 One of the main goals of the Affordable Care Act (ACA) is to control the costs of US healthcare. Channeling patients toward more effective services is one of many approaches being used to control costs while improving health outcomes. This paper reviews value-based insurance design (VBID) concepts and discusses options for states to encourage these designs in the new health insurance exchanges (HIEs).

 We reviewed the literature on VBID as well as the text of the ACA for descriptions of how VBID might be encouraged through the new state health insurance exchanges.

 States, under healthcare reform, are allowed to promote the use of VBID designs in their exchanges. There are 4 broad approaches a state HIE could pursue with regard to VBID, ranging from establishing a process for recommending high- or low-value services and requiring plans to adhere to the recommendations, to offering no guidance to plans. The evidence surrounding how well VBID designs work is growing, but it is still limited. To date there is no evidence that reducing or eliminating copays for preventive services cuts costs in the long term. However, modeling does suggest the potential for such long-term savings,so states should proceed with caution.

Modifying copays, even in small amounts, can send signals to patients about the relative value of drugs and services. However, long-term savings will likely result from higher copays on low-value services. The leadership of each exchange has a unique opportunity to reshape the insurance benefit landscape in its state to improve value and invest in prevention.

Value-based insurance designs (VBIDs) alter copay structures to promote high-value medical services and treatments while discouraging the use of low-value services. Over time, it is possible that VBID will become an important method for changing incentives within the healthcare system in an effort to control growth in healthcare costs. States have several options for promoting these designs within the new health insurance exchanges.Bending the cost curve in American healthcare is a key goal of the Patient Protection and Affordable Care Act (ACA, 2011).

 Projections on the actual dollar amounts saved vary across sources, but the ACA’s emphasis on reducing fraud and abuse in public programs, uniform insurer standards to reduce paperwork, and taxes on generous health insurance benefits are all sources of cost  savings.

 Payment reforms on both the insurer and the delivery side are also key drivers of expected savings.

In implementing exchanges and certifying which plans can participate, states and their health insurance exchanges have the  potential to change the landscape of insurance benefit design. The major reforms to the individual health insurance market involve changing the underwriting practices and establishing state-based health insurance exchanges. Exchanges will function as individual and small group marketplaces for a menu of plans within 4 tiers of coverage with increasing benefits. States may set up their exchanges as extensions of existing government agencies, as nonprofits, or as independent government agencies.

Over 24 million new enrollees are expected to join the exchanges by 2021, so controlling costs will become ever more important.

 Improvements in healthcare costs mostly come from the use of effective services, while some authors note that increasing costs come from services with little value, or value for only some patients.

 Providing incentives to encourage consumers to use high-value services and discourage lowvalue services has been promoted as one way of controlling costs. This strategy is otherwise known as value-based insurance design (VBID). While there is a growing national policy push to adopt VBID designs,

 there has been little attention to date regarding how state health insurance exchanges (HIEs) could or should be involved in promoting VBID for those it will be covering, and whether major cost savings will be achieved.

The exchanges will function as marketplaces for individual and small group health insurance policies. In a central location,  consumers will be able to compare the prices of plans and benefits.

 States have the option of creating separate pools for the individual and small group markets, or combining them. States can also choose to enter into regional exchanges. If states choose not to establish an exchange, they can partner with the federal government or let the federal government run the exchange altogether. In all, 27 states have decided to let the federal government operate their exchanges.

Within the exchanges, 4 tiers of plans will be offered. The tiers are based on actuarial value, which is the amount the plan expects to pay out in medical costs while the other percentage is usually covered by the enrollee. The exchanges will then offer 4 tiers of plans: bronze, silver, gold, and platinum. The tiers increase in actuarial value from 60% to 90%. Categorizing the plans according to actuarial value is designed to make the relative value of the plans more transparent.

 Exchanges can allow any plan wishing to participate, or can select certain plans based on predetermined criteria.

The ACA’s largest boost to VBID is the recommendation of an array of preventive services that must be offered with no cost sharing for all new plans beginning in 2010. The mandated list comprises all of the US Preventive Service Task Force’s A and B recommendations, as well as additional women’s and children’s health services.

 This is an example of VBID helping to signal the relative value of services and treatments to consumers.

VBID discussions in the literature may not always distinguish between designs using different cost-sharing levels as a mechanism to motivate the use of specific services, from other designs using incentives to promote health-related behavioral change. The Affordable Care Act gives employers and insurers more leeway to expand the use of this second type of incentive-based structure to encourage healthier behavior. Under the bill, employers and insurers may offer up to 30 percent of the value of the premium to a consumer as an incentive for meeting certain health goals such as smoking cessation, up from 20 percent currently (§ 2705).

The bill also allows Medicaid programs to experiment with behavioral incentive programs (§ 4108).

Both types of interventions may use financial incentives to influence patient decisions, but incentives for changes such as weight loss or smoking cessation are not focused on individual types or classes of medical services, and are there fore not considered part of the definition of VBID.

Although of interest and worthy of future study, we cite these ACA initiatives to distinguish them from the discussion of the VBID incentives below.

Value-based designs alter incentives for patients to curtail or increase use of specific types of medical services. They are rooted in the health economics literature, which has shown consumers to be consistently responsive to the level of cost sharing for medical services. The best known study of the impact of cost sharing on utilization is the RAND Health Insurance Experiment.

 Although this federally funded study occurred several decades ago, its original intent was to provide input into the design of benefit levels for a national health program such as the ACA. Results showed that higher cost sharing was associated with reduced use of healthcare services, but that patients were just as likely to reduce the use of necessary as unnecessary services.

More recent literature on cost sharing details how these same conclusions from the RAND experiment can be used to encourage prevention of downstream complications in chronic disease.

 Most of the recent VBID initiatives have focused on reducing cost sharing for chronic disease maintenance medications. These policies have usually focused on specific patient subgroups to increase the efficiency of the design.

