Nicole Brainard, PhD, MPH

To understand the expansion of health plan—provider accountable care organization (ACO) partnerships and the evolution of the characteristics of these partnerships. 

We used a 2-pronged data collection approach. We collected qualitative data from the 8 plans who participated in our 2011 study on health plan—provider partnerships using the same set of interview questions. In addition, we fielded a survey on alternative delivery and payment models to 105 member plans from America’s Health Insurance Plans (AHIP), including specific questions on health plan ACO implementation.

 Our survey showed health plan expansion of ACO implementation with increased use of shared risk and widespread provision of analytic reports and other types of technical assistance to providers. The cohort of 8 health plans reported improvements in quality of care since 2011 while evolving their program components, such as use of more readily measureable provider eligibility criteria, ongoing emphasis on health plan—provided technical assistance, and greater collaboration between health plans and their provider partners, to continually re-evaluate quality and utilization measures.

 Our analysis shows maturation of private sector ACOs since 2011, with shared risk playing an increasing role in changes to reimbursement. We also saw continued emphasis on provider readiness to assume downside risk, expansion to smaller provider groups, and an increase in the scope and depth of technical assistance made available to providers in these contracts. Some challenges continue to exist, including alignment of quality measurement and reporting, broader inclusion of smaller practices in alternative delivery and payment models, and developing tools and materials to assist consumers in realizing the benefits of value-based care.

To help accelerate delivery and payment reform across the public sector, HHS Secretary Burwell announced the federal government’s goal to tie 30% of traditional, fee-for-service (FFS) Medicare payments to quality and value through alternative payment models (APMs) by the end of 2016, and 50% of payments by the end of 2018.

 Implementation of APMs in the private sector has been underway for the past 5 to 6 years. Examining the experience of health plans with such models can be useful to policy makers as they seek to transition the healthcare system to value-based payment through implementation of the Health Care Payment Learning and Action Network

 and Medicare Access & CHIP Reauthorization Act of 2015 (MACRA).

In 2011, we published a study on key characteristics and initial lessons from 8 health plans that were early adopters of accountable care organization (ACO) models.

 Re-examining their experience with these models over the past 4 years can yield useful insights for policy makers and other healthcare stakeholders; specifically, we sought to understand how ACOs implemented by these 8 plans have evolved. We were also interested in examining the widespread adoption of ACOs in a more systematic manner. To achieve these objectives, we used a 2-pronged data collection approach. We collected qualitative data from the 8 plans who participated in the 2011 study using the same set of interview questions. In addition, we fielded a survey on alternative delivery and payment models to 105 member plans from America’s Health Insurance Plans (AHIP), including specific questions on health plan ACO implementation. In this article, we describe key findings from the second round of data collection from the original 8 plans and a total of 28 plans who responded to the survey.

Findings from our member plan payment survey showed that all 28 health plans that responded to the survey had ACO contracts in effect. We also found that these ACO contracts were present in 43 states and spanned commercial, Medicaid, and Medicare Advantage product lines. In addition, several of our 8 case study plans, which cover approximately 80 million commercially enrolled lives, reported extending their partnerships beyond the largest health systems and provider groups to smaller physician groups. Some case study health plans have lowered the requirement for minimum patient panel size from 10,000 to 5000 patients. 

The cohort of 8 case study plans with several years of experience implementing ACOs reported improvements in quality and costs. Specifically, some of the observed improvements, shown as ranges experienced by the 8 case study plans, include:

		Improved performance on clinical quality measures for specific conditions (up to 20% improvements), such as diabetes (19%-25% increase) and medication adherence (27% increase).

		Reduced readmission rates (11%-32% improvements).

		Increased prescribing of lower-cost medications (21%-52%) through approaches such as generic prescribing.

		Achieved a significantly higher referral rate to high performing specialists than the market referral rate (81% higher).

		Reduced emergency department visits (19%-50% improvements).

