CMS-10427 Supporting_DocA_10 24 12

PART A

For-Profit PACE Study

Supporting Statement for Paper Reduction Act Submission

June 20, 2012

Nancy Duda

David Jones

Frank Yoon

Jennifer Ott

Contract Number: HHSM-500-2005-00025I/HHSM-500-T0005 Mathematica Reference Number: 06965.400 Submitted to: Centers for Medicare & Medicaid Services 7500 Security Boulevard, WB-06-05 Baltimore, MD 21244 Project Officer: Julia A. Zucco Submitted by: Mathematica Policy Research P.O. Box 2393 Princeton, NJ 08543-2393 Telephone: (609) 799-3535 Facsimile: (609) 799-0005 Project Director: David Jones

PART A For-Profit Pace Study Supporting Statement for Paper Reduction Act Submission June 20, 2012 Nancy Duda David Jones Frank Yoon Jennifer Ott

CONTENTS

Chapter Page

PART A: SUPPORTING STATEMENT FOR PAPERWORK REDUCTION ACT SUBMISSION 1

A. JUSTIFICATION 2

1. Circumstances Making the Collection of Information Necessary 2

2. Purpose and Use of the Information Collection 3

3. Use of Improved Information Technology and Burden Reduction 9

4. Efforts to Identify Duplication and Use of Similar Information 10

5. Impact on Small Businesses or Other Small Entities 11

6. Consequences of Collecting the Information Less Frequently 11

7. Special Circumstances Relating to the Guidelines of 5 CFR 1320.5 11

8. Comments in Response to the Federal Register Notice and Efforts to Consult Outside the Agency 11

9. Explanations of Any Payment or Gift to Respondents 12

10. Assurance of Confidentiality Provided to Respondents 12

11. Justification for Sensitive Questions 15

12. Estimates of Annualized Burden Hours and Costs 16

13. Estimates of Other Total Annual Cost Burden to Respondents and Record Keepers 18

14. Annualized Cost to the Federal Government 18

15. Explanation for Program Changes or Adjustments 18

16. Plans for Tabulation and Publication and Project Time Schedule 18

17. Expiration Date 20

TABLES

A.1 Data Elements in the For-Profit PACE Study Survey 4

A.2 For-profit PACE Study Survey Annual Burden Estimates 16

Part A: SUPPORTING STATEMENT FOR PAPERWORK REDUCTION ACT SUBMISSION

This package requests clearance for a survey to support a study of the for-profit Program of All-Inclusive Care of the Elderly (PACE) plans. The PACE program aims to provide integrated care and services to the frail elderly at risk of institutionalization to enable them to remain in the community. The study is being conducted by the Centers for Medicare and Medicaid Services (CMS); it is being implemented by Mathematica Policy Research (Mathematica).

Under the Balanced Budget Act of 1997 (BBA), the not-for-profit PACE plans were established as permanent providers under the Medicare and Medicaid programs. The BBA also mandated a demonstration of for-profit PACE plans. A 2005-2006 evaluation was conducted of the PACE program, but it focused solely on the not-for-profit PACE sites because not for-profit PACE plans were operational at that time. The objective of this study is to examine the for-profit PACE plans that have enrolled in the demonstration; specifically to understand the extent of variation between the permanent not-for-profit PACE plans and the for-profit PACE plans. The study will address the question of how access to and quality of care delivered to enrollees in these for-profit PACE plans differ from not-for-profit PACE plans.

The study will analyze the outcome measures of access to and quality of care for enrollees of the for-profit PACE plans and a sample of not-for-profit plans. While the study will use secondary data to identify comparable not-for-profit and for-profit PACE enrollee samples, primary data collection is necessary to obtain the data for the outcome measures. We will conduct a telephone survey of the sampled enrollees to collect data on the patient’s health and functional status, access to and satisfaction with health care services, satisfaction with caregivers, and quality of life. The survey for this study is adapted from the survey administered for the evaluation of the not-for-profit PACE programs. This document seeks Office of Budget and Management (OMB) approval to collect data through a telephone enrollee survey.

