1557.⁶³  Assurances also are required by Marketplace issuers and states seeking approval to operate a State-based Marketplace.⁶⁴ 

Covered entities must post a notice regarding their non-discrimination policies by October 16, 2016.⁶⁵  These notices must inform individuals about the bases of discrimination prohibited under Section 1557, the availability of free and timely auxiliary aids and services and language assistance services, how to access those services, contact information for the entity’s employee responsible for Section 1557 compliance, the entity’s Section 1557 grievance procedures, and OCR complaint procedures.⁶⁶  Notices must be available to beneficiaries, enrollees, applicants, and members of the public.⁶⁷  They must be printed in a conspicuously visible font and included in significant communications (such as handbooks and outreach publications⁶⁸ ), in conspicuous physical locations where the entity interacts with the public, and in a conspicuous location on the covered entity’s website accessible from the homepage.⁶⁹  Significant publications that are small-sized, such as postcards and tri-fold brochures, may include a shorter non-discrimination statement in lieu of the full notice.⁷⁰  The final rule contains a sample notice and nondiscrimination statement.⁷¹ 

Covered entities must provide the notice in English and include taglines in the top 15 languages spoken by individuals with LEP within the state.⁷²  Entities also may choose to post the notice in other non-English languages. Small-sized significant publications may include taglines in the top two non-English languages instead of 15.⁷³  HHS will translate its sample notice and taglines into 64 languages.⁷⁴ 

Covered entities that employ at least 15 people must adopt a grievance procedure that incorporates appropriate due process standards and provides prompt and equitable resolution of grievances under Section 1557.⁷⁵  HHS and covered entities with more than 15 employees also must designate at least one employee to coordinate the entity’s efforts to carry out Section 1557 responsibilities, including the investigation of grievances.⁷⁶  The final rule includes a sample grievance procedure.⁷⁷  To ensure compliance with Section 1557, HHS will provide covered entities with a training curriculum on key provisions of the rule.⁷⁸ 

Covered entities must take remedial action as required by the HHS OCR Director if they are found to have discriminated on any basis prohibited by Section 1557.⁷⁹  If discrimination is not found, covered entities may also take voluntary actions to improve operations.⁸⁰  OCR can enforce Section 1557 through informal mediation, reducing or eliminating a covered entity’s Federal financial assistance, or referring matters to the Department of Justice for litigation.⁸¹  The final rule also provides that private individuals and entities can sue in federal court to challenge alleged violations of Section 1557 by entities receiving Federal financial assistance and State-based Marketplaces.⁸²   The final rule clarifies that compensatory damages are available for Section 1557 violations.⁸³  Other existing federal, state, and local anti-discrimination laws also continue to apply.⁸⁴ 

Looking Ahead

HHS’s final rule implementing the ACA’s Section 1557 nondiscrimination protections coordinates existing federal non-discrimination laws, regulations, and policy regarding race, color, national origin, age, and disability as they apply to health coverage and care. Section 1557 also extends sex discrimination protections to health care programs and activities and includes gender identity as a prohibited basis of discrimination. While the rule does not resolve whether sexual orientation should be a stand-alone basis for prohibited discrimination, HHS will recognize Section 1557 sex discrimination claims involving sexual orientation that are based on gender stereotypes. The rule codifies HHS’s long-standing policy guidance on language assistance services for individuals with LEP. It also incorporates existing provisions of Section 504 and the ADA to prohibit disability-based discrimination in health programs and activities, although it defers setting specific standards for accessible medical equipment pending forthcoming standards from the Access Board. The rule also prohibits insurers from using discriminatory benefit designs, although it does not identify specific actions that constitute discrimination such as the placement of all drugs to treat a single medical condition on the highest cost-sharing tier, which is the subject of a recent administrative complaint that is pending with OCR. The final rule takes effect on July 18, 2016, except that provisions that require changes to plan benefit design are effective in the first plan year beginning on or after January 1, 2017.⁸⁵  HHS’s final rule is an additional tool to prevent and remediate discrimination in the health context, and HHS encourages other federal agencies that fund health programs and activities to adopt its standards to fully implement Section 1557.