Health plans reported actively seeking to incorporate patient satisfaction measures into their ACO contracts by conducting patient satisfaction surveys. In conducting these surveys, one plan reported that patients attributed to the ACO experienced better access to urgent care, higher satisfaction with the amount of time they spend with providers, and higher satisfaction with how well their providers listened to them compared with members not attributed to the ACO. Finally, some plans reported observing “spillover effects,” where patients not attributed to the ACO experienced improvements in outcomes, and greater access to care through extended provider office hours due to re-engineering of care delivery.

These improvements in quality of care were also accompanied by other key changes in the design and implementation of ACOs, which are highlighted below. 

As health plans have expanded their ACO contracts, provider eligibility criteria for such contracts have also evolved. Although the 8 case study plans reported using some of the same criteria as in the 2011 study, such as strong leadership and commitment to transformation, several new and more readily measurable criteria have emerged. 

Survey results indicate that health plans use a mix of reimbursement methods in ACO contracts, and our case study plans provide some insight into the rationale for the variation in payment methods. Consistent with previous findings, plans still emphasized the need for flexibility to tailor reimbursement to providers’ ability to assume risk. Health plans reported that it is important for provider organizations to have time to gain both confidence and experience with new approaches before taking on financial risk. As a result, health plans allow for a transition from traditional FFS to APMs by first introducing shared savings only, and then increasing the share of such potential savings as providers gradually accept downside risk. Over time, financial incentives associated with accepting risk significantly outweigh gains from shared savings only and traditional FFS payments.  

Although flexibility is emphasized, we observed increased movement toward shared savings and risk contracts, with 43% of plan respondents including shared savings, and 50% including shared risk, as part of the reimbursement structure.

Data collected through the survey and 8 case study health plans continue to show plan emphasis on technical assistance to providers. Technical assistance usually consists of sharing timely data and analytic reports in addition to infrastructure development and support.   Timeliness of shared information is deemed essential for providers in ACO contracts with financial risk, so that they may quickly identify areas for quality and cost improvement, thus mitigating their risk. 

Survey results indicate that a vast majority of health plans provide technical assistance to provider partners in areas of disease and care management (96%), exchanging health information (96%), population health management (89%), infrastructure needs/data platforms (82%), and managing financial risk (64%). We found that utilization reports, inpatient census reports, and clinical quality rates were the most common types of data shared with ACO providers; data on care gaps, financial targets, and patient experience also were being shared by a majority of the plans surveyed.

Data collected from the 8 case study plans provide a more nuanced view of the scope and breadth of technical assistance made available to their provider partners. 

Within the past several years, greater integration of clinical and electronic health record (EHR) data, as well as medical and pharmacy claims, has allowed plans to identify at-risk members earlier and generate a more comprehensive view of a patient’s health. Beyond acquiring the data, there appears to be interest among some ACOs to use these data and analytics to identify appropriate referral partners. For example, one case study plan reported heightened ACO interest in using the health plan’s data to gauge how a referral partner might perform in a shared savings relationship. As provider organizations have matured, plans place more emphasis on leveraging the combined strengths of both plan and provider groups, with experienced provider groups taking on leadership roles within programs traditionally implemented by plans.

Most health plans continue to use a “quality gate” to ensure that providers achieve performance targets before they earn financial rewards. Additionally, data collected from the case study plans show incorporation of patient experience measures in their ACO contracts either currently or in the near future. We also found that 32% of plans responding to our survey require ACO physicians to meet patient experience targets to be eligible for incentive payments.

As plans have had more years to refine performance measurement and provider partners gain experience with measurement, a collaborative process is established to continually re-evaluate quality and utilization measures to ensure alignment around incentives for quality improvement. One health plan noted that they are increasing the amount of risk tied to performance measures such as potentially preventable events. 

Our analysis shows maturation of private sector ACOs since 2011 with shared risk playing an increasing role in changes to reimbursement. We also saw continued emphasis on provider readiness to assume downside risk, expansion to smaller provider groups, and an increase in the scope and depth of technical assistance made available to providers in these contracts.

Although our data show significant progress and expansion of ACOs, some challenges continue to exist, including alignment of quality measurement and reporting, broader inclusion of smaller practices in alternative delivery and payment models, and development of tools and materials to assist consumers in realizing the benefits of value-based care. Work is currently underway through the Core Quality Measures Collaborative to address quality measure alignment across public and private sectors.