A. JUSTIFICATION

1. Circumstances Making the Collection of Information Necessary

The Program of All-Inclusive Care of the Elderly (PACE) originated with the conviction that elderly persons with disabilities could be maintained in the community for a longer time if an appropriate mix of special services and care coordination were provided to them. The BBA established not-for-profit PACE as a permanent program and established a demonstration of for-profit PACE plans. The BBA included a specific mandate to study the results of the for-profit demonstration. In 2008, an interim evaluation report of the not-for-profit PACE programs was issued. This evaluation did not include the for-profit sites because at that time, there were no for-profit sites in the PACE program. As in the interim report, this study is subject to limitations. Currently, there are only four for-profit PACE plans (under common ownership) that are participating in the demonstration and operating in Pennsylvania. This study compares the care for enrollees in the four for-profit PACE plans with care provided by other not-for-profit PACE plans located in Pennsylvania. Because of the small number of plans in the demonstration and their common ownership, this is essentially a case study of these plans, which limits the generalizability of the findings, but may provide useful insights into potential differences between for-profit and non-profit PACE plans.

As part of a broader study conducted by CMS of care provided to elderly enrollees in for-profit PACE sites, the proposed analysis will address the following research question:

• How does access to and quality of care delivered to enrollees in the for-profit PACE plans differ from not-for-profit PACE plans operating in Pennsylvania?

The extent of variation in access to and quality of care in for-profit and not-for-profit plans is an important policy issue for shaping the future of the PACE program. PACE represents a potentially growing component of the health care delivery system, and the number of for-profit plans is expected to grow in the near future. However, it is unclear whether the quality of care provided by for-profit PACE plans will resemble that of not-for-profit plans. A commonly cited concern in the long-term care field is that for-profit sites might favorably select enrollees who have potentially lower costs or the sites could provide less accessible, lower-quality care.

Although no studies have assessed the delivery of care by for-profit PACE plans, there is some evidence that for-profit status is correlated with access to and quality of care in related settings. For example, some studies have documented that for-profit nursing homes provide lower-quality care than their not-for-profit peers. In a comprehensive literature review, Hillmer et al. (2005) find that not-for-profit nursing homes are less likely to have poor quality-of-care practices and outcomes.

The proposed study will provide some evidence about differences in quality in the PACE setting and will help policymakers decide the future of for-profit PACE plans.

2. Purpose and Use of the Information Collection

Information will be collected and analyzed for this study by Mathematica Policy Research under Contract Number HHSM-500-2005-00025I/HHSM-500 T0005 with CMS, titled “For-profit PACE Study”. Findings from the analysis of the data will be included in briefings to CMS and a final report to Congress.

The survey administered in the not-for-profit PACE evaluation has been updated as part of this study. The data collected in the survey will be used to measure the outcomes of interest- differences in access to and quality of care delivered to PACE enrollees at the selected sites (see Table A1). To measure these key outcomes, the survey will collect data on access to and satisfaction with health care, personal care, and transportation assistance provided by the plans. The study will use these outcomes to estimate the differences in quality and access to care between the for-profit and not-for-profit PACE plans included in the study.

The survey will also collect health and functional status, quality of life, and demographic data. These data will be used to describe differences, if any, between enrollees of the two different types of plans and to improve the precision of the estimate of differences in quality and access to care by including these enrollee characteristics as covariates in the estimation model. All reporting of results will clearly note the limitations associated with having only a small number of for-profit plans with their common ownership.