 For example, Chernew et al, examined copay reductions for the management of chronic diseases in 6 classes of drugs. The employer plan reduced copays on statins, inhaled corticosteroids, angiotensin-converting enzyme inhibitors, angiotensin receptor blockers, beta-blockers, and diabetes medications. The authors found a 7% to 14% reduction in nonadherence to drugs in these classes when copays were reduced 50% for preferred and nonpreferred tiers and to zero for generic drugs.

In a separate study, a large Blue Cross Blue Shield plan eliminated or lowered copays for drugs in 8  classes of medications and adherence rates improved by 2 to 4 percent in the first year of the program.

 Additionally, Choudhry et al found that dropping the copays on prescription drugs for clopidogrel and statins led to 4% and 2.8% increasesin adherence.

While these copay changes to incentivize the use of certain technologies show improvement in some process indicators, they have not yet achieved cost savings.

 For example, the same Blue Cross Blue Shield study found no cost savings.

Another recent article by Choudhry et al (2011) found no significant difference in costs between a group with no copay for cardiovascular drugs and a group with regular cost sharing. 

 The lack of significant cost savings may be due to the short follow-up times in these studies —usually 1 year. More research is needed in the area of cost savings of incentivizing the use of certain services. This will become especially important as the ACA has also required that insurers provide all the USPSTF recommendations free of charge.

In lieu of long-term follow-up, modeling suggests some long-term savings, but these analyses have often focused on raising copays for “low-value” services. For example, the 3 main treatments for prostate cancer vary greatly in their average costs with no evidence that the more expensive treatments result in better outcomes.

 A radical prostatectomy costs $7300, brachytherapy costs $19,000, and radiation therapy costs $46,000 on average. Newer forms of radiation treatment can cost close to $100,000 per case, and have not been shown to have any clinical advantages over any of these less expensive options, including watchful waiting. A simple VBID policy would be to modestly increase the cost sharing for these services to encourage more use of the cheapest and equally effective prostatectomy. The authors of the same prostate treatment study estimate $1.7 to $3 billion could be saved directing patients toward the lower-cost treatments.

To date, there has been little policy work offering guidance on whether or how states should encourage VBID in their HIEs. The Robert Wood Johnson Foundation published a brief on the possible actions HIEs could take and identifies VBID as part of a state’s possible role in active purchasing.

 In selecting plans for participation, an HIE could give special priority to contracting plans applying VBID strategies. While Section 2713 promoted VBID, the preliminary HIE regulations promulgated by the Department of Health and Human Services in July 2011 did not offer substantial guidance for states on implementing value-based designs.

State health exchanges are faced with a spectrum of possibilities. At one extreme, a state could establish mechanisms to identify high- and low-value services, and then mandate that all participating health plans apply different cost sharing levels to the services in these categories. The other extreme is to do nothing, and let insurers continue experimenting with VBID as they wish. Overall, the latitude state health exchanges have in adopting any of the 4 options depends on what level of power state lawmakers give their exchanges to selectively contract with plans. As of January 2013, 7 state exchanges, such as California and Rhode Island, have opted to be “active purchasers,” allowing them more power to control which plans can offer insurance through the exchanges.

Six states, including Colorado, Utah, and Hawaii, opted to organize the exchanges solely as a clearinghouse.

After the November election, many state lawmakers who had been opposed to the ACA opted to allow the federal government to operate their state’s exchange. It is unclear at this juncture whether the federally run exchanges will be involved in promoting certain types of benefit designs or whether they will allow any willing insurer to participate for these 27 states.

Option 1: Require Plans to Structure Copays According to State-Defined Highand Low-Value Services

Under this option, the state would set up a governing body to decide what services would be covered with reduced or increased cost sharing for plans entering into the exchanges. The advantage of this approach is that it would provide consistency with regard to services considered high or low value across VBID plans. Defining the services up front would prevent the plans from structuring copays in ways that might lead to risk selection, such as setting higher costs for certain drugs.

There is some precedence for state involvement in valuebased service determination. For its state-sponsored insurance programs, including Medicaid, Oregon compiles a detailed list of services it deems of high value, and covers as many of the services as possible with the funding allotted in a given year (

). The Oregon Health Plan ranks medical services “in a way that represents the comparative benefits (ie, clinical effectiveness and cost-effectiveness) of each service to the entire population to be served.”

Oregon’s 11-member Health Services Commission has an established process for selecting covered services, starting by ranking clinical areas from highest to lowest benefit. Common treatments and procedures are assigned to the clinical categories ranging from maternity and newborn care (category 1) to “inconsequential services” (category 9).

 These categories are then combined with a series of metrics to derive an overall score: clinical effectiveness, population impact measures, the need for the service, and the net cost (

Having a state agency regulate VBID for private insurers is a politically tough option as there are substantial technical and analytic challenges associated with identifying both highand low-value services. The US Department of Labor recently held an open comment period on VBID, and many patient groups, disease-specific advocacy organizations and even local health departments voiced concerns that VBID would be used as a way to limit access to medically necessary services.

Decisions over the appropriate services to include as high or low value are not without controversy. For example, the USPSTF recently recommended delaying regular mammograms for women until after age 50 years due to the high rate of false positives.

This could leave many women facing higher copays on mammograms conducted at earlier ages in a VBID policy. Oregon’s solution to the change in recommended screening frequency was to cover mammograms for all women aged 40 years and over if the physician advised it. After age 50 years, the state suggested that mammograms should be done every 2 years.

There are several drawbacks to setting up a similar process elsewhere. The first is in timing. Oregon developed its method over a decade and a half, making it difficult to replicate elsewhere in a short period of time.

 There are also potential adverse consequences for states deciding to maintain small group or individual markets outside of the exchanges. Requiring some plans to offer certain services for free could raise the cost, forcing some people out of the exchange. However, the subsidies offered to lower premiums would likely make this result unlikely. More importantly, in order to curb any adverse selection in the exchanges, health plans are required to have similar premiums for similar coverage in and outside of the exchange.