1. 81 Fed. Reg. 31376-31473 (May 18, 2016), available at http://federalregister.gov/a/2016-11458. The proposed rule was published at 80 Fed. Reg. 54172-54221 (Sept. 8, 2015), http://www.gpo.gov/fdsys/pkg/FR-2015-09-08/pdf/2015-22043.pdf. ↩︎
2. 42 U.S.C. § 18116; 45 C.F.R. § 92.101.  In addition to prohibiting discrimination based on an individual’s own race, color, national origin, age, disability, or sex, Section 1557 also prohibits discrimination against an individual known or believed to have a relationship or association with someone else based on that other person’s race, color, national origin, age, disability, or sex. 45 C.F.R. § 92.209.  For example, a primary care physician could not refuse to accept a new patient based on the disability status of one of the patient’s family members.  80 Fed. Reg. 54191. ↩︎
3. 81 Fed. Reg. 31444. ↩︎
4. 45 C.F.R. § 92.1. ↩︎
5. 45 C.F.R. § § 92.2, 92.4.  Consistent with the HHS Office for Civil Rights’ enforcement of other civil rights authorities, the definition of Federal financial assistance does not include Medicare Part B payments.  81 Fed. Reg. 31383. ↩︎
6. 45 C.F.R. § § 92.1, 92.2(a). ↩︎
7. 81 Fed. Reg. 31379. ↩︎
8. 45 C.F.R. § 92.4. ↩︎
9. Id. ↩︎
10. 81 Fed. Reg. 31385. ↩︎
11. Unless the primary purpose of the Federal financial assistance is to fund employee health benefits, Section 1557 does not apply to an employer’s provision of employee health benefits where the provision of those benefits is the only health program or activity operated by the employer. 45 C.F.R. § 92.208; 81 Fed.Reg. 31437. ↩︎
12. 45 C.F.R. § 92.4. ↩︎
13. 81 Fed. Reg. 31386. ↩︎
14.   45 C.F.R. § § 92.4 92.206. ↩︎
15. 45 C.F.R. § 92.206. ↩︎
16. 45 C.F.R. § 92.4; 81 Fed. Reg. 31388. ↩︎
17. 45 C.F.R. § 92.4. ↩︎
18. 81 Fed. Reg. 31390. ↩︎
19. Id. ↩︎
20. Id.  ↩︎
21. 45 C.F.R. § 92.101(b)(3)(iv). ↩︎
22. 81 Fed. Reg. 31409. ↩︎
23.   45 C.F.R. § 92.2(b)(2). ↩︎
24. Id. ↩︎
25. 45 C.F.R. § 92.4. ↩︎
26. 81 Fed. Reg. 31387. ↩︎
27. 45 C.F.R. §92.201(a). ↩︎
28. 45 C.F.R. § 92.201(b)(1). ↩︎
29. 45 C.F.R. § 92.201(b)(2). HHS does not outline minimum requirements for a language access plan but notes that effective plans often address how the entity will determine an individual’s primary language; identify a telephonic interpretation service; identify a translation service; identify the types of language assistance services that may be required; and identify any documents for which written translations should be routinely available.  81 Fed. Reg. 31415. ↩︎
30. 45 C.F.R. § 92.201(c) and (d). ↩︎
31. 45 C.F.R § 92.201(e)(4). ↩︎
32. 45 C.F.R. § § 92.4, 92,201(e)(4). ↩︎
33. 45 C.F.R. § 92.201(f). ↩︎
34. 45 C.F.R § 92.201(g). ↩︎
35. 81 Fed. Reg. 31418-31419. ↩︎
36. 45 C.F.R. § 92.201(e)(1). ↩︎
37. 45 C.F.R § 92.201(e)(2). ↩︎
38. 45 C.F.R. § 92.201 (e)(3). ↩︎
39. 45 C.F.R. § 92.205. ↩︎
40. HHS notes that whether Section 1557 applies to people with chronic conditions depends on whether an individual meets the final rule’s definition of disability, which is based on the ADA/504.  81 Fed. Reg. 31382. ↩︎
41. 45 C.F.R. § 92.202. ↩︎
42. 81 Fed. Reg. 31421. ↩︎
43. 45 C.F.R. § 92.4. ↩︎
44. 45 C.F.R. § 92.202(b). ↩︎
