 To investigate the impact of Part D coverage gap reform on diabetes medication adherence.

 Retrospective data analysis based on pharmacy claims data from a national pharmacy benefit manager.

We used a difference-in-difference-indifference method to evaluate the impact of coverage gap reform on adherence to diabetes medications. Two cohorts (2010 and 2011) were constructed to represent the last year before Affordable Care Act (ACA) reform and the first year after reform, respectively. Each patient had 2 observations: 1 before and 1 after entering the coverage gap. Patients in each cohort were divided into groups based on type of gap coverage: no coverage, partial coverage (generics only), and full coverage.

 Following ACA reform, patients with no gap coverage and patients with partial gap coverage experienced substantial drops in copayments in the coverage gap in 2011. Their adherence to diabetes medications in the gap, measured by percentage of days covered, improved correspondingly (2.99 percentage points, 95% confidence interval [CI] 0.49-5.48, 

 = .019 for patients with no coverage; 6.46 percentage points, 95% CI 3.34-9.58, 

 <.0001 for patients with partial coverage). Patients with full coverage also had lower copayments in the gap in 2011. However, their adherence did not increase (—0.13 percentage point, 

 In the first year of ACA coverage gap reform, copayments in the gap decreased substantially for all patients. Patients with no coverage and patients with partial coverage in the gap had better adherence in the gap in 2011.

The Affordable Care Act mandated that Part D coverage gap be shrunk starting from 2011. The reform had an important impact on adherence to diabetes medications in the gap.

		 The coverage gap reform decreased copayments for diabetes medications substantially in the coverage gap for all patients.

		 Patients with no coverage and patients with partial coverage in the gap had significant improvements in adherence.

		 Patients with full coverage in the gap had unchanged adherence to diabetes medications in the coverage gap in 2011.

Medicare Part D was implemented in 2006 to provide prescription drug benefits for Medicare beneficiaries. The implementation of Part D has led to many positive changes: reduced out-of-pocket expenses, increased medication fills, improved adherence, decreased medical spending, and fewer avoidable hospitalizations.

One controversial aspect of the Part D benefit design is the coverage gap, commonly referred to as the “donut hole.” Patients with a Part D plan that has a standard defined benefit have a copayment of 25% on prescription drugs after they pay the initial deductible. After their total pharmacy spending reaches the coverage gap limit ($2830 in 2010), patients have a 100% copayment for their drugs until they reach the catastrophic limit ($6440 in 2010). In the catastrophic stage patients pay 5% or $2.60 per prescription.

The lack of coverage in the donut hole raises concern that the increase in copayment may have an adverse effect on drug utilization. Research has shown that reaching the coverage gap decreases medication fills

 and lowers adherence to essential medications.

 In response to these concerns, the Affordable Care Act (ACA), passed in March 2010, includes a provision to shrink the coverage gap. Starting in 2011, the copayment for brand drugs in the coverage gap was reduced 50%, and the copayment for generic drugs was reduced 7%. The copayments for brand drugs and generic drugs in the gap will be further reduced after 2012. In the year of 2020, the donut hole will be completely closed with a 25% copayment for both brand and generic drugs.

The closure of the donut hole represents another significant change for Medicare Part D. A solid understanding of the latest reform is critical for the design of Part D benefits. This research contributes by investigating the impacts of the 2011 coverage gap reform on adherence to diabetes medications. We focused on diabetes because it is a highly prevalent disease among Medicare beneficiaries. It is estimated that in 2010, the prevalence rates for people aged 65 to 74 years and >75 years were 20.7% and 18.9%, respectively. The diabetes prevalence rates for the 2 groups are expected to become 30.1% and 32.7% in 2050.

 Effective management of diabetes is vital to the success of Medicare.

There are 4 types of Part D plans: standard defined benefit, actuarially equivalent, basic alternative, and enhanced alternative. Of the 4 types of plans, the actuarially equivalent and basic alternative plans are similar to the standard defined benefit plan in coverage. Patients with an enhanced alternative plan can pay a higher premium for better coverage, such as having coverage in the gap and/or covering more drugs than standard defined benefit plans.

The year 2011 is the first year of coverage gap reform mandated by the ACA. In 2011, patients reaching the gap received a 50% copayment reduction for brand drugs and a 7% copayment reduction for generic drugs. For patients whose Part D plans provided coverage in the gap, discounts were applied to their copayments directly. For example, suppose a patient who had coverage while in the gap in 2010 paid a 25% copayment for both brand and generic drugs in the coverage gap. In 2011, this same patient would pay a 12.5% copayment for brand drugs and a 23.25% copayment for generic drugs in the coverage gap.

Data for this research came from pharmacy claim data from MedImpact HealthCare Systems Inc’s book of business. Med- Impact is a national pharmacy benefit manager. The research data included a 2010 cohort and a 2011 cohort. The 2010 cohort represents the last year before ACA reform and the 2011 cohort represents the first year after ACA reform.

Patients in both cohorts were required to be continuously enrolled in the same health plan for 2 years: 1 year before the cohort year and the cohort year. Patients with diabetes were identified as having at least 2 diabetes drug claims in both the cohort year and the year before the cohort year. It was important to require patients to have used diabetes drugs before the cohort year so that we had a full year to measure adherence in the cohort year. The inclusion of 1 year of data before the cohort year also was important to construct a comorbidity index. All patients had to be at least 65 years old in their cohort year. Patients eligible for a low-income subsidy were removed because their copayment structures are different.