Alternative 2: Require Insurers to Design Their Own VBID Plans

Another approach to promote VBID would be to require HIE-participating plans to offer insurer-created value-based designs. The major advantage of this option is allowing flexibility for insurers. In the Department of Labor’s open comment period on value-based designs, the Blue Cross Blue Shield Association commented that because the field is still emerging, insurers need the flexibility to innovate.

 Researchers at the University of Michigan’s Center for Value-Based Insurance Design also urge flexibility, arguing that mandating too many benefits could limit an insurer’s ability to offer plans in the lower-cost bronze tier.

 If insurers were required to offer too many benefits free of charge (as in Alternative 1), then it could be very hard for the insurer to offer the basic benefits plans that would be part of the exchange’s bronze tier—the 60% actuarial value tier.

Alternative 3: Incentivize/Encourage Insurers to Offer VBID Plans

Incentivizing insurers could promote VBID without facing the potential negative political consequences of mandating non-coverage of low-value services. The University of Michigan’s VBID Center offered a brief piece of guidance on this issue. The authors argue that states need to allow for flexibility in setting up the plans, that states should not set uniform cost-sharing levels, and also suggested that VBID could represent a plan quality performance measure.

 For example, a state could consider whether a plan incorporated VBID as one quality indicator when evaluating which plans are qualified to sell in its exchange. In such a case, plans using VBID could be rewarded with extra “points” in a request for proposal process. Maryland’s lawmakers took this route when writing the legislation enacting the state’s exchange board, indicating that VBIDs could be used as 1 criterion for plans to be selected into the exchanges.

 Exchanges could also encourage the selection of particular services for inclusion in VBID designs. Items a state could incentivize might be zero copays on generic drugs for treating chronic diseases. Massachusetts issues requests for proposals for insurers wanting to participate in its exchange,and scores them on how well they do at providing consumers “good quality and value, according to standards set by the Connector.”

 Plans receive better scores if they have “innovative pharmacy management” and incentives for wellness.

One final example of incentive-based encouragement is the quality ratings the Centers for Medicare & Medicaid Services (CMS) uses for assessing Medicare Advantage plans. The health reform bill allows CMS to make incentive payments to Advantage plans if they meet certain quality benchmarks.

 For example, one of the metrics evaluates the plans’ ability to have patients up-to-date on screenings and vaccinations. While not directly linked to value-based designs in terms of reduced cost sharing for certain services, this program does serve as a possible template in encouraging the use of more highly valued services.

As an alternative to direct incentives, states could also encourage insurers to offer VBID plans in the exchanges. The closest existing example for this approach would be Oregon’s VBID recommendations. Oregon has 2 separate pieces to its VBID processes (see 

). There is the formal priority setting that makes the list of covered services and establishes the varying levels of copays for all of the public plans under the state’s control, as mentioned above. It then created a separate list from this process, which comprises recommended highvalue services for the commercial market. The commercial plans are not required to adhere to the recommendations.

Oregon’s approach for commercial payers is a relatively hands-off one for selecting certain services for reduced copays as part of the plans, but does take advantage of Oregon’s research-based priority setting process for identifying services for reduced copays. While evidence is scarce on whether the private plans in Oregon take up the recommendations, the public employee plans do take the suggestions into account.

The boards overseeing these plans chose to both raise copays on certain undervalued services like emergency department visits and lower them on several highly valued services such as vaccinations. The program was adopted in 2010 and has yet to be evaluated. Oregon’s experience represents a way states could encourage VBID plans.

Alternative 4: Offer No Guidance to the Plans

State exchanges could remain silent on the issue of VBID, which would let insurers decide for themselves what types, if any, to offer beyond what is required in the ACA. This approach would presumably result in limited adoption of VBID unless interest in this model increases substantially. However, many insurers and employers are already experimenting with different types of incentives. In a survey of 1300 large employers, Choudhry et al found that many benefit managers are thinking of implementing some version of valued-linked incentives, either of the behavioral change type, or the reduced copays for high-value medical treatments, or both.

 The authors found that while less than 20 percent of employers in 2007 were currently using some type of incentive, over 81 percent planned to in the next 5 years.

The ultimate impact of VBID designs on health plan premiums has not yet been clearly determined, in part because there is so little well-evaluated experience with initiatives raising cost sharing for low-value services. These are likely to generate most of the savings, but face more implementation challenges. Neumann and colleagues reviewed issues associated with increasing copays for “low-value” services.

 These interventions are much more controversial, both because of the added financial burden on patients and also because it is technically challenging to determine with high confidence which services are low value.

Additional challenges have prevented insurers from widely adopting these designs, including a lack of evidence of what works, and difficulty in determining which services are high or low value.

 These challenges are followed closely by concern over the effective targeting of incentives to the right populations. For example, dropping the copay on all hypertension drugs means the insurer will spend more money on the patients who are already adherent, just to gain improved outcomes for the subset who are not. The lack of continuous enrollment and high turnover rates in US insurance markets means health plans have limited incentives to invest in preventive or behavior-changing initiatives if the enrollees leave the employer or plan before the benefits accrue.

Providers may also object if investing in better preventive care or care coordination means extra expense for them. In an Institutes of Medicine report, providers cited the lack of reimbursement for promoting valuable services or discouraging low-value ones.

 In the report, providers urged a focus on long-term outcomes for patients, such as managing chronic illnesses, as part of an overall agenda of paying for quality. The report also suggests another aspect of VBID could be reducing the overuse or misuse of services.  Separate from helping patients navigate what services will provide the most benefit at the lowest cost, providers also face their own challenges with VBID when plans choose to “tier” hospitals and physicians according to some kind of quality metric. Some initial results from Massachusetts’ experience in this area have shown the copay signals can be counteracted with information from other sources, so more research in this area is needed.

VBID approaches may have the potential to lower healthcare costs, but they must also be designed to account for individual patient needs.

 In some cases, a given service could be of little clinical benefit to one subgroup of patients, but of higher value to patients with a different set of characteristics. Many patient advocacy groups criticized the use of value-based designs in letters written to the Department of Labor during the organization’s comment period on VBID. The groups argue the designs will be used to limit access to certain treatments.