45. 45 C.F.R. § 92.203(a).  Most facilities already are subject to the 2010 Standards.  Those that are not subject to the 2010 Standards as of July 18, 2016, have until January 18, 2018 to comply with those standards for new construction or alternations.  81 Fed. Reg. 32422-23. Construction or alteration that began before July 18, 2016 is deemed to comply with Section 1557 if it meets the 2010 Standards or the 1991 Standards, or the Uniform Federal Accessibility Standards but only if the facility was not covered by the 1991 or 2010 Standards. 45 C.F.R. § 92.203(b). ↩︎
46. 81 Fed. Reg. 31424. ↩︎
47. 45 C.F.R. § 92.204(a). Consistent with the ADA and Section 504, HHS notes that when determining whether an action is an undue burden, a covered entity must consider all resources available for use in the health program or activity’s funding or operation.  81 Fed. Reg. 31425.  The final rule replaces the language defining electronic and information technology with a citation to the Section 508 regulation to ensure that the definition is consistent with other civil rights laws and is not outdated and unduly narrow.  45 C.F.R. § 92.4. ↩︎
48. 45 C.F.R. § 92.204(a). ↩︎
49. 45 C.F.R. § 92.207(b)(1). ↩︎
50. Id.  ↩︎
51. Id. ↩︎
52. 45 C.F.R. § 92.207(b)(2). ↩︎
53. 45 C.F.R. § 92.207(d). ↩︎
54. 81 Fed. Reg. 31434-31435. ↩︎
55. 42 C.F.R. § 92.1. ↩︎
56. 81 Fed. Reg. 31434, n.258. ↩︎
57. Jane Perkins and Wayne Turner, NHeLP and The AIDS Institute Complaint to HHS Re HIV/AIDS Discrimination by Florida Insurers (May 29, 2014), available at  http://www.healthlaw.org/publications/browse-all-publications/HHS-HIV-Complaint#.Vg7Zx03bKUk. ↩︎
58. 45 C.F.R. § § 92.206, 92.207(b)(3). ↩︎
59. 81 Fed. Reg. 31428. ↩︎
60. 45 C.F.R. § 92.207 (b)(4) and (5). ↩︎
61. 81 Fed. Reg. 31429. ↩︎
62. Id. at 31434-31435. ↩︎
63. 45 C.F.R. § 92.5(a). ↩︎
64. Id.  ↩︎
65. 45 C.F.R. § 92.8(b). ↩︎
66. 45 C.F.R. § 92.8(a). ↩︎
67. Id. ↩︎
68. Covered entities may exhaust their existing supply of hard copy publications before complying with this provision.  81 Fed. Reg. 31396. ↩︎
69. 45 C.F.R. § 92.8(f). ↩︎
70. 45 C.F.R. § 92.8(f)(1)(i) and (g). ↩︎
71. 45 C.F.R. Part 92, Appendices A and B. ↩︎
72. 45 C.F.R. § 92.8(b), (d), (f).  Covered entities that serve individuals in more than one state may aggregate the number of individuals with LEP in those states to determine the top non-English languages spoken.  81 Fed. Reg. 31400. ↩︎
73. 45 C.F.R. § 92.8(d)(2). ↩︎
74. 45 C.F.R. § 92.8(c) and (e). ↩︎
75. 45 C.F.R. § 92.7(b).  The preamble indicates that covered entities may use the same grievance procedures and individuals as used to comply with Section 504 and Title IX if their scope is broadened to include all bases of discrimination prohibited by Section 1557.  81 Fed. Reg. 31394. ↩︎
76. 45 C.F.R. § 92.7(a).  OCR will serve as the responsible employee for Section 1557 compliance in HHS programs and activities and the Federally-facilitated Marketplace.  Id.  ↩︎
77. 45 C.F.R. Part 92, Appendix C. ↩︎
78. Id. at 31441. ↩︎
79. 45 C.F.R. § 92.6(a). ↩︎
80. 45 C.F.R. § 92.6(b). ↩︎
81. 81 Fed. Reg. 31439. ↩︎
82. 45 C.F.R. § § 92.302(d), 92.301(b). ↩︎
83. 45 C.F.R. § 92.301(a). ↩︎
84. 45 C.F.R. § 92.3(b). ↩︎
85. 45 C.F.R. § 92.1. ↩︎