Table A.1. Data Elements in the For-Profit PACE Study Survey

Question Number	Data Element	Comments	Use
Introduction and Screener
I1-31d	Introductory remarks.	Obtain contact information if sample member (SM) is in nursing home	Determination of need for proxy and identification of proxy.
I31e	Type of residence		Compare populations
I32	Assisted living		Compare populations
I33	Current living situation (who respondent lives with)		Compare populations
I34	Gender		Compare population/control variable
I35	Closing of introduction		Thank respondent
Section A. Prior Utilization
A1	Introduce questions on use of health care services		Lead in to A1a
A1a-A1d	Prior hospital admission in prior year	If had hospital admission, asks for dates of admission and discharge or length of stay	Compare populations
A2-A2c	Prior nursing home admission in prior year	If had nursing home admission, asks for dates of admission and discharge or length of stay	Compare populations
A19-A20	Satisfaction with rehabilitative care outside of PACE center		Measure of quality of care
A21-A22	Use of PACE center in last month		Measure access to care
A3-A7	Satisfaction with medical care	Satisfaction with medical care; treatment for health conditions; coordination of care; respect that doctors or heath care providers showed for what respondent had to say; and degree of energy to get need health care services	Measure of quality of care
A8-A10	Satisfaction with transportation provided by PACE		Measure of access to and quality of care
A14-A18	Satisfaction with specialist care provided outside of PACE center		Measure of access to and quality of care
Section B. Current Health Problems
B0-B0a	Fatigue and reinforcement probes		Determine respondent fatigue and need for rescheduling or reinforcement
B1-B2	Self-assessment of health status	Assessment of health status compared to other people same age and compared to a year ago	Compare populations
B3	Health conditions		Compare populations and control variable
B4a-B4b	Falls and fall-related injury in past 6 months		Compare populations and measure of quality of care
B5	Check to see if respondent is a proxy		To determine whether B6-B10 should be asked
B6-B10	Assessment of sample member’s behavior	Asked only if respondent is a proxy. Asks about straying, verbally disruptive, physically aggressive, and confusion/delusional behavior.	Compare populations
B13-B14	Weight loss of 10 pounds or more in past 6 months		Measure of quality of care
B15-B15a	Pain in last week		Compare populations
B16-B19	Depression/anxiety		Compare populations
B20-B24	Preventive Care	Asks about flu shot, pneumonia vaccine, hearing and eyesight testing	Measure access to care
Section C. Daily Activities and Caregivers
C0-C0a	Fatigue and reinforcement probes		Determine respondent fatigue and need for rescheduling or reinforcement
CA1	Receive help with eating in past week		Measure access to care and compare population
CA2	Use of special equipment		Measure ADL difficulty
CA7-CA7a	Length of time sample member has not eaten		Compare populations
CA8	Persons who provided help with eating		Determine whether section D should be asked
CA14-CA15	Need for additional help		Measure access to care
CB1	Receive help with getting around indoors in past week		Measure access to care and compare population
CB2	Use of special equipment		Measure ADL difficulty
CB7-CB7a	Length of time sample member has been unable to get around indoors		Compare populations
CB8	Persons who provided help with getting around indoors		Determine whether section D should be asked
CB14-CB15	Need for additional help		Measure access to care
CC1	Receive help with getting dressed in past week		Measure access to care and compare population
CC2	Use of special equipment		Measure ADL difficulty
CC7-CC7a	Length of time sample member has been unable to get dressed		Compare populations
CC8	Persons who provided help with dressing		Determine whether section D should be asked
CC14-CC15	Need for additional help		Measure access to care
CD1	Receive help with bathing in past week		Measure access to care and compare population
CD2	Use of special equipment		Measure ADL difficulty
CD7-CD7a	Length of time sample member has been unable to bathe		Compare populations
CD8	Persons who provided help with bathing		Determine whether section D should be asked
CD14-CD16	Need for additional help		Measure access to care
CE1	Receive help with using toilet in past week		Measure access to care and compare population
CE2	Use of special equipment		Measure ADL difficulty
CCE8-CE8a	Length of time sample member has been unable to use the toilet		Compare populations
C9	Use of urinary catheter or colostomy		Compare populations
C10	Trouble controlling bladder or bowels		Compare populations
CE11	Persons who provided help with using the toilet		Determine whether section D should be asked
CE13-CE15	Need for additional help		Measure access to care
CF1	Receive help with getting in or out of bed in past week		Measure access to care and compare population
CF2	Use of special equipment		Measure ADL difficulty
CF7-CF7a	Length of time sample member has been unable to get out of bed		Compare populations
CF8	Persons who provided help with getting in or out of bed		Determine whether section D should be asked
CF13-CF15	Need for additional help		Measure access to care
Section D. Attitudes and Satisfaction
D0-D0a	Fatigue and reinforcement probes		Determine respondent fatigue and need for rescheduling or reinforcement
D1-D6	Satisfaction with PACE caregivers	Satisfaction with concern for respondent as a person, emotional support provided, attention paid to what respondent had to say, how well needs were taken care, completion of work, and rushing through work.	Measure quality of care
Section E. Overall Quality of Life and Demographics
E0-E0a	Fatigue and reinforcement probes		Determine respondent fatigue and need for rescheduling or reinforcement
E1	Overall satisfaction with quality of life		Measure quality of care
E2	Determine amount of choice sample member has day to day		Measure quality of care
E3-E4	Contact with family and friends in past week		Compare populations
E5	Participation in social activities		Compare populations
E6	Marital status		Compare populations and control variable
E7	Health of spouse		Compare populations
E8	Arrangements for advance directives		Measure quality of care
E9	Level of education		Compare populations and control variable
E10-E11	Ethnicity and race		Compare populations and control variable
E13	Household income		Compare populations and control variable
E16-E17	Closing		Thank respondent
Section F. Interviewer Observations
F1-F5	Interviewer’s assessment of respondent’s accuracy and ability to answer questions		Sensitivity testing

3. Use of Improved Information Technology and Burden Reduction

For the enrollee survey, we have selected technologies that will enable us to obtain reliable information in an efficient way that minimizes respondent burden.

