14. SupportingStatement_PartB_19July2016

OMB Control Number: 0938-1300

Supporting Statement – Part B

Collection of Information Employing Statistical Methods

1. Respondent Universe and Sampling Methods

The Centers for Medicare & Medicaid Services is interested in collecting information to determine beneficiaries’ satisfaction, understanding, and experience with the Healthy Indiana Plan (HIP) 2.0 in the state of Indiana. This data collection effort will help inform beneficiaries’ experiences in general under HIP 2.0, as well as specifically how experiences vary for HIP Basic and HIP Plus enrollees and for key population subgroups (e.g., by health status, income). This data collection effort includes two primary components including the implementation of a beneficiary survey in two waves and a qualitative component encompassing two site visits, key informational interviews, and focus groups. These are detailed below and in the Appendices.

The respondent universe of the beneficiary survey includes current HIP 2.0 enrollees and disenrollees. Table 1 summarizes the number of beneficiaries in the respondent universe. HIP 2.0 enrollees and disenrollees are broken up into eight different subgroups. The enrollee subgroups are defined by the benefits packages available to HIP beneficiaries and the length of time enrolled in HIP 2.0 (i.e., new enrollees). The disenrollee subgroups are defined by actions taken by the beneficiary that would result in disenrollment, being locked out of HIP for six months, or being locked in to HIP Basic for six months.

Table 1

Survey Instrument	HIP Respondent Groups	Members Selected into Sample	Target Completed Responses
HIP Enrollee	Enrollees: Basic	4,570	1,476
HIP Enrollee	Enrollees: Plus	4,570	1,476
HIP Enrollee	Enrollees: w/o ED copay (from control group)	2,285	738
HIP New Enrollee	New Enrollees: Basic & Plus	1,950	630
HIP Disenrollee & Lockout	Basic Disenrollees	282	91
HIP Disenrollee & Lockout	Plus Disenrollees	282	91
HIP Disenrollee & Lockout	Lockouts: Basic enrolled, formerly Plus	1,053	340
HIP Disenrollee & Lockout	Lockouts: Plus	1,053	340
TOTAL		16,045	5,182

Based on previous response rates associated with this target population, we are anticipating a 32% response rate.

The focus groups will include HIP 2.0 enrollees; the informational interviews with key stakeholders will include Medicaid officials, state budget and financial officers, health plan administrators, health care providers, health care industry representatives, consumer advocates, and community based enrollment assistors, among others. All participants selected for the qualitative component will rely on the voluntary participation of the individuals. Throughout the recruitment process as well as at the beginning of each focus group, researchers will emphasize that participation is voluntary and all comments will be kept private to the extent allowed by law.

The participant list identified for focus group recruitment will be derived from enrollment files. We will request from the State of Indiana two randomly drawn lists of current HIP 2.0 enrollees from the State HIP 2.0 enrollment files. These two lists will serve as the sampling frame with 300 currently enrolled individuals per list--one list for HIP Basic and one list for HIP Plus. Apart from being currently enrolled in HIP Basic or HIP Plus, individuals will also meet the following criteria: Adult enrollees (ages 18-64) whose primary language is English; enrolled in HIP 2.0 (Basic or Plus) for at least four (4) months at the time of sampling; home address in a Zip code within Interstate 465 highway encircling Indianapolis. For each sample person we will request the following information from the HIP 2.0 enrollment files: name, contact information (street address, phone number, email address if available, age, gender, race/ethnicity, and income level preferably as a percentage of the federal poverty level (FPL).

The identification and selection of specific individuals for the informational interviews will occur through an iterative process. First, researchers will first examine public records, periodicals, and literature to develop a preliminary roster of individuals and organizations to be considered. Second, a series of telephone conversations will need to occur with state officials in order to draft a roster of stakeholders, verify roles and determine if they currently play key roles relevant to HIP 2.0. Third, after the key individuals are verified and identified researchers will contact key organizations (beyond state officials) to verify the appropriate experts and leaders most important to interview, and to inquire whether there are other stakeholders researchers may have overlooked and should be added to the list. Lastly, when the stakeholder list is finalized interviews will be scheduled to occur during the site visit.