Each data cohort included patients with no coverage, partial coverage (ie, generic coverage only), and full coverage in the gap. In this research, patients with partial and full coverage all came from enhanced alternative plans. Patients with no coverage came from standard defined benefit, actuarially equivalent, basic alternative, and enhanced alternative plans. Theoretically, patients in actuarially equivalent, basic alternative, and enhanced alternative plans could have coverage gap limits different from those of standard defined benefit plans. However, that was not the case in this research. We checked and confirmed that all patients in the data set had the same coverage gap limits.

These identification steps generated 20,709 patients in the 2010 cohort and 

2 patients in the 2011 cohort. We selected only patients who entered the coverage gap but did not enter the catastrophic level to make it easier to compare adherence before the gap with adherence in the gap.

This exclusion led to 6828 patients in the 2010 cohort and 6124 patients in the 2011 cohort.

The outcome variable was adherence to diabetes medications. Diabetes is a long-term and progressive disease. Adherence to diabetes medication is critical for maintaining glycemic control and slowing down disease progression. Poor adherence can lead to inadequate glycemic control, higher probabilities of hospitalization, higher medical cost, and higher mortality.

Adherence is measured by proportion of days covered (PDC), defined as the number of days covered by at least 1 diabetes medication divided by the length of the study period based on medication fill date and days of supply. The PDC methodology has been used in many studies.

The value of PDC is bounded between 0.0 and 1.0.

Difference-in-Difference-in-Difference Method

This research used a difference-in-difference-in-difference method to identify the impact of coverage gap reform on adherence to diabetes medications. In this model, year 2010 data and year 2011 data were used to represent the periods before and after coverage gap reform, respectively. Each patient in a cohort year had 2 observations: one before the coverage gap and the other in the coverage gap. Patients with no coverage or partial coverage were the treatment group; patients with full coverage were the control group. In a straightforward difference- in-difference-in-difference model, the effect of coverage gap reform was captured by the interaction of coverage gap, year 2011, and the type of coverage in the coverage gap.

One problem of a straightforward difference-in-difference-in-difference model is that the outcome variable, PDC, was bounded between 0 and 1. In this research, the mean PDC value was around 80% and the distribution was highly skewed. This can be problematic in an ordinary least squares regression. One way to address this issue is to transform the outcome variable as the change of PDC after reaching the coverage gap, defined as PDC in the coverage gap minus PDC before the gap. After the transformation, the model can be written as:

 equals PDC in the coverage gap minus PDC before the coverage gap for each patient, 

 is the dummy variable indicating no coverage in the gap, 

 is the dummy variable indicating partial coverage in the gap, and 

 is the dummy variable indicating the 2011 cohort.

 are the coefficients of interest. They represent the change in adherence in the coverage gap in response to coverage gap reform for patients with no coverage and partial coverage, respectively.

Through the transformation, each patient’s idiosyncratic propensity to be adherent was removed. Patients’ unchanged characteristics such as age, sex, and geographical location were also removed. As a consequence, the inclusion of covariates such as age and sex was not necessary. The main model was run without covariates. We included covariates in the model as part of the robustness analysis. Control variables included age, sex, geographical location, whether insulin was used in the last year, and comorbidities. Information on patient demographics was obtained from patients’ eligibility files. Information on the use of insulin was available from the previous year’s pharmacy claim file. Comorbidities were measured by RxRisk, a comorbidity measure based on pharmacy claims in the prior year.

Each RxRisk category indicates whether a patient took certain classes of medications (eg, those for hypertension and hyperlipidemia) in the previous year. Only RxRisk categories with prevalence rates higher than 5% were included in the study.

One notable difference between our outcome variable and the comparable measure in the Centers for Medicare & Medicaid Services (CMS) Five-Star Quality Rating System for Medicare Advantage Plans is that the CMS star rating uses only oral diabetes medications in its calculation. This research uses all diabetes medications including insulin. A robustness check was performed at the end of the study by removing patients with a history of insulin use from the model. This robustness check was important also because patients receiving insulin likely had either advanced type 2 diabetes or type 1 diabetes. These patients can be different from patients who do not use insulin, who tend to be patients in early stage of type 2 diabetes.

An alternative model was to use a dichotomous variable of adherence as the outcome variable, with a value of 1 assigned to patients having 

80% PDC. This dichotomous variable of adherence is used frequently in adherence studies.

 One problem of using a dichotomous outcome variable is that the interpretation of interaction terms in a logit/probit model is not straightforward. The interaction effect varies by patients. It can be positive or negative depending on many factors.

 This research uses a continuous outcome variable for ease of interpretation.

A generalized estimating equation was used to model the effect of coverage gap reform. The model was adjusted for health plan clustering effect to account for potential similarities of patients within the same health plan. The whole statistical analysis was performed with SAS 9.3 (SAS Institute, Cary, North Carolina).

Articles

POLICY

The Part D coverage gap reform in 2011 improved adherence to diabetes medications in the coverage gap.

Effects of Coverage Gap Reform on Adherence to Diabetes Medications

 Prescription opioid abuse and addiction are serious problems with growing societal and medical costs, resulting in billions of dollars of excess costs to private and governmental health insurers annually. Although difficult to accurately assess, prescription opioid abuse also leads to increased insurance costs in the form of property and liability claims, and costs to state and local governments for judicial, emergency, and social services. This manuscript’s objective is to provide payers with strategies to control these costs, while supporting safe use of prescription opioid medications for patients with chronic pain.

 A Tufts Health Care Institute Program on Opioid Risk Management meeting was convened in June 2010 with private and public payer representatives, public health and law enforcement officials, pain specialists, and other stakeholders to present research and develop recommendations on solutions that payers might implement to combat this problem.