 Using the previous brachytherapy example, patient heterogeneity must be allowed to be a part of the shared decision making between patient and physician. The advantage of value-based designs, however, is to indicate information about what is high and low value, not limit access to these treatments altogether.

Articles

POLICY

Encouraging Value-Based Insurance Designs in State Health Insurance Exchanges

Because laboratory test results are less available to researchers than claims data, a claims-based indicator of diabetes improvement would be valuable.

 To determine whether a decrease in medication use for diabetes parallels clinical improvement in glycemic control.

 This was a retrospective cohort study using up to 3.5 years of pharmacy and laboratory data from 1 private insurer. Data included 104 patients with diabetes who underwent bariatric surgery and had at least 1 glycated hemoglobin (A1C) test before and after surgery.

 We assigned each A1C test to a 90-day interval before or after surgery. Medication availability was noted for the midpoint of the interval (on insulin, on oral medications, count of medications). Each subject could contribute 1 presurgery and up to 3 postsurgery observations. We recorded the changes in A1C test results and medication use from the presurgery to the postsurgery period. Using the A1C test as the reference standard, positive and negative predictive values of medication-based indicators were calculated.

 After bariatric surgery, A1C test values decreased by more than 1% and the count of unique medications decreased by 0.6. All 3 medication-based indicators had high positive predictive values (0.85) and low negative predictive values (0.20), and count of medications had better performance than the other indicators.

Without clinical information, a decrease in use of medications can serve as a proxy for clinical improvement. Validation of results in other settings is needed.

Diabetes is a major health concern in the United States given its high prevalence and significant morbidity. Many interventions have been implemented to improve glycemic control, especially those targeting weight loss.

 Glycated hemoglobin (A1C) is the standard method of monitoring diabetes mellitus control, including responses to treatment.

 In health services research, clinical data often are not available to researchers, particularly in studies that use administrative claims for outcomes evaluation.

Claims data have been used to describe the outcomes of interventions or as a method for risk stratification among individuals with diabetes.

 However, it is hard to know whether glycemic control has improved without access to laboratory test data. Therefore, in this study we determined whether a decrease in the use of medications for diabetes parallels clinical improvement in glycemic control among individuals with substantial weight loss. We used individuals with diabetes who had undergone bariatric surgery, as they are expected to have improvement in A1C after surgery.

 We hypothesized that a decrease in medication use from the presurgical to the postsurgical period would parallel an improvement in A1C values over the same period.

This was a retrospective cohort study using up to 3.5 years of pharmacy and laboratory data (up to 1 year before and 2.5 years after bariatric surgery).

We accessed claims and laboratory data from 1 Blue Cross Blue Shield plan, as part of a large ongoing collaborative effort. We received data on all members who met at least 1 of the following inclusion criteria at any point during 2002 through 2006: (1) completed a health risk assessment that captured member height and body weight; (2) had an obesity diagnosis reported in aclaim; (3) had a paid or denied claim for bariatric surgery; (4) had a paid or denied claim for a medication for promoting weight loss; or (5) were more than 12 years of age and had a diagnosis of hyperlipidemia, type 2 diabetes, sleep apnea, gall bladder disease or surgery, or metabolic syndrome. These diagnoses were identified in the claims by Current Procedural Terminology codes,

 International Classification of Disease, Ninth Revision, Clinical Modification

) codes, and National Drug Codes or Diagnosis-Related Group codes.

We acquired (1) enrollment files from administrative data; (2) benefits information to determine medical and pharmacy coverage; (3) inpatient, outpatient, and pharmacy claims records containing 

 diagnosis, Current Procedural Terminology codes, prescription National Drug Codes, and the starting date and length of prescriptions; and (4) laboratory records containing titles of test, test dates, and results in a subset of the population.

We restricted analyses to enrollees with bariatric surgery and diabetes. We considered patients to have had bariatric surgery if they had appropriate 

 procedure codes or Current Procedural Terminology codes.

 codes, we required that a code for morbid obesity (

 287.1) be present concurrently with the procedure code. We defined individuals as having type 2 diabetes if they had 1 relevant inpatient code or 2 outpatient codes separated by at least 30 days. The relevant codes were 250.xx, 648.0 (diabetes mellitus with pregnancy), 362.0 (diabetic retinopathy), or 266.41 (diabetic cataract). Individuals with only 250.x3 (type 1 diabetes) were not included. Additionally, any individual filling a prescription for a medication for treatment of hyperglycemia was included. If the prescription was for metformin alone, the individual was also required to have an

 code for diabetes for inclusion in this group.

We classified the dates of the laboratory tests into a 90-day period before surgery and a 90-day period after surgery (time zero). We required an individual to have pharmacy coverage throughout the entire 90-day period for his/her results to be evaluated in that period. We had only the month of the laboratory test result, so we arbitrarily assigned the middle of the month (day 15) to each record and then calculated the mean test result for each period. We limited observations to 360 days before and 900 daysafter surgery, and to enrollees with at least 1 presurgery and 1 postsurgery A1C value. We created pre-post test value pairs by matching any presurgery observation with any postsurgery observation from the same individual. We further restricted analyses to the latest presurgery observation and up to the 3 earliest postsurgery observations from the same individual.

For a 90-day period, we used pharmacy claims to determine whether an enrollee had diabetes drugs in hand on the median day in that period. Diabetes drugs were classified into the following categories: alpha-glycosidase inhibitors, dipeptidyl peptidase inhibitors, exenatide, metformin, nateglinide, sulfonylureas, thiazolidinediones, and insulin. We calculated the number of unique medications as the count of medications from these categories that were in hand on the median day in a period.

Definition of the Change From Presurgery to Postsurgery Periods

We defined change as a 3-level categorical variable: “decrease,” “same,” and “increase” (

). For example, an enrollee would be assigned to the decrease category if he/she switched from having any diabetes medication in the presurgery period to no medication in the postsurgery period, or had a drop in the count of unique medications, or had a postsurgery A1C test value lower than the presurgery value by at least 0.4%.