2. Procedures for the Collection of Information

Statistical Methodology, Estimation, and Degree of Accuracy

The sampling frame consisting of Medicaid beneficiaries in Indiana will be classified into strata representing beneficiary subgroups defined in Table 1 of this document. From the sampling frame, SSS data analysts and statisticians will draw a stratified random sample of individuals, with the number selected corresponding to the respondent-specific estimates listed in Table 1.  If the desired sample sizes are not initially achieved, the SSS survey team will consider an oversample to reach required numbers for a certain subgroup (stratum). Sampling weights would then be applied to the subgroup (stratum) in analysis to account for oversampling.

Unusual Problems Requiring Specialized Sampling Procedures

The different versions of the HIP 2.0 beneficiary survey will require different sampling considerations in order to guarantee the appropriate target populations for each questionnaire. To account for the complex sampling requirements (i.e., to classify the sampling frame into strata), we will primarily be looking at beneficiary enrollment variables. The variables will include, for example: length of time enrolled in HIP 2.0, failure to pay contributions on time, indications of beneficiaries switching between HIP Basic and HIP Plus plans (e.g., due to failing to pay contributions), the type of managed care health plan HIP beneficiaries have, and whether they are a part of the state’s emergency room copay control group (i.e., 5,000 HIP beneficiaries for whom the graduated $25 copay does not apply).

After controlling for the variables needed to identify the appropriate beneficiaries for each survey questionnaire, we will be drawing a stratified random sample of HIP 2.0 enrollees and disenrollees. All standard CAHPS administration specifications will be followed – including removing duplicates and ensuring that only one member of a household is surveyed.

Mode of Administration

The proposed beneficiary survey will be administered through a multi-mode data collection including Priority mail to beneficiaries, phone follow-up to non-responders, and a web survey option. The web option will be optimized to ensure that survey participants can complete the questionnaire on mobile devices (e.g., cell phones, tablets, etc.). All survey questionnaires and materials will be available in Spanish. Bilingual interviewers will also be available at the survey implanting firm Thoroughbred’s phone facilities.

Participants will receive a pre-notification letter inviting everyone to participate in the survey online. The pre-notification letter will be followed by a paper mail-in survey one week later. Non- respondent participants will receive a maximum of two reminder cards and two paper mail-in surveys accompanied with a pre-paid return envelope. Phone follow-up will occur as appropriate throughout the eight week data collection period.

Interviewers will contact participants and will attempt to complete the survey via telephone with non-respondents after the second mail follow-up. Telephone data collection will begin at the appropriate number of days after the second questionnaire is sent. The sample records will be transferred to the Computer Assisted Telephone Interview in (CATI) system. During the entire telephone follow-up period, newly received mail questionnaires will be accepted and scanned, and those respondents will be suppressed from CATI interviewing. Thoroughbred will make the appropriate number of dialing attempts on each sample record.

The CATI system will be programmed with built-in range and consistency checking and with appropriate skip patterns in place. The system will be set up so that it will automatically schedule callbacks, suspend an interview to be resumed at a later date where left off, and handle closed- ended questions, open-ended questions and multiple response questions. Thoroughbred’s CATI system is also designed to accommodate multiple questionnaire versions with different question subsets so that the appropriate version comes up based on the respondent’s tracking number.

Interviewers are allowed to go back as far as needed to correct changed responses or inaccuracies.

3. Methods to Maximize Response Rates and Data Reliability Response Rates

The data collection team will implement a number of procedures to maximize response rates. Pre- notification letters will be mailed to all sample members, followed 1 week later by cover letters and surveys. Mailing reminders and phone follow-ups will occur as appropriate over the course of 5 weeks. Approximately 1 week after mailing the initial survey, a reminder card will be mailed to all sampled members. Five weeks after mailing the initial questionnaire, a second survey and cover letter will be sent to all non-respondents. A second reminder postcard will be mailed to all non-respondents approximately 1 week after mailing the second survey packet. Three weeks after the second questionnaire is mailed, telephone follow-up will be initiated to members who have not yet responded and completed a mail survey (and who have not refused to participate). In addition, we have the capacity to support bilingual interviewing at our phone facilities. We anticipate that these procedures, the offer of multiple options for completion, and the incentive payment will result in an overall response rate in the 30 percent range, with some variation among program groups.