 Going forward, increased emphasis on patient engagement and communication strategies will be central to the success of these alternative delivery and payment models. Finally, additional scientific research is needed to identify which combinations of interventions have led to the reported improvements in quality and cost measures among private sector ACOs, as well as to identify the types of data most useful to providers in achieving these improvements.

 America’s Health Insurance Plans (AH, KS, GP, NB, KD), Washington, DC.

 Ms Dawson is an employee of Health Care Service Corporation as of February 2016; however, the study was conducted while she was employed by America’s Health Insurance Plans. The remaining authors 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 (AH, KD, GP); acquisition of data (KD); analysis and interpretation of data (AH, KS, GP, NB, KD); drafting of the manuscript (KS, GP, AH, NB); critical revision of the manuscript for important intellectual content (AH, NB, KS); statistical analysis (NB); administrative, technical, or logistic support (KP, NB); supervision (AH).

Aparna Higgins, MA, America’s Health Insurance Plans, 601 Pennsylvania Ave NW, Washington DC 20004. E-mail: ahiggins@ahip.org

1. Better, smarter, healthier: in historic announcement, HHS sets clear goals and timeline for shifting Medicare reimbursements from volume to value. http://www.hhs.gov/news/press/2015pres/01/20150126a.html. Washington, DC: HHS; January 26, 2015. Accessed November 27, 2015.

2. Fact sheet: Health Care Payment Learning and Action Network. CMS website. https://www.cms.gov/Newsroom/MediaReleaseDatabase/Fact-sheets/2015-Fact-sheets-items/2015-03-25.html. Washington, DC: CMS; March 25, 2015. Accessed December 1, 2015.

3. H.R. 2: Medicare Access and CHIP Reauthorization Act of 2015. GovTrack.us website. https://www.govtrack.us/congress/bills/114/hr2. Published 2015. Accessed February 23, 2016.

4. Higgins A, Stewart K, Dawson K, Bocchino C. Early lessons from accountable care models in the private sector: partnerships between health plans and providers. 

. 2011;30(9):1718-1727. doi:10.1377/hlthaff.2011.0561.

5. Conway PH; Core Quality Measures Collaborative Workgroup. The Core Quality Measures Collaborative: a rationale and framework for public-private quality measure alignment. 

 blog. http://healthaffairs.org/blog/2015/06/23/the-core-quality-measures-collaborative-a-rationale-and-framework-for-public-private-quality-measure-alignment/. Published June 23, 2015. Accessed December 14, 2015.

Articles

Although health plan accountable care models have evolved provider readiness, data, analytics, and the use of performance measurement are important components of plan-provider partnerships.

Health Plan—Provider Accountable Care Partnerships: How Have They Evolved?

Policy makers have growing interest in price transparency and in the kinds of tools available to consumers. Health plans have implemented price estimator tools that make provider pricing information available to members; however, systematic data on prevalence and characteristics of such tools are limited. The purpose of this study was to describe the characteristics of price estimator tools offered by health plans to their members and to identify potential trends, challenges, and opportunities for advancing the utility of these tools.

Between 2014 and 2015, we conducted a national Web-based survey of health plans with commercial enrollment (100 plans, 43% response rate). Descriptive analyses were conducted using survey data. 

Health plan members have access to a variety of price estimator tool capabilities for commonly used procedures. These tools take into account member characteristics, including member zip code and benefit design. Despite outreach to members, however, challenges remain with respect to member uptake of such tools. 

Our study found that health plans share price and provider performance data with their members. 

	We conducted a national survey of health plans to examine characteristics of provider price estimator tools available to the commercially insured population. Health plans offer these tools to their enrollees and take into account member characteristics such as zip code and benefit design. 

		Price estimates are available for commonly used services, including outpatient surgery and radiology. 

		Plans use various approaches, including digital and social marketing, to increase member use of tools. 

		Early evidence shows the positive impact of such tools on cost savings. Additional research is needed to increase the use of these tools and broaden their impact.

In an effort to slow the rising costs of healthcare, consumer groups, state governments, and health plans have called for greater price transparency in the healthcare market.