The enrollee survey will be administered as a telephone survey with in-person follow up for the non-respondents. Based on the sample characteristics, we have determined that telephone administration of the survey is the least burdensome approach. A telephone survey is well-suited for this sample which may have difficulty reading and filling out a self-administered hard copy survey due to physical and cognitive impairments and may be less likely to complete a web survey due to limited computer literacy or computer access. With the expected high level of cognitive impairments among this population, we will require significant participation of proxy respondents. We will work with the PACE sites to identify and obtain contact information for enrollees’ caregivers. However, in cases where this information is not available, determining the need for and identification of the appropriate proxy respondent is most efficiently conducted in a telephone survey.

The telephone survey will use a variety of techniques to reduce respondent burden: (1) Computer Assisted Telephone Interviewing (CATI), (2) automated dialing, and (3) automated call scheduling. The CATI system limits respondent burden by automating interviewer instructions and skip logic, so that interviews progress quickly and smoothly from question to question. These features will be very helpful for the initial screening process and for the daily activities section of the survey which has complicated skip logic. Computer assisted interviewing minimizes interviewer error and provides consistency checks and probes. This technology eliminates the need to call back respondents to obtain missing data as errors and inconsistencies as these items are corrected as they occur during the actual interview. These systems also assist survey management by producing reports on screening and interview outcomes, yield rates, item nonresponse rates, and interviewer productivity.

The autodialer, which is linked to the CATI system, will dial all calls; this virtually eliminates dialing errors and improves interviewer efficiency.

The automated call scheduler manages telephone interviewer assignments by scheduling and rescheduling calls to ensure that they are made according to optimal calling patterns, that all appointments are kept, and that cases requiring refusal specialists or foreign language speakers are routed to the appropriate interviewers.

We will offer in person administration of the survey to respondents who may have difficulty speaking on the phone or who may be fearful of providing this information to an unknown person on the telephone. We will also use in-person follow up for sample members who we have not been able to contact by telephone. Depending on the preference of the sample member, we will conduct these in-person interviews at the PACE site or at the sample member’s home.

4. Efforts to Identify Duplication and Use of Similar Information

No other study of for-profit PACE plans has been conducted. Furthermore, this information cannot be collected from any other source. The outcomes of interest, access to and quality of care provided by the PACE programs, are subjective data based on perception and opinion. This information is uniquely related to this program and would not be available in any administrative or other secondary dataset.

While the Medicare Health Outcomes Survey – Modified (MHOS-M), OMB 0938-0701, is administered annually to a random sample of PACE enrollees, the MHOS-M survey does not collect the full range of data required for the study. Although the survey includes several items that are drawn from the MHOS-M instrument, we will be unable to use this data as the MHOS-M and for-profit PACE study will be administered to different samples of enrollees.

The survey collects data on hospital and nursing home utilization which may be captured from administrative databases maintained by the PACE plans. However, information from such reports has not been found to be consistently reliable.

5. Impact on Small Businesses or Other Small Entities

This data collection does not involve small businesses or other small entities.

6. Consequences of Collecting the Information Less Frequently

This study of the for-profit PACE plans is mandated by the BBA. If the survey is not conducted, the study could not effectively measure the quality and access to care for the not-for-profit and for-profit PACE plans included in the analysis. Without such data, it will be difficult to determine whether these for-profit PACE programs are providing quality and access to care comparable to their not-for-profit counterparts in the state. Policymakers will not have the results needed to inform the decision as to whether the for-profit PACE plans should be made permanent providers under the Medicare and Medicaid programs.

7. Special Circumstances Relating to the Guidelines of 5 CFR 1320.5

There are no special circumstances involved with the data collection.