We further collapsed these 3 categories into a binary variable: “favorable” or “unfavorable” (Table 1). Observations in the decrease category were considered favorable while those in the increase category were unfavorable. For observations classified into the “same” category, the use of no diabetes medications in both periods was classified as favorable, whereas others were unfavorable. For “same” observations of A1C test values, postsurgery test values lower than the presurgery values,by any amount, were considered favorable; the rest were unfavorable.

Using the binary change in A1C test results as a gold standard, we calculated the positive predictive value and negative predictive value for 3 binary medication-based indicators. We also calculated the Spearman’s correlation coefficient for 2 continuous variables: change in the count of unique medications and change in the test value (a simple difference between the presurgery and postsurgery values).

Given the high prevalence of being a nonuser of oral medication/insulin or having a zero count of medications in both periods, we tested the results while restricting the analyses to enrollees who were active users in at least 1 period to identify whether these medication-based indicators performed better among active users (restricted set; 

). In addition, we also tested the outcomes with the sample limited to 1 pre-post observation pair (the latest presurgery observation and the earliest postsurgery observation; 104 pairs in total; 

). Furthermore, when calculating drug in hand, instead of using the median day only, we also investigated whether a longer period (15 days; from a week before to a week after the median day in a period) would result in differences in the performance of medication-based indicators; the results are shown in 

The data were de-identified in accordance with the Health Insurance Portability and Accountability Act’s definition of a limited data set. The Johns Hopkins University Office of Research Subjects deemed the study to be exempt from federal regulations because the research activities were considered to be of minimal risk to subjects, since the subjects were not identifiable.

 shows the flow chart for sample selection.

There were 208 paired presurgery and postsurgery observations (A1C and medication use) from 104 enrollees when including up to the 3 earliest postsurgery observations (

). The mean age of the 104 eligible enrollees was 47.9 years, and females accounted for 80% of the sample. Of 208 pairs in the presurgery period, the mean A1C test value was 7.5%. About half of the individuals used oral medications, 12% used insulin, the average number of unique medications was 1.1, and 41% were nonusers. We observed reductions in medication use and A1C in the postsurgery period: the mean A1C value decreased to 6.1%, only 19% had oral medications, only 3% had insulin, the average number of unique medications was 0.4, and close to 80% were nonusers.

Distribution of the Change in Test Indicators

For the 3 medication-based indicators, the majority of pairs were assigned to the same level categorically (54%, 90%, and 47% for the change in oral medication, insulin, and count of unique medications, respectively) (

). Among the rest, more pairs were in the decrease category than in the increase category (39%, 10%, and 48% for the change in oral medication, insulin, and count of unique medications, respectively). For the change in A1C test value, about three-fourths were in the decrease category; only 7% were in the increase category.

As for binary indicators of change, more than 80% of 208 pairs had favorable change across all 4 measures (81%, 97%, 84%, and 87% for the change in oral medication, insulin, count of unique medications, and A1C value, respectively).

Performance of the Medication-Based Indicators

All 3 medication-based indicators had high predictive positive values and low negative predictive values under most circumstances (

). The positive predictive values were similar across all 3 indicators, with more variation in the negative predictive values. The predictive positive values of all 3 indicators were about 0.87. In analyses restricted to patients with any medication use, the positive predictive values ranged from 0.80 (insulin) to 0.91 (oral medication) (eAppendix A). The negative predictive values ranged from 0.18 (oral medication) to 0.33 (insulin); however, the 95% confidence intervals associated with insulin were substantially wider than others. When we restricted the analyses to a single postsurgery observation for each individual, results were minimally changed.

When using a 15-day period to calculate drug in hand, we found that the actual counts of medication in hand from the same period increased compared with those calculated by using just the median day; for example, the proportion of subjects on oral medication in the presurgery period among these 208 pairs increased from 50.5% to 59.1% (Table 1 in eAppendix C), which could be expected because the length of time used to determine drug in hand was much longer. However, when it came to the distribution and performance of the medication-based indicators, both sets generated very similar results (Tables 2 and 3 in eAppendix C).

All 3 medication-based indicators had very high sensitivity (

0.25). There was a modest but statistically significant correlation between the change in count of unique medications and the change in A1C test value. Spearman’s correlation coefficient was 0.24 without restriction (Table 4) and 0.33 after restriction (Table 2 in eAppendix A).

METHODS

Without clinical information, a decrease in use of medications can serve as a proxy for clinical improvement.

Change in Claims-Based Diabetes Medications Is a Diabetes Improvement Indicator

 To determine the willingness of accountable care organizations (ACOs) to bear financial risk for the healthcare they provide.

 Structured interviews conducted between January and June 2012 with 57 ACOs led by hospitals and physician groups located throughout the United States. Findings are based on the 38 ACOs that were actively providing care under an ACO payment arrangement at the time of the interview.

Among these ACOs, 71% cover a portion of their ACO population with contracts that put the ACOs at some financial risk, while 45% have risk-based contracts for their entire ACO population. Payments based on fee-for-service (FFS) billing still dominate, as 92% of ACOs use FFSbased billing for at least a portion of their ACO population and 71% are fully reimbursed  using FFS-based billing.

Under the auspices of an ACO, providers are accepting some financial risk for their accountable care patient population. There is still strong reliance on FFS-based billing methods as providers experiment with different payment models.

Provides an overview of how accountable care organizations (ACOs) are experimenting with risk-bearing reimbursement agreements.n ACOs are willing to bear some risk for a portion of the care they provide.

		Movement toward more risk will be slow as ACOs develop needed skill sets to better manage risk; the movement will not mimic the rush to capitation in the 1990s.

		ACO reimbursement models largely rely on fee-for-service billing.

Accountable care organizations (ACOs) have become part of the American health system lexicon with active backing from Medicare

 They have been tabbed as a means to improve the value of healthcare by improving the experience of care, improving population health, and lowering health costs.

 Whether they will succeed at these  goals remains to be determined, but a small number of providers have begun to actively experiment in accountable care.

