CMS-10575 Final SS - Part B - 1 21 16 clean

Health Care Payment Learning & Action Network

Supporting Statement - Part B

Collections of Information Employing Statistical Methods

CMS requests a generic clearance to assist in monitoring and characterizing the adoption of alternative payment models in order to meaningfully inform 1115A model decisions on the design, selection, testing, modification, and expansion of innovative payment and service delivery models in accordance with the requirements of section 1115A, which would in turn contribute to progress toward the Secretary’s goal to increase the percentage of payments tied to alternative payment models. Each of the four types of information requests, and the statistical methods behind them, are described below.

a) LAN Registration Information

1. Describe (including a numerical estimate) the potential respondent universe and any sam­pling or other respondent selection method to be used. Data on the number of entities (e.g., establishments, State and local government units, households, or persons) in the universe covered by the collection and in the corre­sponding sample are to be provided in tabular form for the universe as a whole and for each of the strata in the proposed sample. Indicate expected response rates for the collection as a whole. If the collection had been conducted previously, include the actual response rate achieved during the last collection.

Response: The first type of information request involves collecting self-reported identification and opinion information from participants when they sign up for the LAN and associated events (LAN Summits, webinars, and other meetings). This information is being collected to understand the types of stakeholders participating in LAN activities (providers, payers, consumers, states, etc.). Although the potential universe of respondents includes all interested U.S. healthcare stakeholders (from physicians to payers to health care consultants), this collection effort does not involve any sampling or other respondent selection methods, because CMS will only collect information from organizations and individuals who voluntarily register for the LAN and associated LAN events. Therefore, the response rate will necessarily be 100%. However, this respondent information for the LAN and each LAN event will be used to estimate attendance at future LAN events and to segment respondents by stakeholder type, which will inform our ability to segment and conduct information requests for the other three LAN information requests described later in this document.

To date, approximately 4,670 public and private health care organizations (payers, providers, individuals, employers, consumers, local, state and federal agencies) interested individuals have signed up to participate in the LAN since its launch on March 26, 2015. We anticipate an additional 3,000 participants will sign up for the LAN during 2016. From this pool of the LAN participants, subsets of LAN participants are expected to sign-up for LAN in-person events, webinars, and other meetings. For example, at the LAN Summit held in October 2015 approximately 800 LAN participants registered for the event. Based on these past registrations for LAN events in 2015 and the number of LAN events planned for 2016, CMS is aggressively estimating that LAN participants will register for LAN events on no more than 30,000 occasions during 2016.

2. Describe the procedures for the collection of information including:

- Statistical methodology for stratification and sample selection,

- Estimation procedure,

- Degree of accuracy needed for the pur­pose described in the justification,

- Unusual problems requiring specialized sampling procedures, and

- Any use of periodic (less frequent than annual) data collection cycles to reduce burden.

Response: This information collection type will not be employing any statistical/sampling methodologies or estimation procedures for this information request. However, we do expect the burden for this information request to decrease over time, because organizations and individuals who have already registered for the LAN will be able to automatically populate LAN event registration forms.

3. Describe methods to maximize response rates and to deal with issues of non-response. The accuracy and reliability of information collected must be shown to be adequate for intended uses. For collections based on sam­pling, a special justification must be provid­ed for any collection that will not yield 'reliable' data that can be generalized to the uni­verse studied.

Response: LAN participant organizations who do complete registration forms correctly will not be counted as registrants. Maximizing response rates and sample methodologies are not applicable to this information request.

4. Describe any tests of procedures or methods to be undertaken. Testing is encouraged as an effective means of refining collections of information to minimize burden and improve utility. Tests must be approved if they call for answers to identical questions from 10 or more respondents. A proposed test or set of tests may be submitted for approval separate­ly or in combination with the main collection of information.

Response: For LAN registration information collections, CMS and its contractor (CAMH) will build and internally test the registration data fields and associated collection instruments for functionality and ease of use, to include the estimated amount of time required to complete each registration form. For these simple information requests, we do not plan to work with more than 10 respondents to estimate the burden. Instead, this kind of testing will be conducted internally.