8. Comments in Response to the Federal Register Notice and Efforts to Consult Outside the Agency

a. Federal Register Notice

A 60-day notice to solicit public comments was published in the Federal Register on April 4, 2012. Comments received in the first comment period were addressed prior to submission to the Office of Management and Budget.

b. Efforts to Consult Outside the Agency

A number of individuals outside CMS have consulted on the updating of the not-for-profit PACE survey instrument for administration in this study. These individuals are researchers who have designed or conducted similar surveys for other CMS demonstration evaluations. They are:

• David Jones, Ph.D., Mathematica Policy Research

• Nancy Duda, Ph.D., Mathematica Policy Research

• Robert Schmitz, Ph.D., Mathematica Policy Research

• Karen Cybulski, M.S., Mathematica Policy Research

9. Explanations of Any Payment or Gift to Respondents

No payment will be offered to respondents or proxies who complete the survey.

10. Assurance of Confidentiality Provided to Respondents

Mathematica will follow procedures for assuring and maintaining confidentiality, consistent with provision of the Privacy Act of 1974, as amended (45 CFR 5b). Prior to the start of data collection, each sample member will be sent an advance letter and information brochure describing the study and informing the sample member of the following:

• participation in the survey is voluntary

• information provided is held in strict confidence and used only for study

• names will never be used in reporting the results

• answers will not affect eligibility for any federal, state, or local government programs or receipt of benefits from such programs

Once data collection begins, every respondent to the survey will again receive the information regarding confidentiality protection at the beginning of his or her interview, as part of the interviewer’s introduction.

a. Mathematica’s Confidentiality Policy

It is the policy of Mathematica to efficiently protect confidential information and data, in whatever medium it exists, in accordance with applicable federal and state laws and contractual requirements. In conjunction with this policy, all Mathematica staff is required to:

• Comply with the Mathematica Confidentiality Pledge and the Mathematica Security Manual procedures to prevent the improper disclosure, use or alteration of confidential information. Staff may be subjected to disciplinary and/or civil or criminal actions for knowingly and willfully allowing the improper disclosure or unauthorized use of confidential information.

• Access confidential information and proprietary information only in performance of assigned duties.

• Notify their supervisor, the Mathematica project director, and the Mathematica Security Officer if confidential information has been disclosed to an unauthorized individual, used in an improper manner, or altered in an improper manner.

• Report immediately to both the Mathematica project director and the Mathematica Security Officer all contacts and inquiries concerning confidential or proprietary information from unauthorized staff and non-Mathematica personnel.

Survey data for this and all other Mathematica projects are maintained in the Survey Operations Center, access to which is limited. During working hours only Mathematica personnel have access to the center. Visitors are required to report to the center managers or supervisors upon entering the building and must sign a confidentiality agreement. The Survey Operations Center is locked during nonworking hours and access is limited to the center managers, supervisors, and Mathematica senior systems analysts. Interviewers and other Mathematica personnel do not have access to the center after working hours.

To ensure that data are protected and that the confidentiality of sample members is maintained, access to identifying information is limited to project personnel for the time it is needed. The Survey Operations Center is equipped with locked file cabinets for storing hardcopy questionnaires and other paper documents. Access to these storage units is limited to project managers and supervisors. Once an interview is complete, identifying information is not kept with completed interview data, whether the data are in hardcopy or machine-readable form and is not included on the data files provided to clients. Hardcopy questionnaires and other paper documents are routinely returned to the client or destroyed after completion of the project in accordance with contract specifications. Similarly, data are purged from the computer system at the conclusion of a project or maintained for a set period as required by contract specifications.

The Survey Operations Center also houses Mathematica’s CATI computer system and interviewing stations. Data security barriers guard against unauthorized access to all files and directories on the CATI system, including read and write access to both data and program files and unauthorized execution of programs. Data security barriers included the login ID/password barrier and access-permission barriers assigned to files and directories. Access permissions grant read, write, and execute access to individual users or members of a group of users. Only users who have the correct access permission can access files and directories.

Access to Mathematica’s CATI system differs for its three types of users—interviewers, interviewer supervisors, and systems analysts/programmers. Interviewers have access only to the cases for the survey to which they been assigned and only to those cases that are still in progress. Interviewer supervisors are also restricted in their access to the CATI system. The system’s analysts/programmers have access to most of the user files on the system, including the standard unrestricted UNIX functions and all CATI survey instruments and cases. This access is required for survey monitoring, maintenance and quality assurance. The system administrator, as the superuser, has unrestricted access to the entire system and is responsible for the daily backup of the computer system among other tasks.

b. Basis for Confidentiality Assurance

Assurance of confidentiality for the survey is made on the basis of the Privacy Act of 1974, as amended (45 CFR 5b) which stipulates that information may be released by the Department of Health and Human Services without written consent for a purpose compatible with the purpose for which the information was collected.