Since late 2010 we have been tracking the accountable care movement and have identified, to date, 428 organizations that are currently either operating as ACOs or are in the process of adopting accountable care. In tracking these entities, we use a broad definition and include organizations that self-identify as being an ACO and those that may choose a different name but seek the goals of accountable care

 and oversee the provision of health services delivered to a defined population with financial responsibility for that care. ACOs on this master list are identified via, among other means, press reports, news articles, government announcements, conferences, personal and industry interviews, and other public records.

 While the number of ACOs indicates strong interest in accountable care, it still represents a small minority of providers within the United States. To better understand what accountable care means for these organizations and the patients they serve, we are conducting interviews with these ACOs to gain insight into how they are becoming “accountable,” what risk they are bearing, how they are coordinating care, and how they are measuring and seeking to improve quality.

From January to June 2012, our team conducted structured interviews

 with 57 ACOs from across the United States that are engaged with public or private payers in accountable care programs, drawn from a sample (at that time) of 221 ACOs. This series of interviews represents the first round of an ongoing study and includes 45 hospital system—led ACOs and 12 physician group–led ACOs.

 Of those interviewed, 49 percent are participating in the Medicare Shared Savings Program (MSSP) or Medicare Pioneer ACO initiatives, with the remainder involved in ACOs with private payers and/or Medicaid programs; 68 percent of the MSSP/Pioneer ACOs also have private or Medicaid contracts. At the time of the interview, 67 percent of these ACOs were operational (meaning they were already accepting payments under their ACO contract) and 25 percent more expected to be operational within a year. The results of this paper are limited to the 38 operational ACOs which comprise 28 led by hospital groups and 10 led by physician groups.

Interviews with ACOs consisted of interviewers asking structured questions where specific pieces of information were sought, which was followed by qualitative discussions of the responses. The questions relevant to this study are as follows:

(1) How far away is your ACO from being fully operational (ie, formally providing care to ACO patients) [currently operational, 1 year, 2-4 years, 5+ years, don’t know]?

(2) What is the nature of your payment arrangement with participating [providers/payers] [fee-for-service, episodic bundled payment/DRG, capitation, shared savings (upside only), shared savings (up and downside), other]?

(3) What is the ballpark mix of the ACO’s covered lives by payer [commercially insured %, Medicare %, Medicaid %, Medicare Advantage %, other %]?

(4) How many individual patients are served by the ACO?

An important distinction for ACOs is the difference between the ACO population and the total patient population. The ACO population includes all individuals whose care is reimbursed under an ACO payment arrangement. In addition to their ACO population, all of these providers have non-ACO patients that they still serve. One weakness of this study is that we were not able to consistently collect the size of the non-ACO population. Among ACOs where we did collect data on the size of the total patient population, there is a considerable range of the ACO population as a percent of the total patient population, ranging from very small (less than 5%) to a majority of patients (greater than 50%), with smaller values being much more common. The results in this study, then, are limited to the ACO patient population.

The population of patients covered by ACOs varied considerably, with 26 percent of ACOs only working with a fee-for-service (FFS) Medicare or Medicare Advantage population, 21 percent only working with commercially insured patients, and 8 percent only working with their state Medicaid population. Eight percent of ACOs had contracts in place covering all major payer types (Medicare, Medicare Advantage, commercial, and Medicaid). Within these groupings, covering some commercial patients or Medicare Advantage patients within the ACO does not mean the ACO covers all commercial or Medicare Advantage patients, as providers must individually negotiate with each payer. 

 contains information on the percent of ACOs that cover variouspopulations under their accountable care contracts. The “Other” category includes populations such as Tricare, self-pay, and the provider’s own employees.

ACOs have multiple types of payment arrangements, with many (42 percent) being reimbursed for their ACO population in 2 or more ways. Types of payment arrangements include (1) capitation (global payments for all the healthcare needs of the covered population); (2) bundled payments (fixed payments for episodes of care); (3) shared savings (using FFS payments as a basis, but with retrospective adjustments if the total annual cost for a population is more or less than certain thresholds), which can be broken down into 2-sided risk contracts (where providers are required to repay a portion of any cost overage but will share in “savings” if the population cost is lower than expected) and upside-only contracts (where the provider will only share any savings); and (4) pay for performance (P4P) variants (where providers are paid based on FFS but can receive bonus payments or increases in their base fee schedule conditioned on reaching performance benchmarks). 

 contains information on the percent of ACOs that have entered into different categories of payment arrangements. We did not consistently track the percent of the ACO population that was covered by each type of payment arrangement.

We consider providers to be accepting “risk-based” payments if they could potentially lose or be required to repay money, which includes payment types 1, 2, and 3. The amount of risk varies based on the type of payment. We define FFSbased billing as those models that are reliant on traditional FFS coding and billing and include types 3, 4, and 5. Shared savings with 2-sided risk is both a risk-based model because the provider can be required to repay some of the population’s costs and an FFS-based billing model because it relies on traditional FFS billing practices before the annual adjustment.

Capitation represents providers bearing full financial risk for the care of a population. While 24 percent of ACOs cover some portion of their ACO population with a capitation arrangement, only 1 ACO covers its entire ACO population under a capitated global payment arrangement. Some providers view capitation as the ultimate goal of accountable care and are actively preparing for it while others have taken more of a “wait and see” approach to evaluate whether narrow network capitation is the ultimate outcome of the accountable care movement or if a different, hybrid payment model will evolve.

As a way to prepare for more risk, some providers are experimenting with episodic bundling where they receive a capitated payment for specific procedures or episodes of care. Bundled payments do represent full risk, but are limited in time and scope. By choosing service lines where they are confident they can be profitable, providers justify the added expense of implementing the necessary infrastructure to expand their risk. Other providers are experimenting with bundled payments on a small scale as a way to learn about accountable care with minimal financial risk while laying the groundwork for potential future ACO efforts. Still others have consciously avoided bundled payments because they lack the technology to accurately measure the cost of providing a bundle of services.