Issues of Non-Response

Subsequent to the fielding of the beneficiary surveys we intend to conduct a nonresponse analysis. Survey non-respondents will be compared to survey respondents on demographic and other characteristics to see if there is a systematic difference between those who respond and those who do not. Sampling weights will adjust for any differential response, as needed.

Reliability of Data Collection

The beneficiary survey was developed from a number of previously vetted and fielded beneficiary surveys. Several existing beneficiary surveys were examined during survey development including: The Consumer Assessment of Healthcare Providers and Systems (CAHPS), CAHPS Qualified Health Plan Survey, Nationwide Medicaid CAHPS Survey, CAHPS Supplemental Items for Adult Questionnaires (CAHPS Healthy Plan Survey 4.0), Behavioral Risk Factor Surveillance System (BRFSS), National Health Interview Survey (NHIS), Iowa Wellness Plan, Healthy Indiana Plan 1.0 Beneficiary Surveys: Enrollee and Leaver Survey, and the Healthy Indiana Plan 2.0 Beneficiary Surveys: HIP Basic; HIP Plus; Never member, no POWER account contribution made. Revisions to the survey instruments have been made based on the results of the instrument pretest, feedback from CMS, review of survey experts, and public comments received from the publication of the 30-day Federal Register Notice.

The use of a programmable survey will help to ensure the consistency of the data, and will provide another option for participants to complete the survey. During the production phase, programmers create an image of an actual survey from the production print run. The programmer will first identify unique text within the survey that can be used as registration points so the software can identify all data capture areas as they relate to these registration points and identify each data capture location and assign the values to be captured along with any instructions for data cleansing. Once the program is completed, Quality Control will run a full test batch through the entire data capture process. The data captured is then compared to the actual survey to make sure each field was captured correctly. The final step is to review the program definitions for each field to ensure that they are programmed correctly.

Once in production, the Kofax Capture software will attempt to recognize respondent marks as well as text by conducting optical mark recognition (OMR) and Optical Character Recognition/Intelligent Character Reading (OCR/ICR). It is thereby able to read respondents’ marks in check boxes along with any numeric responses they make. It can also read pre-printed barcodes or other pre-printed, numeric information on the survey. The verifications and consistency checks are built into the system and standardize the procedures, all of which helps ensure the reliability of the data collection methods and the data collected through those methods. Survey information will be collected electronically by the system Survox Web Survey version

2. The information will be stored on the Thoroughbred Research Group secure servers and then uploaded to the SSS Secure Data Center via secure file transfer.

Research Goals and Intended Use of Data Collected

Our evaluation of HIP 2.0 has three main research goals:

• Provide in-depth understanding of the design, implementation, ongoing operation and impacts of HIP 2.0

• Provide a detailed examination of beneficiary experience under HIP 2.0

• Estimate impacts HIP 2.0 on health insurance coverage, access, service use, affordability as well as quality of care and health and health behaviors

Collecting both qualitative and quantitative data will help to inform and achieve these goals.

The qualitative components (focus groups and informational interviews with key stakeholders) will enable us to assess how the launch of HIP 2.0 proceeded and also to identify the successes and challenges in establishing and administering HIP 2.0. Additionally, the informational interviews will provide important insights into how major stakeholders to HIP 2.0 perceive the operations and effectiveness of the program, from its beginnings to its maturity. Focus groups will further enrich the evaluation by capturing the “voices” of adults affected by HIP 2.0, providing valuable details about their experiences and concerns, details that cannot be obtained in the beneficiary survey. Apart from providing an in-depth understanding of the design, implementation and operations of HIP 2.0 and a detailed consumer perspective on the program, data from informational interviews will inform our impact analyses by guiding these analyses and providing valuable context for interpreting the results.