While protecting access to prescription opioids for patients with pain, private and public payers can implement strategies to mitigate financial risks associated with opioid abuse, using internal strategies such as formulary controls, claims data surveillance, and claims matching; and external policies and procedures that support and educate physicians on reducing opioid risks among patients with chronic pain.

Reimbursement policies, incentives, and health technology systems that encourage physicians to use universal precautions, to consult prescription monitoring program (PMP) data, and to implement Screening, Brief Intervention, and Referral to Treatment protocols have a high potential to reduce insurer risks while addressing a serious public health problem.

Many payers in the healthcare industry lack the resources to comprehensively assess the risk of prescription opioid abuse, diversion, and addiction on the health of their members and on their corporate margins. Although greater research remains to be performed, strategies have been identified that companies and government agencies can use now to manage the risks involved. It is critically important that payers stay abreast of new developments in this field and implement risk management strategies on an ongoing basis if they are to stem the growing tide of financial losses associated with this serious societal problem.Access to prescription opioid (RxO) medications is essential to the care and daily function of millions of patients who live with constant pain. The use of prescription opioids has increased tremendously in the past decade (from 174 million in 2000 to 257 million in 2009)

 due to the widespread availability and variety of prescription opioid products, and changes in treatment paradigms. While most patients use these medications appropriately, some do not. RxO misuse, abuse, addiction, overdose, and other health and social consequences of inappropriate RxO use are taking a rapidly growing toll on individuals and institutions in the United States. It is estimated that 2.2 to 2.4 million individuals initiate non-medical use of RxOs in the United States each year,

and non-medical RxO use now exceeds use of many conventional street drugs, including cocaine and heroin.

 Overdose deaths from prescription drugs have exceeded those from street drugs since 2002,

 and have surpassed traffic accidents as a cause of accidental death.

 In 2007, over 850,000 of 1.9 million emergency department (ED) visits associated with drug abuse or addiction involved non-medical use of pharmaceuticals, and over one-third of these involved RxOs.

Much of the financial cost of RxO abuse, addiction, and associated health consequences is borne by private and public health insurers. Individuals addicted to RxOs, as well as those who sell them, obtain them by diverting prescriptions, frequently paid for by insurers or government programs.

 In addition to the cost of the drugs themselves, individuals who do not adhere to opioid treatment (likely due to abuse or addiction) incur higher total healthcare costs than adherent patients,

probably due to the serious health consequences of opioid abuse and addiction. Overall, the medical and prescription costs associated with opioid addiction and diversion have been estimated at $72.5 billion annually for private and public healthcare payers.

 Reports from the Government Accountability Office (GAO) on RxO fraud and abuse in Medicaid

have documented the substantial size of this problem and recommended solutions; yet it does not appear that such recommendations have been adopted among public or private payers.

A Tufts Health Care Institute (THCI) Program on Opioid Risk Management summit meeting was convened in June 2010 to bring together various stakeholders to prePrescription sent their perspectives and research, to raise awareness of the complex issues involved, and to consider potential solutions that healthcare payers might implement to combat this difficult and costly problem.

Prescription drug costs represent about 10% of US healthcare spending, and the share of that cost that is borne by healthcare payers has increased significantly in recent years.

 In 1990, prescription drug costs were about $40 billion, with 56% paid directly by consumers and 26% paid by private insurers.

 In 2006, this had risen to $216.7 billion a year, with 22% paid by consumers and 44% by private insurers. During the same period, the government share of prescription drug costs rose from 18% to 34%.

Prescriptions for controlled substances represent a significant proportion of the rising costs for health insurers, reflecting an increase in the therapeutic use of opioids to manage chronic pain.

Total opioid prescriptions rose over 200% between 1992 and 2002, with hydrocodone and oxycodone posting nearly 400% increases.

 Most patients who seek RxOs are looking for pain relief, but some obtain opioids for pleasure seeking or financial gain. Moreover, some individuals are initially prescribed opioids for pain but become addicted and need larger quantities of opioids because of ever-increasing tolerance. Some individuals visit multiple physicians without the others’ knowledge, a practice known as “doctor shopping,” or fill prescriptions at multiple pharmacies, making them more difficult to track. These prescriptions may be used by addicted individuals themselves, diverted to family members or friends, or sold on the street. Physicians, pharmacists, and clinics may also engage in fraudulent activities, selling prescriptions or medications to non-legitimate patients, while billing the costs to private insurers or the government.

 Such activities can be quite costly. A review of prescription data from Medicaid in Massachusetts (MassHealth) found that the top 50 members with questionable behaviors suggestive of abuse and diversion cost the program $161,000 in aggregate for controlled substance prescriptions in 2009, with total healthcare costs averaging $72,000 for each member. This led the agency to implement treatment, formulary, data matching, lock-in, and other controls detailed in this document. Owing to these measures, MassHealth has experienced a significant reduction in cost and potential/ probable program abuse.

RxO abuse is also thought to have a large impact on other types of insurers.13 Automobile, property, and workers compensation insurers pay both directly and indirectly for accidents, theft, property damage, and workers compensation claims from intoxicated or addicted individuals. In 2009, an estimated 10.5 million people, or 4.2% of the population 12 years or older, reported drugged driving.

 The rate was highest among young adults aged 18 to 25 years (12.8%).

Opioid abuse, addiction, and diversion also have important costs to the public sector, at local, state, and federal levels.

 Locally, there are substantial costs to law enforcement agencies, court systems, and emergency service providers.

 Direct costs are incurred for responses to fraud, theft, or illegal actions by individuals, as well as responses to injury and rescue calls. Costs for state and local governments include court systems, departments of correction, drug prevention and rehabilitation programs, and other public services.

Many individuals with pain or addiction to RxOs are unable to work, increasing costs for federal or state income assistance and support services for dependents.