Of the 2.6 million people who died in the U.S. in 2014,¹  2.1 million, or eight out of 10, were people on Medicare,²  making Medicare the largest insurer of medical care provided at the end of life. Spending on Medicare beneficiaries in their last year of life accounts for about 25% of total Medicare spending on beneficiaries age 65 or older.³  The fact that a disproportionate share of Medicare spending goes to beneficiaries at the end of life is not surprising given that many have serious illnesses or multiple chronic conditions and often use costly services, including inpatient hospitalizations, post-acute care, and hospice, in the year leading up to their death.

This data note provides a snapshot of Medicare beneficiaries who died at some point in 2014 and spending at the end of life. It examines Medicare per capita spending in 2014 and trends since 2000, both overall and by type of service, for beneficiaries in traditional Medicare who died at some point during the year (referred to here as “decedents”) compared to those who lived for the entire year (referred to here as “survivors”). The analysis includes beneficiaries who are under age 65 with disabilities, those who qualified for Medicare due to having end-stage renal disease, and those who are age 65 or older.⁴  The analysis of per capita spending excludes beneficiaries enrolled in Medicare Advantage because data on Medicare spending, overall and by type of service, are not available for these enrollees.⁵  The analysis is based on data from a 5% sample of Medicare claims for services covered under Parts A, B, and D for traditional Medicare beneficiaries from the Centers for Medicare & Medicaid Services (CMS) Chronic Conditions Data Warehouse (CCW) from 2000 to 2014.

What are the characteristics of medicare beneficiaries who died at some point in 2014?

More than half of Medicare decedents were age 80 or older in 2014.

• Of the more than 2.1 million Medicare beneficiaries who died at some point in 2014—representing 4.0% of the total Medicare population that year—over half (55%) were age 80 or older, which is more than double their share of the Medicare population overall (22%) (Figure 1). Just over half (52%) of decedents were women and eight out of 10 were non-Hispanic white (81%), roughly comparable to their shares of the overall Medicare population (54% and 77%, respectively).

• More than seven in 10 (72%, or 1.5 million) Medicare decedents were in traditional Medicare in 2014, and the remainder (28%, or 0.6 million) were enrolled in Medicare Advantage plans, reflecting overall enrollment patterns in traditional Medicare and Medicare Advantage.
• Diseases that were highly prevalent among decedents in traditional Medicare in 2014 include hypertension (67%), ischemic heart disease (53%), chronic kidney disease (51%), congestive heart failure (48%), Alzheimer’s disease or dementia (43%), diabetes (38%), and cancer (17%). The prevalence of each of these conditions was higher among beneficiaries who died at some point in 2014 than among beneficiaries overall, in some cases substantially higher (Table 1). For example, more than 4 in 10 decedents had Alzheimer’s or dementia, compared to only 9% of beneficiaries overall, and more half of decedents had ischemic heart disease, compared to one fourth of beneficiaries overall.

How does medicare per capita spending differ for decedents and survivors in 2014 and over time?

Average total Medicare per capita spending was nearly four times higher for beneficiaries in traditional Medicare who died at some point in 2014 than for those who lived the entire year.

• Average Medicare per capita spending on services covered under Parts A, B, and D for traditional Medicare beneficiaries who died at some point in 2014 was $34,529—nearly four times higher than per capita spending for survivors ($9,121) and more than three times higher than the average among all beneficiaries in traditional Medicare ($10,126) (Figure 2).

• In 2014, beneficiaries who died at some point during the year accounted for 4% of all beneficiaries in traditional Medicare, but 13.5% of traditional Medicare spending. This amount is disproportionate to the decedent share of beneficiaries overall, but it accounts for a relatively small share of total spending that year. This estimate is lower than the 25% estimate cited earlier because it is based on Medicare spending for people who died at some point in a given calendar year (in this case, 2014), rather than the last 12 months of spending for people who died.⁶ 

The share of total traditional Medicare spending on beneficiaries who died at some point during the year has decreased over time.

• The share of total traditional Medicare spending on beneficiaries who died at some point during the year has dropped over time, from 18.6% in 2000 to 13.5% in 2014 (Figure 3). This drop is likely due to a combination of factors affecting total traditional Medicare spending over time and spending on decedents, including: growth in the number of Medicare beneficiaries overall, particularly in recent years as the baby boom generation ages on to Medicare, which means more younger, healthier beneficiaries, on average; longer life expectancy, which means people are living longer and dying at older ages (as seen in a decline in the share of traditional Medicare beneficiaries who die at some point in a given year—from 4.9% in 2000 to 4.0% in 2014); lower average per capita spending on older decedents compared to younger decedents (as described further below); and slower growth in the rate of annual per capita spending for decedents than survivors (also described further below).

Between 2000 and 2014, the annual rate of growth in average Medicare per capita spending was lower among decedents than survivors.

• Average Medicare spending among decedents was 80% higher per person in 2014 ($34,529) than in 2000 ($19,130), while average spending among survivors more than doubled between 2000 and 2014, from $4322 to $9,121 (Figure 4).