Respondent confidentiality will be assured by adherence to Section 903(d) of the Public Health Service Act (42 USC 299a-1(s)) as follows:

“No information, if an establishment or person supplying the information or described in it is identified, obtained in the course of activities undertaken or supported under this title may be used for any purpose other than the purpose for which it was supplied unless such establishment or person has consented (as determined under regulations or the Secretary) to its use for such other purpose. Such information may not be published or released in other form if the person who supplied the information or who is described in it is identifiable unless such person as consented (as determined under regulation of the Secretary) to its publication or release in other form.”

11. Justification for Sensitive Questions

The survey contains questions that are routinely asked by providers or health plans when obtaining a patient history for assessing the need for care and would not commonly be considered sensitive when asked in that context. However, in the survey, there are four types of questions that might make some individuals uncomfortable because they deal with highly personal and sensitive issues. These questions are about (1) toileting, (2) incontinence, (3) Living Will/Durable Power of Attorney, and (4) income.

Information on physical functioning provide important measures for assessing patient well-being and burden of disease in the analysis of quality of care, especially for individuals with chronic illness, so questions about incontinence and using the toilet provide key data for the study. We have found that elderly or disabled sample members rarely object to question on incontinence or toileting if asked as part of a battery of questions on physical functioning.

Questions about a Living Will and a Durable Power of Attorney might also be considered sensitive by the general population. However, the study population will be used to answering questions dealing with this topic having been asked similar questions on previous occasions by health care and supportive services providers.

Income is another important variable because the financial resources available to sample members may affect their ability to purchase services not covered by Medicare or Medicaid. We will ask only for the range within which the respondent’s income falls, rather than the exact amount. Respondents may choose to refuse to answer any question in the survey including the questions described in this section.

These questions were adapted, without any or with very slight modifications, from the survey administered for the not-for-profit PACE evaluation. A review of the item response frequency for these items showed there were very few refusals to answer these questions.

12. Estimates of Annualized Burden Hours and Costs

Table A.2 presents our burden estimate for the survey.

Table A.2. For-profit PACE Study Survey Annual Burden Estimates

a. Maximum Total Respondent Costs, 100% proxy responses

Data Collection Effort	Estimated Number of Respondents	Number of Responses Per Respondent	Total Number of Responses	Average Burden Hour Per Response	Total Burden Hours	Hourly Wage Rate	Total Respondent Costs
For-profit PACE enrollees	406	1	406	.55	223	20.70*	$4,616
Not-for-profit PACE enrollees	407	1	407	.55	224	20.70*	$4,637
Total	813	1	813	NA	467	NA	$9,253

b. Total Respondent Costs based on 80% survey response rate, 100% proxy responses

Data Collection Effort	Estimated Number of Respondents	Number of Responses Per Respondent	Total Number of Responses	Average Burden Hour Per Response	Total Burden Hours	Hourly Wage Rate	Total Respondent Costs
For-profit PACE enrollees	325	1	325	.55	179	20.70*	$3,705
Not-for-profit PACE enrollees	325	1	325	.55	179	20.70*	$3,705
Total	650**	1	650	NA	358	NA	$7,410

c. Total Respondent Costs based on 80% survey response rate, 60% proxy responses

Data Collection Effort	Estimated Number of Respondents	Number of Responses Per Respondent	Total Number of Responses	Average Burden Hour Per Response	Total Burden Hours	Hourly Wage Rate	Total Respondent Costs
For-profit PACE enrollees	325	1	325	.55	179	20.70*	$2,220***
Not-for-profit PACE enrollees	325	1	325	.55	179	20.70*	$2,220***
Total	650**	1	650	NA	358	NA	$4,440

* The hourly rate for the proxy respondents is based on the average hourly wage in Pennsylvania.

** Assuming an 80 percent response rate, 650 respondents of the 813 enrollees, 325 for-profit and 325 not-for-profit enrollees, will complete the survey.

** The total respondent costs for the enrollees is based on the assumption that 60 percent of the respondents will be proxy respondents at the hourly wage rate of $20.70 and the remaining 40 percent of the respondents will be elderly respondents who are retired and have an hourly rate of zero.