5. Provide the name and telephone number of individuals consulted on statistical aspects of the design and the name of the agency unit, contractor(s), grantee(s), or other person(s) who will actually collect and/or analyze the information for the agency.

Response:

Name	Telephone	Organization
Darren DeWalt	410-786-9731	CMS/CMMI/LDG
Alison Shippy	410-786-0438	CMS/CMMI/LDG
Dustin Allison	303-844-7031	CMS/CMMI/LDG
Kris Prendergast	443-200-4039	Mitre-CAMH
Julia Rollison	443-200-4039	Ripple Effect Communications

b) LAN Participant Surveys

1. Describe (including a numerical estimate) the potential respondent universe and any sam­pling or other respondent selection method to be used. Data on the number of entities (e.g., establishments, State and local government units, households, or persons) in the universe covered by the collection and in the corre­sponding sample are to be provided in tabular form for the universe as a whole and for each of the strata in the proposed sample. Indicate expected response rates for the collection as a whole. If the collection had been conducted previously, include the actual response rate achieved during the last collection.

Response: The second type of information request involves administering surveys to LAN participants to understand LAN participant opinions, priorities, and issues with respect to how to best increase the adoption of alternative payment models. Surveys will inform decision making about future LAN activities, including LAN Conference sessions, webinar topics, and feedback on LAN work group ideas. The potential universe of respondents includes all LAN participants, though each survey will likely be limited to a particular LAN event. For example, in connection with a LAN webinar focused on sharing a white paper, a simple evaluation form would be sent to all attendees to better understand what went well, what could be improved, and so on. In our experience at CMS, these kinds of simple surveys can, at most, expect a 33% response rate. For this reason, we estimate that 10,000 LAN participants will respond to these type of surveys during 2016, out of the total estimate of 30,000 LAN event registrants.

2. Describe the procedures for the collection of information including:

- Statistical methodology for stratification and sample selection,

- Estimation procedure,

- Degree of accuracy needed for the pur­pose described in the justification,

- Unusual problems requiring specialized sampling procedures, and

- Any use of periodic (less frequent than annual) data collection cycles to reduce burden.

Response: This information collection type will not be employing any statistical/sampling methodologies or estimation procedures for this information request. CMS will continually strive to increase the proportion of the LAN participants who respond to these surveys, but our intended use for informal decision making will not require generalizable data.

3. Describe methods to maximize response rates and to deal with issues of non-response. The accuracy and reliability of information collected must be shown to be adequate for intended uses. For collections based on sam­pling, a special justification must be provid­ed for any collection that will not yield 'reliable' data that can be generalized to the uni­verse studied.

Response: Maximizing response rates and sampling methodologies are not applicable to this information request. CMS will continually strive to increase the proportion of the LAN participants who respond to these surveys, but our intended use for informal decision making will not require generalizable data.

4. Describe any tests of procedures or methods to be undertaken. Testing is encouraged as an effective means of refining collections of information to minimize burden and improve utility. Tests must be approved if they call for answers to identical questions from 10 or more respondents. A proposed test or set of tests may be submitted for approval separate­ly or in combination with the main collection of information.

Response: For LAN surveys, CMS and its contractor (CAMH) will build and internally test the survey data fields and associated collection instruments for functionality and ease of use, to include the estimated amount of time required to complete each registration form. For these simple information requests, we do not plan to work with more than 10 respondents to estimate the burden. Instead, this kind of testing will be conducted internally.

5. Provide the name and telephone number of individuals consulted on statistical aspects of the design and the name of the agency unit, contractor(s), grantee(s), or other person(s) who will actually collect and/or analyze the information for the agency.