• Although average per capita spending was higher for decedents than survivors in each year between 2000 and 2014, the average annual rate of growth in spending over this time period was lower for decedents (4.3%) than for survivors (5.5%).

What services account for the difference in medicare per capita spending between decedents and survivors?

Higher Medicare per capita spending among decedents than survivors in traditional Medicare is primarily driven by much higher spending on inpatient hospital services.

• Spending on inpatient hospital services accounted for the largest amount of per capita Medicare spending by type of service among decedents in traditional Medicare in 2014 (Figure 5; Table 2), and was the primary reason for the substantial difference in spending between decedents and survivors. Per capita inpatient hospital spending among decedents was $17,574 in 2014, on average, seven times higher than among survivors ($2,497). Decedents also incurred much higher spending on post-acute care (skilled nursing facility (SNF) and home health services) and hospice care than survivors in 2014.

• Inpatient hospital services accounted for the largest share of average Medicare per capita spending by type of service for traditional Medicare beneficiaries overall in 2014 (31%), but the share of per capita spending on inpatient hospital services was particularly large for decedents, accounting for just over half (51%) of total spending in 2014 (Figure 6).

• For decedents, the next largest service categories were physicians, providers, and supplies combined and post-acute care (skilled nursing facilities and home health services), at 13% each, followed by hospice care (10%). For surviving beneficiaries in traditional Medicare, inpatient hospital services accounted for 27% of total per capita Medicare spending in 2014, followed closely by spending on physicians/providers/supplies (23%) and Part B and Part D prescription drugs (21%).

How does Medicare per capita spending differ for decedents under and over age 65, and what accounts for the difference in spending?

Medicare per capita spending was higher for decedents under age 65 in 2014 than for those over age 65.

• Among decedents in traditional Medicare in 2014 who were younger than age 65, Medicare per capita spending was $41,950, on average (Figure 7). This amount is 25% more than average per capita spending among decedents over age 65 in 2014 ($33,676).

• The fact that Medicare per capita spending is higher for decedents under age 65 than those over age 65 is related to the fact that a larger share of traditional Medicare beneficiaries under age 65 who died at some point in 2014 is dually eligible for Medicare and Medicaid than of decedents who are over age 65, and their Medicare per capita spending is significantly higher, on average, than dually eligible beneficiaries over age 65. Among dually eligible decedents in 2014, average Medicare per capita spending among those under age 65 was $51,997, compared to $36,037 among those over age 65.
• In terms of Medicare spending by type of service, higher per capita spending among decedents under age 65 than among those over age 65 was driven by higher spending on inpatient hospital services ($24,515 versus $16,701, respectively), along with higher spending on outpatient hospital, physician/providers/supplies, and Part B/D prescription drugs (Figure 8). By contrast, among decedents over age 65, spending was higher on post-acute care and hospice care.

How does medicare per capita spending differ among decedents over age 65?

Per capita Medicare spending generally declines with age among decedents in traditional Medicare who are over age 65.

• Among decedents over age 65, Medicare per capita spending was highest for those in their early 70s and then declined with each year of age in 2014 (Figure 9). Among survivors over age 65, the opposite pattern occurred, with spending rising steadily with each year of age in 2014.

What services account for the difference in spending by age among decedents over age 65?

The decline in Medicare per capita spending by age for decedents over age 65 in 2014 was mainly due to lower inpatient spending.

• Per capita Medicare spending on inpatient hospital services decreased steadily with age among decedents in 2014 (Figure 10; Table 3), declining from more than $20,000 for decedents in their late sixties and seventies to around $10,000 or less among decedents older than age 90. In contrast, per capita spending for hospice services increased with age among decedents, from around $2,000 to $5,000 or more between the ages of 66 and 100. Spending on post-acute care (SNF and home health) also increased with advancing age.

• In contrast, spending increased with each year of age among survivors over age 65 in 2014, primarily due to increasing spending on post-acute and hospice care (Figure 11; Table 4). Inpatient hospital care was the costliest type of service for survivors in traditional Medicare up to the early nineties, after which age spending on post-acute care was the largest amount.

Discussion

Much attention has been focused lately on end-of-life care, the services patients receive at the end of life, and Medicare spending on these services. This analysis contributes a number of important findings to the discussion surrounding these issues.