 Policy makers and researchers have posited that increasing price transparency will afford consumers the opportunity to comparison-shop among providers, which could result in higher-cost providers lowering their prices to improve value.

 As enrollment in high-deductible health plans has grown, consumers in these plans have a greater financial incentive to compare prices across providers because they have higher out-of-pocket obligations.

 Even consumers not in high-deductible health plans may now participate in benefit designs offering incentives and opportunities to select higher quality and lower cost (ie, high-value) providers.

 Consumers want to understand their out-of-pocket expenses for specific services for a given provider, as well as the costs for a complete episode of care.

 To help meet consumer needs, health plans have implemented tools that allow consumers to compare and select providers based on providers’ quality performance and prices for specific healthcare services. Using such tools can benefit consumers in terms of lowering their out-of-pocket costs and selecting higher-quality providers.

Recent evidence shows that availability of price information can impact consumer behavior and promote informed decision making. In a study of 1421 employees, researchers found that presenting easy-to-understand, high-quality information alongside pricing estimates increased the likelihood of an individual choosing a high-value provider.

 Increased likelihood of visiting a new provider, as well as lower payments for clinical care, have also been attributed to accessibility of price estimates.

 Much attention has been given to the topic of price transparency; however, studies to date have primarily focused on price information made available through state requirements for mandatory reporting.

 Meanwhile, in the private sector, health plans have begun to offer provider-level pricing information to their members using Web-based price estimator tools.

 Although some information is available, we are unaware of any studies that have systematically collected data on the characteristics of health plan price estimators.

 The purpose of this exploratory study was to describe the characteristics of price estimator tools offered by health plans to their members and to identify potential trends, challenges, and opportunities for advancing the utility of these tools.

For the purposes of this study, “price estimator tools” refer to tools that allow consumers to obtain estimates of prices associated with specific healthcare services that could be either specific to a provider or based on geography.

Data for this study were collected using a Web-based survey instrument. To develop the survey instrument, we conducted a targeted literature review of health plan price estimator tools, reviewed health plan price estimator tools that were publicly available, and conducted preliminary interviews with representatives of 4 plans to get a broad understanding of these tools. We pilot-tested the survey with these same 4 health plans and conducted telephone interviews with these plans to get additional feedback on the survey.

Final survey questions focused on the following 7 aspects of health plan price estimator tools: 1) core capabilities and characteristics taken into account, 2) data used in price estimator tools, 3) display of price estimates, 4) services for which estimates may be obtained, 5) provider information, 6) member engagement, and 7) outcomes and challenges. The questions were a mixture of open-ended and binary yes or no responses. For open-ended questions, we asked plan representatives to elaborate on characteristics pertaining to their price estimator tool that were not included as a response option on binary yes or no questions. Twenty-two binary questions had corresponding open-ended questions to which the plan representatives could respond.

We invited 106 health plans that met the following criteria: a) listed as a member of America's Health Insurance Plans in 2014, and b) offered products in the commercial insurance market. Of the 106 plans, 4 plans were not eligible to participate because they no longer offered commercial products at the time of the survey, and 2 had recently been acquired by another health plan. Using a key informant approach, we e-mailed invitations to the chief medical officers of the remaining 100 health plans, who then shared the survey with their teams, as appropriate. Data from this survey were collected in 2 phases between April 2014 and July 2015.

To be included in our analyses, plans had to have a price estimator tool available to their members. For the analyses, we characterized health plans by total size of enrollment (using quintiles), time since their price estimator tool became available (<1 year, 1-3 years, >3 years), and vendor type for their tool (health plan, third-party, or a combination of both). We examined the overall frequencies and Fisher’s exact tests for each price estimator tool characteristic, stratified by health plan characteristics (not all results are reported). Following the calculation of frequencies, we compiled and summarized all open-ended responses. Any plans that did not provide a price estimator tool were excluded from analyses.

A total of 43 plans responded to the survey, resulting in a 43% response rate. The characteristics of the responding plans can be found in the 

). Of the 43 plans that responded during both phases of data collection, 11 did not provide a price estimator tool to their members, although 4 of these 11 intended to provide a tool within 12 to 24 months. One plan of the 43 had made a tool available during our second phase of data collection in 2015, but did not provide updated responses to the survey. Our final analytic sample is thus composed of the remaining 31 plans that provided price estimator tools to their members.