Response:

Name	Telephone	Organization
Darren DeWalt	410-786-9731	CMS/CMMI/LDG
Alison Shippy	410-786-0438	CMS/CMMI/LDG
Dustin Allison	303-844-7031	CMS/CMMI/LDG
Kris Prendergast	443-200-4039	Mitre-CAMH
Julia Rollison	443-200-4039	Ripple Effect Communications

c) LAN Payer APM Tracking

1. Describe (including a numerical estimate) the potential respondent universe and any sam­pling or other respondent selection method to be used. Data on the number of entities (e.g., establishments, State and local government units, households, or persons) in the universe covered by the collection and in the corre­sponding sample are to be provided in tabular form for the universe as a whole and for each of the strata in the proposed sample. Indicate expected response rates for the collection as a whole. If the collection had been conducted previously, include the actual response rate achieved during the last collection.

Response: The third type of information request involves collecting information from LAN payers to track the adoption of alternative payment models. Requested information may also include estimated counts of U.S. beneficiaries/consumers who are covered by alternative payment models and for each respondent, the percentage of payments made through alternative models. This information will help the LAN understand general market trends and the pace of progress toward alternative payment model adoption across public and private payers.

The LAN adapted the CMS payment taxonomy¹ and expanded it by introducing refinements that describe health care payment through the stages of transition from pure fee-for-service to APMs and, ultimately, population based payments. The resulting APM Framework classifies payment models into four categories according to how providers are paid:

• category 1—fee-for-service with no link of payment to quality;

• category 2—fee-for-service with a link of payment to quality;

• category 3—alternative payment models built on fee-for-service architecture;

• category 4—population-based payment.

Because the goal of the LAN is to match or exceed the Secretary’s Medicare FFS goal for the entire U.S. health care system, the LAN plans to use a calculation similar to the following:

Denominator: The total estimated annual payment amounts that U.S. health plans/payers participating in the LAN made to providers in a given year. To the extent to which these participants encompass of all plans/payers, it can be used to estimate the total annual U.S. health care spending to ensure representativeness)

Numerator: The total estimated annual payment amounts of category 3 and 4 payments that health plan/payers participating in the LAN made to providers, such as the amount of payments made in 2015 to providers participating in a patient centered medical home with quality gates and upside risk. (See the LAN’s recently released APM Framework White Paper to get a better sense of what plans should count as a category 3 or 4 payment model)

To obtain this kind of information, U.S. health plans/payers participating in the LAN must be willing to categorize their payments to providers in accordance with the APM Framework and then be willing to share that categorized and aggregated data with the LAN and the public. The LAN does not have any regulatory levers at its disposal to compel participating plans to submit these data – it will be purely voluntary. That said, the majority of national payers are participating in the LAN and have played significant roles in developing a revised APM Framework built off the CMS payment taxonomy (see APM Framework White Paper).

Given health plan involvement in the LAN events to date (estimated at nearly 100 organizations), CMS is confident a sufficient number of health plans will participate in this data collection effort so that the resulting data would give a strong indication of the national direction. Although CMS does not expect to engage in sampling a certain proportion of health plans for these data, the LAN will continually engage health plans throughout the life of the LAN to increase the number of participating health plans in the LAN. The ultimate goal is for all U.S. health plans (the potential respondent universe), both public and private, to participate in the LAN and commit to sharing aggregate payment data on an annual to bi-annual basis with the LAN and the public. Currently, CMS aggressively estimates approximately 100 U.S. payers will participate in the LAN and participate in this information request (i.e., agree to submit aggregate payment data) in 2016.

2. Describe the procedures for the collection of information including:

- Statistical methodology for stratification and sample selection,

- Estimation procedure,

- Degree of accuracy needed for the pur­pose described in the justification,

- Unusual problems requiring specialized sampling procedures, and

- Any use of periodic (less frequent than annual) data collection cycles to reduce burden.

Response: For this data collection, sampling is not planned. Instead, the LAN will continually strive to increase the number of payers participating in the LAN who have committed to 1) matching or exceeding the Secretary’s Medicare FFS goal; and 2) submitting aggregate data that categorizes how plans are paying providers. At the end of each data collection, the LAN plans to say that X plans participated that represent Y beneficiaries (Z% of U.S. population) within a given calendar year. Among this proportion, A% of payment is in category 1, B% in category 2, and C% in categories 3 & 4. Subsequent data collections (at least annually) will report on the change among those plans that reported each year, and the number of new plans that are now reporting their payment categories.