Shared savings, consisting of 2-sided risk and upside-only payments, is the most common payment approach, with 79 percent of ACOs participating in 1 or both forms. Under a shared savings model, providers are able to continue to operate under FFS billing but with retrospective adjustments if total cost differs from a predetermined baseline. This allows providers to maintain their existing billing practices and focus their attention on efforts to lower costs, such as through care coordination. Two-sided shared savings accounts for less provider risk than capitation because any cost overages are shared with the payer, but the upside benefit is similarly limited. While most providers with shared savings arrangements agreed to take on downside risk, 21 percent have only agreed to upside payments. These ACOs with upside-only contracts recognize that they will likely be forced to bear some financial risk in the future, probably by agreeing to downside shared savings risk. This emphasis on shared savings models, though, is viewed by many of these providers as a temporary measure as the system moves toward some other, longer-term model. The vision of what the ultimate model is varies considerably among providers.

Payment approaches based on P4P payments represent the most basic foray into accountable care and are the first step  toward other payment models.

 These payment models, which vary significantly, encourage certain results by modifying the FFS pay schedule or paying bonuses based on achieving measured outcomes or reaching performance targets, but do not dramatically disrupt the existing FFS payment system. Eight percent of ACOs use only a payment arrangement based on P4P, and some reliance on this payment model was due to the inability or unwillingness of payers to enter into other agreements.

We classified 3 payment types as risk-bearing arrangements (meaning the provider could potentially lose or be required to repay money): capitation, bundling, and 2-sided shared savings, with capitation and bundling representing full provider risk for the defined care and shared savings representing/shared risk between the provider and payer. Of the ACOs we interviewed, 71 percent have some risk-bearing arrangement while 45 percent have only risk-bearing arrangements covering their ACO population. Only 8 percent of ACOs, though, cover their entire ACO population with the full-risk capitated or bundled payments. Of ACOs that bear risk, 81% share some of that risk with payers. Most ACOs that are experimenting with capitation and bundled payments limit those arrangements to just a subset of their ACO population which, in turn, is only a subset of their total patient population. While providers are experimenting with risk-based contracts, the lessons of the managed care movement of the 1990s are still very fresh in the minds of many ACO leaders, and few organizations are willing to dive into full-risk payment arrangements for their entire patient population headfirst.

ACO Use of FFS Billing With Risk-Based Payments

We defined FFS-based payments as those that rely on classic FFS- and relative value unit (RVU)-based reimbursement which allow providers to use existing billing infrastructure while experimenting with accountable care. Both varieties of shared savings payments rely on FFS billing and payments and then reconcile the population’s cost at the end of the accounting period. Similarly, the models based on P4P utilize existing FFS billing methods and modify the reimbursement factors for individual providers or add bonus payments. Only the capitation and bundled payments represent departures from FFS billing, as they consist of a single payment for all the care provided (capitation) or all the care related to a single episode of care (bundling). For their ACO population, 92 percent of ACOs still rely on FFS-based billing for a portion of their ACO payments and 71 percent use only FFS-based billing. Many ACOs have plans to adopt some of these new billing approaches, but implementing new reimbursement models can be slow and complex.

ACO leaders, almost universally, recognize the need for financial risk to be coupled with reimbursement that does not emphasize volume. Providers, though, are hesitant to drop their FFS reimbursement models, which incent higher volumes of care,8 even while experimenting with accountable care contracts. Shared savings with downside risk, the most common payment arrangement (58% of ACOs), represents a good middle ground for many providers to experiment with bearing risk while simultaneously maintaining their FFS-based billing systems. In this way, if their ACO proves workable, they can move to new billing models, and if the ACO fails, they can easily return to the world of traditional FFS reimbursement. Adopting some risk-based contracts while maintaining FFS payment contracts helps minimize provider risk but also poses a major challenge for ACOs as they attempt to move to value-based accountable care while maintaining short-term viability in a volume-based FFS world.

The development of the accountable care movement will be determined by early results. The majority of ACOs are using shared savings models, and most are committed to evaluating their financial returns for 2 to 3 years before moving away from these payment arrangements. ACO leaders speak of a desire to improve the value they provide, but they are hesitant to adopt any wholesale movement away from FFS-based billing as they are not fully convinced that full provider risk and capitation/bundling is the inevitable conclusion of the accountable care movement. In the short term, however, providers are continuing to experiment with new payment models and new approaches to providing and coordinating care. The end result of the ACO movement, particularly relating to provider risk and reimbursement, is still undecided, but the consensus among these organizations is that the value-based focus of accountable care is here to stay. 

 From The Ohio State University (DBM), Columbus, OH; Leavitt Partners (DBM, AAC, TPM), Salt Lake City, UT.

 The authors (DBM, AAC, TPM) report no relationship or financial interest with any entity that would pose a conflict of interest with the subject matter of this article.

Concept and design (DBM, AAC, TPM); acquisition of data (DBM, AAC, TPM); analysis and interpretation of data (DBM, AAC, TPM); drafting of the manuscript (DBM); critical revision of the manuscript for important intellectual content (AAC, TPM); administrative, technical, or logistic support (AAC); and supervision (AAC).

David B. Muhlestein, JD, MHA, 200D Cunz Hall, The Ohio State University, 1841 Neil Ave,  Columbus, OH 43210. Email: muhlestein.1@osu.edu; muhlestein@leavittpartners.com.1. Berwick DM. Making good on ACOs’ promise — the final rule for the Medicare Shared Savings Program. N Engl J Med. 2011;365(19):1753-1756.

2. Muhlestein D, Croshaw A, Merrill T, Peña C. Growth and Dispersion of Accountable Care Organizations: June 2012 Update. Leavitt Partners; 2012. http://leavittpartners.com/wp-content/uploads/2012/06/Growthand-Dispersion-of-ACOs-June-2012-Update1.pdf. Accessed December 10, 2012.

3. Berwick DM, Nolan TW, Whittington J. The Triple Aim: care, health, and cost. Health Aff. 2008;27(3):759-769.

4. Interviewed ACOs are located in the following Census Regions: 19 in the Midwest, 14 in the Northeast, 9 in the South, and 15 in the West. http://www.census.gov/geo/www/us_regdiv.pdf for a list of states included in each region. All hospital-based ACOs are led by general hospitals, 2 of which are general children’s hospitals.