Individual practitioners, many of whom received little or no training in pain management in medical school, have become reluctant to prescribe opioids and manage their consequences. 

 This situation creates a 2-sided problem: on 1 side, the potential for underutilization of opioids where appropriate, leaving pain untreated, and on the other side, the potential failure to monitor, detect, and proactively manage harmful consequences of opioid use.

Private and public healthcare payers can consider a wide range of internal and external procedures and policies, to lower financial risks associated with RxO abuse. They can establish and maintain internal procedures, such as claims reviews, claims matching, and formulary controls, that are likely to help identify individuals who abuse RxOs and reduce inappropriate prescribing. They can also establish external policies that promote and support appropriate prescribing behavior by physicians and other healthcare providers. The overarching goals of these strategies should be to minimize RxO abuse while ensuring that legitimate patients with pain have access to these medications, that patients with addiction receive appropriate treatment, and that the privacy of all patients remains protected. In addition, policies should ensure that these measures do not increase administrative burdens on prescribers.

 A number of payers have already implemented processes that can help limit abuse of RxOs. Formulary controls that limit reimbursement can help ensure that higher-risk opioids are not prescribed unless the risks outweigh the benefits, and that appropriate monitoring is implemented. Claims review procedures can be designed to question potentially inappropriate prescriptions. Several studies have shown that individuals at high risk for RxO fraud or abuse can be identified using claims data or state prescription monitoring program (PMP) data.

 Individuals who are suspected of doctor shopping can be “locked in” to the use of a single pharmacy and/or single prescriber to minimize inappropriate or fraudulent prescriptions.

 For example, Oklahoma’s SoonerCare pharmacy lock-in program showed reductions in RxO use, visits to multiple pharmacies and physicians, and ED visits, with estimated cost savings of greater than $31,500 for 52 members after 12 months.35,36 Such interventions can include appropriate referral of patients to pain management specialists or addiction treatment services if necessary.

Promoting the Use of Opioid Formulations That Potentially Minimize Abuse. 

Individuals who abuse RxOs often do so by crushing and injecting, snorting, or chewing oral formulations. Several opioid formulations have been designed to prevent or reduce these forms of abuse, including 2 that have been approved by the US Food and Drug Administration,

 although the abuse deterrence of these formulations is not supported by claims in the label. Because approval of these formulations is recent, data regarding their potential benefits are not available yet. A budget impact model to quantify the potential cost savings associated with a hypothetical opioid formulation designed to resist or deter common methods of extraction estimated potential cost savings to third-party payers at 0.6 to 1.6 billion dollars per year.

Prescription claims data can be reviewed for overuse of RxOs, including use of multiple prescribers, multiple pharmacies, and/or early refills.

 MassHealth, for example, reviews prescriptions for controlled substances to identify cases of inappropriate prescribing, including a review of clinical outcomes and medically appropriate use [GG, personal communication]. A few companies have instituted procedures for matching prescription, medical, behavioral, and substance abuse claims that can identify patterns or “red flags” suggestive of abuse or fraud. At Wellpoint, a private health benefit company, a comprehensive internal mechanism has been developed that proactively identifies members suspected of inappropriate use of pharmacy or medical benefits. Once identified, each case is carefully researched, the member and the healthcare provider involved are contacted, and interventions can range from provider and patient education, to addiction treatment services, to involvement of law enforcement if necessary.

Claims data can also help identify individuals at high risk for opioid abuse. The Trends and Risks of Opioid Use for Pain (TROUP) study, a large observational study of claims data, identified a number of factors that were associated with a high risk of opioid abuse, including younger age, back pain, and multiple pain complaints.

Mental health and substance use disorders have also been shown to be strongly associated with increased risks of opioid abuse or addiction.

Claims matching can be difficult to implement because many claims databases do not interface with one another. Thus payers should develop procedures for handling potential fraud cases that are effective but still allow for individualized consideration of each case, including clinical oversight to ensure that patients and providers are not penalized for appropriate medical uses of opioids.

 It is also important that controls be in place to protect patient privacy where needed.

The GAO in its 2009 report on fraud and abuse in the Medicaid system recommended that, at the time of dispensing an opioid analgesic at the pharmacy, a fraud and abuse screen be automatically conducted, including checks of PMP data, validity of Drug Enforcement Administration (DEA) registrations of prescriber and pharmacist, vital status registries to ensure that both prescriber and patient are alive, and federal debarment and exclusion databases (

). While such an intervention would appear to be relatively inexpensive, minimally burdensome, and result in substantial return on investment, it does not appear to have been implemented. Requiring the use of tamper-resistant prescription pads, another fraud-reduction approach, has recently been implemented by Medicaid in New York State and has led to dramatic reductions in money spent on fraudulent prescriptions.

 This measure could be expanded to other payers.

Cooperation and Data Sharing With Other Types of Insurance Providers.

 Automobile, property, and workers compensation insurers have a wealth of data on claims information involving individuals who have incurred damage and liability losses, some of whom abuse RxOs. The property/ casualty insurance industry maintains common databases, but currently such databases are not linked to databases of health insurance claims or prescriptions. Future industry standards could be developed that would allow not only cross-industry data matching, but also integration with public sector databases, facilitating identification of abuse, addiction, and fraud. Implemented with appropriate privacy controls, such efforts would reduce costs, benefiting both private and public payers as well as patients who might be identified earlier as in need of addiction treatment services.