The 31 plans included in the analyses account for 140.8 million commercial enrollees, corresponding to 75.9% of the total national commercial enrollment for 2014.

 Thirty-nine percent of these plans launched their price estimator tools more than 3 years ago, and just under half of the 31 plans (45%) offered multiple tools to their members—both provided by the plan directly and through arrangements with third-party vendors.

Capabilities of health plan price estimator tools. 

Our results indicated that health plan members have access to a variety of price estimator tool capabilities (

). Ninety-four percent of plans allowed for provider comparison shopping and about 58% displayed estimates for prescription drug costs. In generating price estimates, plans account for various member characteristics depending on the type of service being estimated. For instance, a member’s zip code (94%), product type (77%), and benefit design (77%) are commonly used characteristics for provider comparison shopping. In contrast, about one-third of plans account for member zip code (35%), product type (29%), and benefit design (29%) for cost estimates of prescription drugs (Table 1).

Services for which estimates may be obtained. 

Eighty percent of health plans choose which services to provide estimates for based on the most commonly used procedures and services. Of the most commonly used procedures, we found that nearly all of the plans provided estimates for elective outpatient surgery (97%), radiology services (eg, x-rays, computed tomography scans) (97%), and inpatient surgical services (97%) (

A majority of health plans also provided estimates for physician services (71%) and services associated with select chronic conditions (61%), such as glycated hemoglobin testing for patients with diabetes. Open-ended responses revealed that many members might also obtain estimates for preventive services, including wellness visits, preventive screenings, and behavioral health services. Less common, are estimates for services at retail/convenience clinics (32%), services provided in the emergency department (26%), and telemedicine (6%).

Data used to generate price estimates for the services primarily consisted of a health plan’s historical paid claims for a specific geographic area (55%), the health plan’s historical allowed rates (the maximum a health plan will pay an in-network provider for a service) for specific providers (52%), and a plan’s current negotiated rates (ie, the price negotiated between the health plan and the provider for a service) for specific providers (45%). These data are generally updated quarterly or semiannually (65%).

Health plans price estimator tools display a number of specific elements primarily for in-network providers, and members have the ability to compare prices for specific healthcare services across individual providers (68%). Seventy-one percent of plans also included price estimates for treatments at the individual provider level. Potential out-of-pocket costs that a member may incur are displayed by 90% of plans; such potential out-of-pocket costs could include co-pays, coinsurance, and the deductible for which the member is responsible. Fewer plans display information on reference-based pricing for services (13%), health savings account (HSA) balances (29%), and links to an outside HSA administrator (10%).

Alongside price estimates, plans share other information on providers, such as geographic location, performance data, designation as a Center of Excellence, etc (

). As shown in Table 2, the extent to which these additional pieces of provider information are fully integrated into the price estimator tool varies across plans.

The most common ways to direct members to use a price estimator tool are through messaging on a health plan portal (97%), outreach through employers (71%), digital communication—which includes e-mail, social media, and text messaging&mdash;(68%), and postal mail (58%) (

). Open-ended responses yielded additional details on innovative methods that plans are using to engage members; such methods include technological platforms like e-mail and text messaging, in which plans use alerts to prompt members’ awareness of “ways to save” on healthcare. Also technologically based are gamified programs like “Healthcare University,” in which members receive information through interactive videos and quizzes. Person-to-person communication, including the work of schedulers and financial counselors, is another way in which health plans engage members. On a larger scale, plans cite the implementation of social marketing campaigns to connect with members on price transparency.

Plans were also asked to identify features of their price estimator tool that members have found most useful. Estimates for out-of-pocket costs based on member benefits are recognized as one of the most useful features by the majority of plans (77%). Equally useful were the ability to compare prices for services across providers (77%) and having access to estimates for a wide range of services (68%) (Table 3).