5. These first interviews represent a convenience sample as they were the organizations that we were first able to interview. Approximately 40 percent of the total ACOs we have identified are led by physician groups compared with the 21 percent in our sample.

6. Fisher ES, Staiger DO, Bynum JPW, Gottlieb DJ. Creating accountable care organizations: the extended hospital medical staff. Health Aff. 2007;26(1):w44-w57.

7. Hussey PS, Ridgely MS, Rosenthal MB. The PROMETHEUS Bundled Payment Experiment: slow start shows problems in implementing new payment models. Health Aff. 2011;30(11):2116-2124.

8. Medicare Payment Advisory Commission. Report to the Congress: Reforming the Delivery System; 2008.

TRENDS FROM THE FIELD

The authors present a brief summary of the types of payment arrangements that early accountable care organizations are adopting.

Risk Bearing and Use of Fee-for-Service Billing Among Accountable Care Organizations

To evaluate the impact on smoking status documentation of a payer-sponsored pay-for-performance (P4P) incentive that targeted a minority of an integrated healthcare delivery system’s patients.

Three commercial insurers simultaneously adopted P4P incentives to document  smoking status of their members with 3 chronic diseases. The healthcare system responded by adding a smoking status reminder to all patients’electronic health records (EHRs). We measured change in smoking status documentation before (2008-2009) and after (2010-2011) P4P implementation by patient P4P eligibility.

The P4P-eligible patients were compared primarily with a subset of non—P4P-eligible patients who resembled P4P-eligible patients and also with all non–P4P-eligible patients. Multivariate models adjusted for patient and provider characteristics and accounted for provider-level clustering and preimplementation trends.

Documentation increased from 48% of 207,471 patients before P4P to 71% of 227,574 patients after P4P. Improvement  from 56% to 83% occurred among P4P-eligible patients (adjusted odds ratio [AOR], 3.6; 95% confidence interval [CI], 2.9-4.5) and from 56% to 80% among the comparable subset of non—P4P-eligible patients (AOR, 3.0; 95% CI, 2.3-3.9). The difference in improvement between groups was significant (AOR, 1.3; 95% CI, 1.1-1.4; P = .009).

A P4P incentive targeting a minority of a healthcare system’s patients stimulated adoption of a systemwide EHR reminder and improved smoking status documentation overall. Combining a P4P incentive with an EHR reminder might help healthcare systems improve treatment delivery for smokers and meet “meaningful use” standards for EHRs.

Routine documentation of smoking status improves clinicians’ delivery of tobacco treatment and is mandated in “meaningful use” standards for electronic health records butcan be challenging to accomplish.

		A payer-sponsored pay-for-performance incentive prompted systemwide action by the healthcare system, resulting in improved documentation among both targeted and nontargeted patients.

		The effect was greatest among targeted patients, suggesting the financial incentive added value as well as prompting the reminder.

		Combining a performance incentive with an electronic reminder could help healthcare systems to improve treatment for tobacco use and enable population health interventions for smokers.

Smoking kills more than 440,000 people in the United States annually and remains the leading preventable cause of death.

Despite the availability of effective treatments for tobacco dependence, physicians assess smoking at only 63% of visits and  offer counseling to only 21% of smokers.

 Identification and documentation of smoking status are the first steps to addressing tobacco dependence and have been shown to increase physicians’ delivery of treatment.

 Clinical practice guidelines for treating tobacco dependence recognize the importance of documentation and call for routine smoking status screening and documentation by healthcare systems. The US government’s “meaningful use” electronic health record (EHR) incentive program requires smoking status identification in a coded field.

 Yet achieving a high rate of smoking status documentation is a challenge for many healthcare systems.5

Efforts to improve smoking status documentation have included electronic or paper-based reminders, performance feedback, and a simple vital sign stamp.

 Pay-for-performance (P4P) incentives have been used in some systems to promote delivery of guideline-based tobacco treatment, or as a component of broader quality improvement efforts.

 These studies have examined performance incentives for various guideline-based treatment activities, including smoking status documentation, documentation of physician-delivered counseling, referral to telephone counseling, and payments to providers for patients’ tobacco abstinence. Prior studies of P4P for smoking status documentation were limited to US physician groups or healthcare delivery  systems outside the United States.

 Relatively little is known about the effects of P4P programs in large, multipayer, integrated US healthcare delivery systems in which P4P incentives do not apply to all patients in the system. A payer-sponsored P4P incentive in a multipayer system may only be effective for the patient population to whom it applies. However, practice changes stimulated by a targeted P4P measure can have broader benefits that improve documentation and treatment delivery for targeted and non-targeted patients.

A P4P incentive that was introduced in 2010 in a large multipayer healthcare delivery system rewarded practices for documenting the smoking status of patients with specific commercial insurers and 3 chronic diseases. The organization’s multiple payers and diverse provider groups with varied practice styles and patient populations make standardized quality improvement efforts especially challenging. Our objective was to study the effect of a targeted, payer-sponsored P4P incentive payment on smoking status documentation across the healthcare system.

A payer-sponsored pay-for-performance incentive for smoking status documentation prompted implementation of a systemwide electronic reminder and improved documentation among all healthcare system patients.

July 2013

Encouraging Value-Based Insurance Designs in State Health Insurance Exchanges

Change in Claims-Based Diabetes Medications Is a Diabetes Improvement Indicator

Risk Bearing and Use of Fee-for-Service Billing Among Accountable Care Organizations

Healthcare System Effects of Pay-for-Performance for Smoking Status Documentation

Healthcare's 'Big Data' Challenge

Previously Unrecognized Trends in Diabetes Consumption Clusters in Medicare

Physician Assistants and Their Intent to Retire

The Patient-Centered Medical Home in the Veterans Health Administration

Budget Impact Analysis of 8 Hormonal Contraceptive Options

Early Results From the Hospital Electronic Health Record Incentive Programs