Encouraging Use of Risk Management Strategies Among Healthcare Providers

 The most common source of abused RxOs is, directly or indirectly, by prescription.3 Thus, much of the responsibility for safe prescribing rests upon clinicians, including primary care practitioners, pain management specialists, dentists, and others. By accepting the fact that any patient could potentially abuse RxOs, clinicians can adopt tools and processes, termed a “universal precautions” approach,

 which aim to mitigate misuse, abuse, and diversion of RxOs by patients. These universal precautions include careful screening and risk stratification of patients; effective patient education and counseling that maximizes patient involvement in treatment; individualized treatment that is periodically reassessed, using urine drug testing, pill counts, or other measures if deemed necessary; and careful documentation of the pain management process (see

 Guidelines for managing risk in opioid prescribing are available from various medical societies.

 Payers, in turn, can limit their own risk by promoting and supporting use of these precautions among healthcare providers in their systems.

Surveys of physician attitudes about RxOs have shown that, while many practitioners recognize the effectiveness of these medications, they are justifiably concerned about creating addiction in their patients.

A number of studies have found that previous substance abuse, including alcohol abuse, represents a strong, overlapping risk factor for abuse of prescription medications.

A wide range of other risk factors for opioid abuse have been identified, including age, gender, family history of substance abuse, presence of a major psychiatric disorder, a history of aberrant behaviors, and others.

 Payers can promote patient screening by encouraging practitioners to use some of the many screening tools that are available for this purpose, and reimbursing for the time involved. Medicaid (“H”), Medicare (“G”), and Current Procedural Terminology codes designed specifically to reimburse for substance abuse screening and brief interventions became effective in 2008.

Detection of these risks in many patients through automatable processes at the payer level can provide prescribers with useful risk information on their patients without increasing the prescriber’s burden.

Screening, Brief Intervention, and Referral to Treatment. 

Considerable research has been devoted to Screening, Brief Intervention, and Referral to Treatment (SBIRT), a comprehensive, integrated public health approach to identifying those who have or are at risk for problematic substance abuse or substance abuse disorders, and providing early intervention. SBIRT is designed to be implemented in primary care centers, hospital EDs, trauma centers, and other community settings where substance abusers often first interact with the healthcare system. Patients are screened for substance abuse problems, then provided with a brief intervention or treatment (for low- to moderate-risk cases) or referred to specialists (for high-risk cases). Use of SBIRT is associated with reduced substance abuse, fewer ED visits, and fewer hospital days, and several large studies have suggested that it is cost-effective for payers.

 Payers can support the use of SBIRT by aligning their reimbursement and training policies with the services required to follow this method.

The increased use of opioid analgesics and concomitant rise in abuse and addiction underlie another issue that may be amenable to payer support: case management. In particular, offering support to primary care physicians and specialists may help reduce clinical dependence on opioid use for persistent pain. Programs including medication review and promotion of nonpharmacologic therapies (eg, acupuncnture, massage, health/wellness classes) may shift the focus away from simply prescribing opioids as a means of pain relief.

 At least 40 states now have operational PMPs or have enacted PMP legislation, covering 87% of the US population,

 and the White House plan has committed federal resources to expanding them.

PMPs collect data from pharmacies on controlled substance prescriptions, including information on who has been prescribed what substance and by whom. These data can be provided to prescribers, pharmacists, law enforcement officials, professional licensing boards, public health researchers, and others in order to identify individuals engaged in prescription drug diversion and to perform research in this area. Even early versions of today’s PMPs have been shown to reduce prescription drug diversion; for example, implementation of a PMP in New York State in 1978 reduced the percentage of forged or counterfeit schedule II prescriptions from 12% to less than 1% within 5 years.

 Payers can promote the use of this important resource by strongly encouraging, facilitating through technology, or even requiring prescribers to check the PMP if available, before prescribing controlled substances. In the future, they could make it easier for time-challenged physicians to use these data by linking the PMP to their electronic prescribing tools, or to pharmacy computers at the time of dispensing. Payers in the public sector can also use PMP data to investigate whether clinicians’ prescribing patterns meet the standard of care in a given community, and to identify geographic areas where doctor shopping and adverse outcomes are higher than average and warrant closer surveillance.

 Current guidelines encourage practices to engage the patient receiving RxOs, as part of the risk management process,

 including patient counseling, informed consent, and treatment agreements that outline patient expectations and responsibilities regarding opioid therapy.45 Poor patient knowledge and lack of communication between patient and clinicians have been recognized as barriers to effective RxO utilization.

 Patient education is time-consuming. Supportive reimbursement policies could play a role in facilitating the development of time-efficient and effective approaches to patient education.

To control costs, clinicians trained in pain management

can provide education, support, guidance, and monitoring of patients on opioid analgesics. A model is diabetes educators who provide medication education and monitor for compliance, side effects, and tolerability in individual or small support group settings. This approach was shown to improve disease status, increase cost-effectiveness, and reduce lengths of hospital stays.

 Following this model has the potential to improve communication and patient awareness of risks and benefits of pain therapy. A recent Veterans Administration study on the role of nurses in pain management found that nurses with more experience and higher levels of self-efficacy and confidence were more likely to address patient fears of addiction, supporting the use of trained nurses to counsel patients.

Payers could be instrumental in enhancing nurse training in RxO abuse and addiction.

UDM includes testing for illicit drugs, confirming use of the prescribed RxO, and detecting non-prescribed medications and controlled substances with potentially dangerous interaction with RxOs. UDM is considered an objective indicator of drug abuse or diversion that complements patient self-report, physician assessment, and behavioral monitoring.

 To this end, many organizations recommend UDM in their guidelines for RxO management in patients with chronic pain, though the effectiveness of UDM is not fully proven. It has been estimated that clinical methods alone (without laboratory corroboration) miss about half of patients who are misusing opioids.