3. Describe methods to maximize response rates and to deal with issues of non-response. The accuracy and reliability of information collected must be shown to be adequate for intended uses. For collections based on sam­pling, a special justification must be provid­ed for any collection that will not yield 'reliable' data that can be generalized to the uni­verse studied.

Response: This data collection uses those plans/payers participating LAN health plan payments as a surrogate for the US Health Care system as a whole, thereby allowing CMS gauging progress towards the Secretary’s 30% goal in a very straight forward manner. Complex statistical analyses will not be conducted on these data, and any discussion of LAN APM results will come with important caveats. There are a limited number of health plans in the United States (~200) with a variety of characteristics. We will only be able to describe the characteristics of LAN health plans and their percentages and acknowledge that the estimate could be too high or too low. That said, because the LAN has substantial engagement from many health plans already, CMS expects it will be straight forward to get plans representing well over 50% of the healthcare market to participate in this data collection in year 1. As such, it will help to understand directionality of payment methods.

As such, maximizing response rates and minimizing non-response bias is an important aspect of this collection. The LAN initially plans to pull together a cohort of LAN health plans to participate in a collaborative in early 2016. The purpose of this payer collaborative will be to both increase understanding of how the LAN proposes to measure progress towards APM adoption and build commitment from payers to respond to this information collection effort when asked. The LAN expects a response rate of 100% from all payers participating in this LAN payer collaborative and will thus continually strive to increase the number payers formally participating in the LAN payer collaborative.

4. Describe any tests of procedures or methods to be undertaken. Testing is encouraged as an effective means of refining collections of information to minimize burden and improve utility. Tests must be approved if they call for answers to identical questions from 10 or more respondents. A proposed test or set of tests may be submitted for approval separate­ly or in combination with the main collection of information.

Response: The LAN plans to pull together a cohort of LAN health plans to participate in a collaborative in early 2016. The purpose of this payer collaborative is to educate health plans on the APM Framework White Paper to ensure alignment on APM definitions across payers and to increase understanding with respect to how the LAN proposes to measure progress towards APM adoption. From this initial cohort of health plans who agree to participate in the LAN payer collaborative (approximately 20 health plans are slated to participate so far), the LAN will select fewer than 10 health plans to gather feedback on the proposed metrics, test the feasibility of the proposed survey instrument, develop more accurate collection burden estimates, and finalize an approach for a full-scale nationwide data collection effort of private and public health plans.

We anticipate a kick-off conference call towards the end of January 2016 for the payer collaborative, followed by two to three additional meetings. Once the pilot test is completed and the approach for full scale nationwide data collection is finalized, the details of that approach will be submitted to OMB and released to the public for a 30 day comment period through the LAN website, blog, newsletter, and other communication mediums as needed, such as webinars and listening sessions.

5. Provide the name and telephone number of individuals consulted on statistical aspects of the design and the name of the agency unit, contractor(s), grantee(s), or other person(s) who will actually collect and/or analyze the information for the agency.

Response:

Name	Telephone	Organization
Darren DeWalt	410-786-9731	CMS/CMMI/LDG
Alison Shippy	410-786-0438	CMS/CMMI/LDG
Dustin Allison	303-844-7031	CMS/CMMI/LDG
Kris Prendergast	443-200-4039	Mitre-CAMH
Julia Rollison	443-200-4039	Ripple Effect Communications
Andrea Caballero	714-815-8425	Catalyst for Payment Reform

d) LAN APM Characteristics Tracking

1. Describe (including a numerical estimate) the potential respondent universe and any sam­pling or other respondent selection method to be used. Data on the number of entities (e.g., establishments, State and local government units, households, or persons) in the universe covered by the collection and in the corre­sponding sample are to be provided in tabular form for the universe as a whole and for each of the strata in the proposed sample. Indicate expected response rates for the collection as a whole. If the collection had been conducted previously, include the actual response rate achieved during the last collection.