An immediate objective of the implementation of the beneficiary survey (quantitative component) will be to evaluate the waiver of the non-emergency medical transportation (NEMT) benefit for all new adults (those at or below 138% of FPL) in HIP 2.0. The survey includes a section focusing solely on the beneficiary experience and understanding of their transportation to and from their health care visits. The survey questions are organized to identify individuals with and without the NEMT benefit and to determine if they have experienced a barrier in accessing health care services due to transportation. For example, we ask beneficiaries “In the last 6 months, was there any time when you needed health care but did not get it because you could not pay for transportation or could not get transportation?” If respondents indicate that they have experienced a barrier to care in the last 6 months, we proceed with asking them about what specific types of health care they were unable to get (1) because they could not pay for transportation, (2) could not get transportation, (3) if they had no trouble with transportation, or (4) if it was not applicable. Additionally, we also have a survey question assessing if beneficiaries had any transportation (in general) to get to and from their doctor’s office to get health care services they needed in the last 6 months. The data collected in this survey section will be crucial in helping to identify whether there is a barrier to care for HIP2.0 beneficiaries due to lack of transportation. Most importantly if there is such a barrier, to identify whether that barrier plays a significant role in health outcomes for beneficiaries.

In addition to the policy focus on the NEMT waiver, the beneficiary survey will serve as a vital data collection opportunity for information on other policies being tested under the HIP 2.0 demonstration including the HIP 2.0 beneficiary lockout population¹. The beneficiary survey instruments have been designed to adequately collect beneficiary experience and understanding around several issues concerning the lockout population. Given that a portion of the lockout population is disenrolled from HIP 2.0 (i.e., no access to health insurance available under HIP 2.0), we have designed the disenrollee and lockout beneficiary survey instrument to inform descriptive analyses focusing on HIP 2.0 elements, including beneficiary understanding and awareness of the consequences of non-payment. For example, we ask if respondents experienced any barriers to care due to transportation and cost after they were no longer enrolled in HIP 2.0. Additionally, we ask detailed questions about POWER accounts and contributions with respect to affordability, understanding, preventive health care services, and consequences of non-payment. The informational interviews and focus groups will also explore these other policies. This additional information is being collected during the same period as the NEMT survey questions to improve efficiency of resources.

4. Tests of Procedures or Methods

Emergency OMB clearance was obtained to conduct testing of the beneficiary survey questionnaires. The survey testing period was used to refine the beneficiary survey questions, minimize burden on respondents, and improve utility. A total of 15 HIP 2.0 enrollees and disenrollees participated in the beneficiary survey instrument testing and respondent debriefing.

¹ The federal evaluation of HIP 2.0 characterizes the ”lockout population” as individuals who failed to pay their contribution and have either been moved down to HIP Basic (at or below 100% FPL) or locked out of HIP 2.0 for six months (above 100% FPL).

Participants completed the questionnaires in the Briljent firm’s local offices in Fort Wayne and Indianapolis, Indiana, and were debriefed on their understanding of the survey instruments.

As part of the survey testing, interviewers monitored how long it took for respondents to complete the survey. Participants were debriefed on any survey questions and/or sections that were unclear. The average survey completion time aligned with the projected 15 minute average. After completing the survey instrument, the interviewer asked participants a series of debriefing questions that focused on the survey overall, survey sections, and specific survey questions. All interviews were recorded, transcribed, and coded to capture participant comments on the survey structure and any questions that required revisions. The data collection did not focus on the actual survey instrument responses but rather on how well the questions were understood and how well they conveyed the research intentions. All identifiable information was kept private to the extent allowable by law and destroyed subsequent to the conclusion of all interviews.

Revisions to the questionnaire were made based on the results of the survey instrument testing. Few revisions were needed and are documented in the HIP 2.0 Beneficiary Survey crosswalk. The crosswalk identifies question changes, additions and deletions, and provides justification for the revisions. Several of the survey revisions aligned with the recommendations obtained during the public comment period. The survey development team has also proposed additional minor refinements that also are described and documented in in the survey crosswalk.

5. Individuals Consulted on Statistical Methods

The following persons outside of CMS contributed to, reviewed, and/or approved the design, instrumentation and sampling plan:

Name	Affiliation	Telephone Number
Niranjana Kowlessar, PhD	Social & Scientific Systems, Inc.	301-628-3275
Linda Piccinino, MPS, DrPH (ABD)	Social & Scientific Systems, Inc.	301-628-0244
Eugene Elbert, MS	Social & Scientific Systems, Inc.	301-628-3039

2