A review of studies using UDM in 2007 to 2009 showed that at least 11% of patients with chronic pain were misusing opioids, and at least 12% were not adhering to a prescribed medication.

However, no high-quality studies have examined whether UDM has an impact on opioid abuse, addiction, and/or overdoses. A recent systematic review found only 8 fair to poor quality studies of UDM that examined these outcomes, and these studies showed a modest 7% to 23% reduction in abuse in the context of multicomponent monitoring programs.

 Payers could set standards for use (yet to be determined) of UDM in their reimbursement policies. Recommendations regarding who to test, how often to test, and what types of tests to use could be modeled on some of the most recent consensus recommendations.

 It is also likely that UDM will need to be limited to specific patients and circumstances; thus, payers will need to be flexible in their policies. It remains to be seen how this important tool will be used most effectively.

When patients test positive for an illicit drug or an unexpected opioid, health insurers should support clinicians in providing affordable access to substance abuse treatment services to supplement medical care. While discharging patients from practice is sometimes necessary, addicted patients who are discharged will tend to seek prescription medications in another practice or in EDs, a behavior that simply moves the problem but does not address it. Opioid agonist treatment, such as with methadone or buprenorphine, is effective at treating opioid dependence and is increasingly utilized for RxO addiction. More research is needed to identify the most effective strategies to prevent and treat addiction to RxOs that is identified during treatment for chronic pain. The National Drug Abuse Treatment Clinical Trials Network recently completed a trial that provides important evidence regarding the length of pharmacotherapy, effects of intensive counseling, and role of chronic pain in modulating the effectiveness of treatment for RxO addiction.

 This was the largest study ever examining treatment for this population, and results are expected soon. Health insurers and government entities have an important role in ensuring access to effective and accessible addiction-treatment services.

 There are several potential barriers to implementation of recommendations. Among prescribers, perceived logistical barriers include lack of time, limited ancillary support, lack of appropriate pain management and addiction referral options, limited information on diagnostic workup, and limited insurance coverage for pain management services.

 In primary care settings, financial constraints are of particular concern. Pressure on reimbursement and financial constraints may limit the time and support staff devote to opioid risk management. Physicians may opt not to treat pain if it is perceived that responsibilities surrounding opioid prescribing are too time consuming.

 Payers can relieve some of these barriers by increasing reimbursement for pain management services and ancillary support; providing automated risk information, PMP data, urine drug testing results, and claims data for adverse patient outcomes (eg, ED visits) integrated into the patient management work flow; and providing practical training on managing chronic pain, opioid use, and its complications (

MANAGERIAL

This article outlines strategies insurers can use to mitigate their risks related to prescription opioid abuse by members, while addressing this serious public health problem.

Prescription Opioid Abuse: Challenges and Opportunities for Payers

To estimate the long-term comparative effectiveness of first-line treatment in patients negative for anti-JC virus (JCV) antibodies with glatiramer acetate (GA), fingolimod, or natalizumab for relapsing-remitting multiple sclerosis (RRMS).

 We developed a simulation model to estimate the average 20-year clinical risks and benefits of GA, fingolimod, and natalizumab for RRMS patients initially negative for anti-JCV antibodies.

Model inputs included published natural history progressions of the Expanded Disability Status Scale (EDSS), treatment effects from randomized controlled trials on slowing disease progression and reducing relapse rates, risk of progressive multifocal leukoencephalopathy (PML), and utility preference scores. Outputs were long-term risks (PML risk and other non-PML risks), benefits (average relapse rate and time to disability [EDSS 

7]), and quality-adjusted lifeyears (QALYs).

 Compared with GA, natalizumab resulted in 4.6 fewer relapses, 0.6 more years of disabilityfree time, 0.0165 more cases of PML per treated patient, and an incremental 1.2 QALYs gained. Compared with fingolimod, natalizumab resulted in 1.7 fewer relapses, 0.1 more years of disabilityfree time, 0.0165 more cases of PML per treated patient, and an incremental 0.4 QALYs gained. The probability that incremental QALYs favored natalizumab over GA was 0.963 and natalizumab over fingolimod was 0.720.

Average QALYs, a measure that aggregates across risks and benefits, favored natalizumab, suggesting more aggressive early intervention with natalizumab in the negative anti-JCV population. For certain decision makers, more evidence may be needed to further reduce the uncertainty in these comparative projections prior to making population-based adoption decisions.

Long-term clinical benefits were favored by natalizumab with minimal increased risks in the negative anti-JC virus population.

		Average quality-adjusted life-years, a measure that aggregates across risks and benefits, favored natalizumab, suggesting more aggressive early intervention with natalizumab in the negative anti-JC virus population.

		 In the absence of long-term trials, we used traditional and standardized methodology from the cost-effectiveness field to quantify comparative effectiveness research goals.

		 Comparative effectiveness models like this may be used to provide projections about long-term risks and benefits in many different patient subpopulations, estimate the uncertainty in the comparative findings, and guide the design of future evidence generation.

Multiple sclerosis (MS) is a chronic and morbid condition of the central nervous system. Recent US approval of disease- modifying therapies (DMTs) including fingolimod and natalizumab for relapsing-remitting MS (RRMS) provide new options for the management of the disease. The ultimate treatment goal of any DMT is to delay or prevent the long-term disability of MS while minimizing DMT-related risks. This long-term treatment goal often remains untested in randomized controlled trials.

Randomized controlled trials of DMTs in RRMS lack evidence of long-term clinical benefits and risks, as follow-up periods are typically limited to 1 to 2 years. Markov models are commonly used to project the long-term cost-effectiveness of DMTs for RRMS.