Our analysis shows that Medicare per capita spending for beneficiaries in traditional Medicare who died at some point in 2014 was substantially higher than for those who lived the entire year, as might be expected. It also shows that Medicare per capita spending among beneficiaries over age 65 who die in a given year declines steadily with age. Per capita spending for inpatient services is lower among decedents in their eighties, nineties, and older than for decedents in their late sixties and seventies, while spending is higher for hospice care among older decedents. These results suggest that providers, patients, and their families may be inclined to be more aggressive in treating younger seniors compared to older seniors, perhaps because there is a greater expectation for positive outcomes among those with a longer life expectancy, even those who are seriously ill.

In addition, we find that total spending on people who die in a given year accounts for a relatively small and declining share of traditional Medicare spending. This reduction is likely due to a combination of factors, including: growth in the number of traditional Medicare beneficiaries overall as the baby boom generation ages on to Medicare, which means a younger, healthier beneficiary population, on average; gains in life expectancy, which means beneficiaries are living longer and dying at older ages; lower average per capita spending on older decedents compared to younger decedents; slower growth in the rate of annual per capita spending for decedents than survivors, and a slight decline between 2000 and 2014 in the share of beneficiaries in traditional Medicare who died at some point in each year.

This analysis focuses exclusively on beneficiaries in traditional Medicare, excluding the roughly one in three beneficiaries who are enrolled in Medicare Advantage plans,⁷  because comparable spending data for Medicare Advantage enrollees are not available. With research showing significant differences in certain demographic and health status characteristics between decedents in traditional Medicare and in Medicare Advantage,⁸  it is possible that spending and service use patterns may differ as well. While the majority of Medicare beneficiaries who died at some point in 2014 were in traditional Medicare, the inclusion of Medicare Advantage enrollees would provide a better understanding of whether and how the experiences of traditional Medicare and Medicare Advantage beneficiaries differ at the end of life.

Decisions pertaining to end-of-life care are among the most difficult for patients, families, and health care providers. The recent change in Medicare payment policy to reimburse physicians for conversations about advance care planning with their patients⁹  could bring about changes that better align services delivered with patient preferences and that also potentially reduce the costs associated with care at the end of life.

1. National Center for Health Statistics, Health, United States, 2015: With Special Feature on Racial and Ethnic Health Disparities, 2016, Table 19: “Leading causes of death and numbers of deaths, by sex, race, and Hispanic origin: United States, 1980 and 2014,” available at http://www.cdc.gov/nchs/data/hus/hus15.pdf#019. ↩︎
2. Kaiser Family Foundation analysis of a five percent sample of 2014 Medicare claims from the Centers for Medicare & Medicaid Services’ Chronic Conditions Data Warehouse. ↩︎
3. Gerald F. Riley and James D. Lubitz, “Long-Term Trends in Medicare Payments in the Last Year of Life,” Health Services Research April 2010; 45(2):565-576. ↩︎
4. We exclude beneficiaries who are exactly age 65 from the analysis of spending by year of age and spending comparisons between beneficiaries under age 65 and over age 65 because many are enrolled for less than a full year, resulting in less than an entire year of Medicare spending data. ↩︎
5. Medicare spending for Medicare Advantage enrollees takes the form of monthly capitation payments which are not based on actual service utilization. ↩︎
6. We are unable to examine spending data that spans the last 12 months of life for decedents in our analysis because we are unable to disaggregate spending data for a given beneficiary for periods of time less than one calendar year. Had we been able to calculate Medicare spending for a 12-month period for both survivors and dependents, the difference in per capita spending between decedents and survivors and the estimate of spending on decedents as a share of total Medicare spending would have been larger. ↩︎
7. Gretchen Jacobson, Giselle Casillas, Anthony Damico, Tricia Neuman, and Marsha Gold, “Medicare Advantage 2016 Spotlight: Enrollment Market Update,” Kaiser Family Foundation, May 2016, available at https://modern.kff.org/medicare/issue-brief/medicare-advantage-2016-spotlight-enrollment-market-update/. ↩︎
8. Elena Byhoff, John A. Harris, and John Z. Ayanian, “Characteristics of Decedents in Medicare Advantage and Traditional Medicare,” Research Letter, JAMA Internal Medicine June 2016; Published online ahead of print: E1-3. ↩︎
9. For more information, see “10 FAQs: Medicare’s Role in End-of-Life Care,” Kaiser Family Foundation, November 2015, available at https://modern.kff.org/medicare/fact-sheet/10-faqs-medicares-role-in-end-of-life-care/. ↩︎