Nearly half of the responding plans (45%) (

) had yet to formally evaluate outcomes from member use of price estimator tools, but they intended to in the near future. Those that had evaluated their price estimator tool reported increased use of lower-cost providers (19%), member out-of-pocket savings (16%), and employer cost savings (16%) (Table 4). The most common challenges reported were limited member uptake (58%) and lack of member awareness about the tool (58%) (Table 4).

Potential Relationship Between Key Features and Time Since the Tool Became Available

As part of our analysis, we wanted to understand if there were differences in tool characteristics given a health plan’s enrollment size, how long their tool has been available, and the type of tool vendor. Our results indicated a possible relationship between the length of time a tool has been available and the presence of specific tool features and challenges. An examination of this relationship can help to identify areas for additional research with a focus on understanding the complex implementation process of price estimator tools. To further examine this potential relationship, we calculated Fisher’s exact tests for tool features and challenges by time since the tool became available. Due to the exploratory nature of our study, we reported tool features and challenges, yielding a Fisher’s exact test significant at 

We observed the strongest relationships between time since the tool became available and estimation of physician services (

 <.03), displaying patient-related costs (

 <.05), and availability of delivery and postpartum service estimates (

 <.05), which indicated a significant difference in the likelihood of these elements being present in a tool given how long the tool has been available. Closely following in significance were limited member uptake (

 <.09), assistance provided by a nurse care manager (

 <.10), and displaying episode-of-care costs (

 <.14). Of the remaining features, we saw significance among availability of provider performance data (

 <.17), displaying estimates for prescription drugs (

 <.21), phone outreach by nurse care managers (

 <.18), availability of provider-level metrics (

As price estimator tools continue to evolve to meet the needs of consumers, the data that are displayed may affect the overall utility of these tools. Our results indicated that health plans are responding to consumer needs by providing price estimator tools that include information on out-of-pocket costs, with many of these price estimates taking into account individual benefit design.

 When price information is displayed without accompanying quality information, consumers potentially perceive higher price as being equivalent to higher quality. If consumers associate high price with high value, this does not lead to selection of high-value providers and may even deter consumers from seeking care altogether.

 Our study found that two-thirds of responding plans shared provider performance data with their members, with half of these plans integrating such data into their price estimator tool.

Despite providing price estimator tools that also include access to provider performance data, health plans in our study encountered challenges related to lack of member uptake. This finding coincides with the Catalyst for Payment Reform’s 2013 National Scorecard, which reported that 86% of plans provide cost information to members relative to co-pays and cost sharing, but only 2% of their members use these resources.

 Our results support the need for additional research on the appropriateness of outreach methods and how to better understand reasons for lack of engagement, including low health literacy and language barriers.

 A better understanding of these reasons can facilitate consumer uptake and help the evolution of these tools.

We also know from previous research that simply having access to a tool with such information does not increase the likelihood of using the tool, nor does it necessarily lead to lower costs.

 The way in which cost and quality data are visually displayed merits further exploration, as it impacts how well consumers are able to comprehend and use the information.

 Research suggests that quality data be presented with a clear format and descriptive labeling, and in minimal volume to not overwhelm consumers reviewing information.

First, with a response rate of 43%, these findings may not be fully generalizable to all health plans. However, we are comfortable with our findings since the plans included in our analysis account for 75.9% of total national commercial enrollment and, therefore, capture tool availability to a large proportion of commercial enrollees.

 Secondly, this study was not designed to draw conclusions on how tool characteristics impact cost outcomes or consumer behavior, but rather to describe the current characteristics and capabilities of health plan price estimator tools. These limitations notwithstanding, we believe our study contributes several interesting findings on the characteristics and capabilities of health plan price estimator tools.

Our study found that health plans provide price and provider performance data to their members. Additional research is needed to inform further development of these tools so they can better engage consumers, who can then make more informed and cost-effective decisions regarding their healthcare. 

 America’s Health Insurance Plans (AH, NB, GV), Washington, DC.

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

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

A national survey of health plans shows that price estimator tools available to commercial plan enrollees provide price information for a variety of services at the provider level.

Nicole Brainard, PhD, MPH

Articles

Health Plan—Provider Accountable Care Partnerships: How Have They Evolved?

Characterizing Health Plan Price Estimator Tools: Findings From a National Survey