Thompson and colleagues used a Markov model to examine the long-term risk-benefit or comparative-effectiveness of natalizumab as compared with interferon beta-1-alpha prior to available evidence on the risk stratification of patients treated with natalizumab.

Natalizumab treatment based on post-marketing surveillance is associated with a risk of progressive multifocal leukoencephalopathy (PML) of 2.13 per 1000 patients treated per year.

 PML is often a fatal viral disease characterized by progressive inflammation and damage to the white matter.

PML is caused by a pathogenic form of the JC virus (JCV).

 The increased risk of PML has limited natalizumab’s use as a first-line DMT in RRMS. A 2-step assay was developed for detecting and confirming the presence of anti-JCV antibodies in human serum and plasma.

 As described in the January 2012 revised natalizumab US Food and Drug Administration (FDA) label,

this assay can stratify the risk of PML for prospective natalizumab patients where the majority of PML risk falls on approximately 55% of RRMS patients who test positive for anti-JCV antibodies.

 The ability to stratify prospective natalizumab patients by JCV status gives rise to the following question: should natalizumab be considered first-line therapy for JCV-negative patients?

There is a need to fill critical gaps in the comparative effectiveness of DMTs for RRMS to facilitate clinical decision making. Available evidence suggests the ability to minimize PML risk for natalizumab by targeting negative anti-JCV antibody patients. Further, the long-term risks and benefits associated with natalizumab treatment should be compared with other DMTs, including fingolimod, the first oral DMT approved by the FDA in 2010. Our objective was to estimate the long-term comparative effectiveness of first-line treatment in RRMS patients initially negative for anti-JCV antibodies with glatiramer acetate (GA), fingolimod, or natalizumab.

We created a Markov simulation model spreadsheet that followed good practices for decision analytic modeling in health technology evaluations.

 We used this model to project the long-term clinical benefits, risks, and quality-adjusted life-years (QALYs) of a cohort of RRMS patients negative for anti-JCV antibodies who were DMT treatment—naïve and had begun treatment with GA, fingolimod, or natalizumab (

). We selected GA as a representative of standard first-line agents because it is a market leader in the United States and 3 head-to-head trials comparing GA with interferons suggest no significant difference in clinical efficacy.

We simulated a cohort of patients who progressed through health states defined by the Expanded Disability Status Scale (EDSS).

 Each year, patients could remain in their current EDSS state, progress to the next EDSS state, experience PML (for natalizumab-treated patients), or experience death due to all other causes. EDSS states were in 0.5 increments between 0 and 10 (no EDSS state of 0.5 existed and an EDSS of 10 was death due to MS). In all EDSS states, patients could experience an MS relapse or not experience a relapse during each yearly cycle.

This research was considered exempt by the Colorado Multiple Institutional Review Board.

• During a 1-year cycle, patients could not improve in terms of their EDSS state (patients could not transition to a lower EDSS state).

• Patients were measured over a 20-year time period (consistent with a prior risk-benefit model).

• The average starting age was 30 years (consistent with DMT clinical trials and prior models).

• There was a 0% discount rate for future years of life (cost-effectiveness models typically discount future costs and clinical consequences at 3% per year in order to estimate the net present value of costs and consequences). Since our objective was to project the findings in clinical terms (comparative effectiveness, not cost-effectiveness) over the 20-year time horizon, a 0% discount rate was warranted. The 0% discount rate is synonymous with projecting 20-year survival (with quality adjustment) and therefore holds more face validity than discounting survival for this clinical objective.

• The average clinical efficacy for patients negative or positive for anti-JCV antibodies would be the same as observed from clinical trials.

• There was a 2% annual seroconversion risk of JCVnegative to JCV-positive status over the 20 years,

and constant annual risk of PML from years 2 to 20 (PML incidence is shown to be constant from 2 to 6 years of natalizumab exposure time).

• There was no prior immunosuppressant use in those that seroconverted to JCV-positive status (immunosuppressant therapy in combination with JCV-positive status is associated with higher risk of PML, but 90.4% to 97.3% of patients followed in large MS cohort studies do not have prior immunosuppressant use).

• There was no additional risk of cardiac or other death risk related to fingolimod.

• Equal effect on relapse rate and EDSS changes regardless of relapse rate or EDSS level at outset, ie, clinical effectiveness is the same for individuals with less or more severe disease.

• Continuous treatment with each agent, and no switching to an alternative agent over time, regardless of clinical change.

The London, Ontario, Canada, MS cohort study tracked the 25-year progression of RRMS patients prior to DMTs (best supportive care only).

Best supportive care in the London, Ontario, cohort was defined as symptom control, physiotherapy, psychiatric and social support, and disability aids. A patientlevel analysis of the London, Ontario, MS cohort was done by Tappenden and colleagues for the purposes of estimating EDSS initial state distribution, as well as the EDSS transitions for best supportive care.

 Tappenden and colleagues combined the information from the London, Ontario, cohort on best supportive care with other literature sources to estimate the cost-effectiveness of beta interferons and GA compared with best supportive care.

 We used Tappenden’s EDSS initial state distribution for all comparators, as well as the EDSS transitions for best supportive care as a foundation for EDSS transitions of DMTs added on to best supportive care, 

The initial EDSS state distribution was consistent with DMT RCTs for GA, fingolimod, and natalizumab. Approximately 88% of patients in the model started in EDSS states 0 through 2.5 (minimal to mild disability) for all DMT scenarios. The average EDSS scores in previous RCTs of DMTs are in the 2 to 3 range.

 We used the National Center for Health Statistics estimates for the annual gender- and agespecific all-other cause mortality.