The survey will take an average of 33 minutes to complete. We estimate that the maximum total burden associated with the survey is 467 hours. The survey will be administered to a maximum of 813 respondents (406 for-profit PACE enrollees and 407 not-for-profit PACE enrollees), for a maximum cost total of $9,253 (Table A.2.a). Assuming an 80 percent response rate, the survey will be administered to 650 respondents (325 for-profit PACE enrollees and 325 not-for-profit PACE enrollees), for a burden total of 358 hours and cost of $7,410 (Table A.2.b). We anticipate that 60 percent, or 390 respondents, of the 650 respondents will be proxy respondents who are working and 40 percent of the respondents are the retired elderly, reducing the total estimated cost to $4,440 (Table A.2.c).

13. Estimates of Other Total Annual Cost Burden to Respondents and Record Keepers

There are no direct costs to individual respondents.

14. Annualized Cost to the Federal Government

The total cost to CMS of conducting this study is $524,922. This total included the cost of developing the survey instrument ($35,805), the cost of the data collection ($213,817), and the costs of the project design report, analysis of survey data, preparation of report and data file, and presentation to CMS ($230,784). These estimates are based on the contractor’s costs for collecting and tabulating survey data, including labor and travel; other direct costs for computer, telephone, postage, reproduction, fax, printing, and survey facilities; and indirect costs for fringe benefits, general and administrative costs, and fees. In addition to the contractor costs, the total includes the cost of 1 FTE Federal employee, GS-13 (locality pay area of Washington-Baltimore-Northern Virginia) at 25% effort for 2 years, or $44,516.

15. Explanation for Program Changes or Adjustments

This is a new collection.

16. Plans for Tabulation and Publication and Project Time Schedule

a. Plans for Tabulation and Analysis

We will employ a two-part analytic framework to determine whether the for-profit PACE plans taking part in the demonstration deliver access to and quality of care that differ from care provided by not-for-profit plans in the state. We will begin with a descriptive analysis that presents the differences in mean access and quality measures by plan type. We will test the statistical significance of the differences while accounting for the sample design described in Section B. The measures of access to and quality of care will be taken from the survey (for example, questions on patients’ ability to access various services care and their satisfaction with care).

The second component of the framework is a multivariate analysis, which will estimate the differences in quality and access in the for-profit and not-for-profit plans included in the study, while accounting for patient and site characteristics. The patient characteristics will be collected primarily through the survey (for example, health conditions and difficulty with activities of daily living (ADLs)). The plan characteristics will come from CMS administrative data (for example, number of enrollees and tenure of the plans).

The key analytic challenge in measuring differences in access and quality by PACE plan type is accounting for the potential differences between the patient populations at the for-profit and not-for-profit PACE plans included in the study. This concern is somewhat mitigated by the fact that we are comparing PACE enrollees with other PACE enrollees. By selecting sites located in areas with similar demographic characteristics, we take the first step in accounting for possible differences in the broader local populations. The matching process described in Section B is the next step in ensuring that the for-profit patient population is similar to the comparison group along as many health and demographic dimensions as possible. Finally, by controlling for population and site characteristics in the multivariate analysis, we control for observable differences between the sites and the patients.

b. Plans for Publication

Mathematica will deliver a report, compliant with requirements of Section 508, that describes the study design and summarizes the differences between access to and quality of care for enrollees in the for-profit and not-for-profit PACE plans in Pennsylvania. This report will include the findings in both the descriptive and multivariate analyses. A draft report will be submitted to CMS in August 2013. The final report will be delivered to CMS in September 2013. Mathematica will also present the findings in a briefing to CMS at the conclusion of the contract. All reporting of results will clearly note the limitations associated with having only a small number of for-profit plans with their common ownership. The final report will be disseminated through the Centers for Medicare & Medicaid (CMS) Web site. This report will also be posted on the Mathematica’s website. Manuscripts on the study findings may be submitted for publication in leading health policy journals.

c. Project Time Schedule

The survey will be fielded from November 2012 through March 2013. The draft report will be submitted in August 2013 and the final report will be delivered in September 2013.

17. Expiration Date

We will display the expiration date of the OMB approval in all notification letters sent to respondents and eligible proxies soliciting their cooperation. As this is a telephone survey, the expiration date of the OMB approval will not need to be included on surveys.

www.mathematica-mpr.com

Improving public well-being by conducting high-quality, objective research and surveys

Princeton, NJ ■ Ann Arbor, MI ■ Cambridge, MA ■ Chicago, IL ■ Oakland, CA ■ Washington, DC

Mathematica^® is a registered trademark of Mathematica Policy Research