Response: The fourth type information request is intended to track other characteristics of alternative payment model adoption among certain LAN participant types—employers, providers, states, and so on. This information will help the LAN understand how other stakeholders are contributing to market trends and the pace of progress toward alternative payment model adoption across the U.S. healthcare system.

The universe of respondents for these type of information requests will be limited to LAN participants and/or subsets of LAN participant types. An example could be an information request on employer practices around purchasing health care for employees, to include the types of payment models used. In this case, the respondent universe would be limited to participating LAN employers, and CMS would not expect to engage in sampling a certain proportion of U.S. employers to obtain generalizable data. Instead, the LAN would continually engage employers to participate in the LAN and to increase the number of respondents.

To date, approximately 4,670 individuals, employers, payers, consumers, providers, local, state and federal agencies have signed up to participate in the LAN. Each LAN organization type could be asked to provide information related to participation in alternative payment models, depending on the purpose of the information request. More information will be provided on the potential respondent universe as the actual information requests are developed in the coming weeks and months.

2. Describe the procedures for the collection of information including:

- Statistical methodology for stratification and sample selection,

- Estimation procedure,

- Degree of accuracy needed for the pur­pose described in the justification,

- Unusual problems requiring specialized sampling procedures, and

- Any use of periodic (less frequent than annual) data collection cycles to reduce burden.

Response: The respondent universe for this information request will be limited to participating LAN stakeholders, and CMS does not expect to engage in sampling a certain proportion of these LAN participants to obtain generalizable data. Instead, the LAN will continually engage stakeholders to participate in the LAN and to increase the number of potential respondents. More information will be provided on the information collection procedures as the actual information requests are developed in the coming weeks and months.

3. Describe methods to maximize response rates and to deal with issues of non-response. The accuracy and reliability of information collected must be shown to be adequate for intended uses. For collections based on sam­pling, a special justification must be provid­ed for any collection that will not yield 'reliable' data that can be generalized to the uni­verse studied.

Response: The overarching reason for this type of data collection is for the LAN to understand how other stakeholders are contributing to market trends and the pace of progress toward alternative payment model adoption across the U.S. healthcare system—such as, in 2015, 35% of participating LAN employer health care purchases were tied to APMs. Complex statistical analyses will likely not be conducted on these data, and any discussion of LAN APM results will come with important caveats.

Maximizing response rates and minimizing non-response bias will be important. The LAN will continually engage stakeholders relevant to these information requests to participate in the LAN and to increase the number of potential respondents. More information will be provided on the information collection procedures as the actual information requests are developed in the coming weeks and months.

4. Describe any tests of procedures or methods to be undertaken. Testing is encouraged as an effective means of refining collections of information to minimize burden and improve utility. Tests must be approved if they call for answers to identical questions from 10 or more respondents. A proposed test or set of tests may be submitted for approval separate­ly or in combination with the main collection of information.

Response: The LAN plans to select fewer than 10 LAN participants, depending on the survey, to gather feedback on the proposed metrics, test the feasibility of the proposed survey instrument, develop more accurate collection burden estimates, and finalize an approach for a full-scale LAN data collection effort for information requests of this type. Once these pilot tests and the approach for full-scale data collection are finalized, the details for each approach will be submitted to OMB and released to the public for a 30 day comment period through the LAN website, blog, newsletter, and other communication mediums as needed, such as webinars and listening sessions.

5. Provide the name and telephone number of individuals consulted on statistical aspects of the design and the name of the agency unit, contractor(s), grantee(s), or other person(s) who will actually collect and/or analyze the information for the agency.

Response:

Name	Telephone	Organization
Darren DeWalt	410-786-9731	CMS/CMMI/LDG
Alison Shippy	410-786-0438	CMS/CMMI/LDG
Dustin Allison	303-844-7031	CMS/CMMI/LDG
Kris Prendergast	443-200-4039	Mitre-CAMH
Julia Rollison	443-200-4039	Ripple Effect Communications
Andrea Caballero	714-815-8425	Catalyst for Payment Reform

1 Rajkumar R, Conway PH, Tavenner M. CMS: Engaging multiple payers in payment reform. JAMA. 2014 May 21: 311(19):1967 8.

9