We used relapse rates specific to EDSS state for best supportive care that were consistent with Tappenden and colleagues from the Patzold data

9). Patzold and colleagues reported relapse rates for best supportive care over a 19-year time period and showed that relapses decreased with increased disease exposure. We tested the validity of the modeled relapse rates by comparing the 1- and 2-year projected average annual relapse rates of best supportive care with that of the placebo arms of a natalizumab RCT.

DMT Projections—Modeling Treatment-Specific Risks and Benefits

We compared 3 DMT strategies: GA 20-mg subcutaneous administration daily, fingolimod 0.5-mg oral (pill) administration daily, and natalizumab 300-mg intravenous administration every 4 weeks. We used the relative changes in relapse rates and EDSS progression from the pivotal RCTs for GA versus placebo,

). Due to a concern that the underlying risk of relapses may have decreased over calendar time, we modeled relative changes in risks because relative changes are thought to be more robust across ranges of underlying risk.

The annual risk of PML per natalizumab-treated patient with negative anti-JCV status was 0.00009 (95% confidence interval [CI], 0.000001 to 0.00048), whereas natalizumab-treated patients that seroconverted to positive anti-JCV status (and were assumed to have no prior immunosuppressant use) had an annual risk of PML of 0.00056 (95% CI, 0.00036-0.00083) for <2 years of natalizumab exposure and an annual risk of PML of 0.0046 (95% CI, 0.0037-0.0056) for 

2 years of natalizumab exposure (Table 1). We tested the average risk for patients regardless of JCV antibody or immunosuppressant use status as a sensitivity analysis (2.13 per 1000 patients treated per year).

CLINICAL

We estimated the long-term risks and benefits of disease modifying therapies. Benefits were favored by natalizumab with minimal increased risks in the negative anti-JC virus population.

Comparative Effectiveness of Early Natalizumab Treatment in JC Virus-Negative Relapsing-Remitting Multiple Sclerosis

Making the healthcare system more affordable will determine whether employers can continue to provide coverage, whether individuals can purchase it, and whether important public programs can be sustained. Health plans are playing a vital role in reducing the cost of care and improving value by changing how they pay providers. At the same time, they are creating new benefit designs to encourage patients to choose high-performing clinicians and hospitals and take advantage of care coordination and case management for chronic conditions. This article describes health plans’ innovations, while also teeing up a policy agenda that would advance these changes.

Health Plan Innovations in Delivery System Reform

Health plans are redesigning payment mechanisms to move away from the practice of rewarding volume through fee-for-service payments and toward encouraging better outcomes and improved efficiency through accountable care organizations (ACOs), patient-centered medical homes, and bundled payments. In moving away from retroactive payment to a prospective design, these new models are built on accountability, shared risk, and population-based care.

At America’s Health Insurance Plans (AHIP), our experts have been mapping these developments and will be releasing a new state-by-state compendium of activities around the country. In addition, over the past 18 months, we have convened 3 invitational summits bringing together health plans and their provider partners to discuss how they have restructured their payment contracts, outline key features of their programs, and present results.

What we are not seeing are cookie-cutter, 1-size-fits-all approaches. Rather, payment redesign is being structured and calibrated across the country to take account of providers’ readiness to accept financial risk. Unlike previous attempts at payment redesign, these current efforts are aimed at tackling both primary and specialty care. While they are at various stages of development and implementation, 2 distinct features are fundamental to the new models being launched across the country: there is strong evidence of collaboration between health plans and their provider partners and both quality performance and cost reduction goals are being negotiated. This allows health plans to engage in meaningful population-based measurement and providers to have confidence that performance metrics are transparent and fair.

Another major change is that health plans are redesigning benefit structures at the same time they are changing payment mechanisms. These changes are designed to work synergistically to reward providers for achieving results, while also rewarding patients for making choices to use higher-performing hospitals and physicians and regularly obtaining services that are crucial for chronic care management. Strategies advancing either payment restructuring or benefit design cannot work optimally if they are working alone. To maximize results, they need to be aligned and coordinated, and health plans are in a unique position to make that happen.

In building new payment models, health plans are offering their provider partners more data, as well as decision-support tools. These data help physicians recognize gaps in care, such as which patients need comprehensive case management, which patients are most at risk of developing serious conditions, and which are in need of immunizations and preventive care.

From our research, we have noted several characteristics that are present in today’s plan-provider collaborative models that are yielding promising results. Buy-in for these new arrangements must start with leadership. Clinical integration, a culture of initiating change, a robust health information technology (IT) infrastructure, and acceptance of new payment arrangements are all key criteria.

 In addition, a relationship of 3 or more years is critical to achieving efficiencies among all partners.

Preliminary data suggest that new private-sector ACO models are off to a strong start, with initial quality improvements of approximately 10%; a 15% decrease in readmissions and total inpatient days; and an initial annual savings of $336 per patient.

Health plans are fundamentally transforming the delivery of healthcare for patients by reforming payment models, coordinating care, and partnering with providers.

April 2013

Effects of Coverage Gap Reform on Adherence to Diabetes Medications

Prescription Opioid Abuse: Challenges and Opportunities for Payers

Comparative Effectiveness of Early Natalizumab Treatment in JC Virus-Negative Relapsing-Remitting Multiple Sclerosis

Health Plan Innovations in Delivery System Reforms

Disease Management Programs: Barriers and Benefits

Use of Web-Based Shared Medical Records Among Patients With HIV

Retail Clinic Utilization Associated With Lower Total Cost of Care

More Comprehensive Discussion of CRC Screening Associated With Higher Screening

Primary Nonadherence to Statin Therapy: Patients' Perceptions

Trends in Hospital-Based Childbirth Care: The Role of Health Insurance