The Colorado Department of Health Care Policy and Financing (HCPF) administers the Medicaid and Child Health Plan Plus (CHP+) programs as well as a variety of other health care programs for Colorado’s low-income families, the elderly and persons with disabilities. The total population of clients included 283,687 Medicaid clients and the survey population included 118,839 Adults and Elderly, Blind and Disabled individuals. On April 12, 2011, the HCPF issued a Documented Quote to solicit proposals for the Health Outcome Measure Initiative Survey. The solicitation was on closed on April 26, 2011 and the P.O. was signed on May 16, 2011 authorizing Corybant to administer health outcome surveys, analyze the responses and report the findings to the Department before June 24, 2011.

Using Corybant's IVEENA® Service Delivery Platform and Survey System, the survey was conducted in seven (7) calendar days, and the total duration of the project was thirty five (35) calendar days from contract award to final reports. Corybant configured and deployed IVEENA® to implement and host an inbound Interactive Voice Response (IVR) for the SF-12v2® survey instrument as well as the required information management support services.

The total population of clients included 283,687 Medicaid clients and the survey population included 118,839 Adults and Elderly, Blind and Disabled individuals. A personalized invitation by HCPF was mailed to the survey population in three batches on June 6th, June 7th and June 8th. Each letter included a survey access code and instructions for calling the toll-free number Corybant had established for the survey. As an incentive to increase participation, those participants who completed a survey were entered into a drawing for a grocery gift card and instructed that the drawing would take place on or around June 30, 2011.

The survey was conducted via telephone using an automated IVR implementation of the SF-12v2® survey instrument. The project goal was to collect data on the physical and mental health of beneficiaries in two Medicaid populations in the state of Colorado: low income Adult, and the Elderly, Blind and Disabled. The survey was conducted automatically by phone and included a sampling of each population. Each completed survey was scored with each score consisting of a physical and mental health component.

Survey participation was treated anonymously and each participant caller was required to log into the survey using the access code number that was provided in the invitation letter. Callers were presented with a branded greeting for HCPF and detailed instructions for taking the survey, as well as an option to hear frequently asked questions about the survey. The survey was available in English and Spanish, 24 hours a day via an automated telephone system. During the survey, callers were asked to confirm each response to the survey questions. Including the greeting, detailed survey instructions, and confirmation of each response, the average call length for a completed survey was approximately 9 minutes. A total of 4,161 surveys were collected in 7 days.

Corybant anonymously matched the survey response to the Gender, Language, Age Group, Eligibility Type, county, and general category of Adult with Medicaid versus Elderly, Blind and Disabled populations. Corybant performed the scoring of the surveys using the automated scoring service that was licensed to HCPF by the provider of the SF-12v2® survey instrument.

SURVEY METHODOLOGY

Project Responsibilities

HCPF was responsible for contacting Medicaid clients by mail requesting their participation in the survey, and providing instructions for participating in the survey. Instructions included a randomly generated number to identify the client, which the client entered into Corybant's inbound IVR system upon calling. HCPF was responsible for sending letters to Medicaid clients and for providing Corybant with the SF-12v2® survey questionnaire and scoring mechanism to the Contractor. Participants who completed a survey were entered into a drawing for a grocery gift certificate, the drawing to be held on or about June 30, 2011.

Corybant was responsible for administering surveys to Colorado Medicaid clients to determine the health status of the clients and to obtain outcomes of the health care services provided or administered to the clients using the Short Form 12, version 2 (SF-12v2®) survey and scoring mechanism that are licensed to the HCPF.

Corybant provided a toll free number for participants to call 24-hours a day. The survey was available in English or Spanish with high quality voice recordings and Corybant worked with HCPF to create branded dialog scripts. The system was configured to accept both DTMF (key pad entry) and voice commands. The IVR program included sensitivity settings and error detection capabilities to ensure unambiguous responses. Survey responses were captured and posted immediately. Results were correlated to the Gender, Language, Age Group, Eligibility Type, County, and general category of Adult with Medicaid and Elderly, Blind and Disabled populations. The following paragraphs include additional detail about the survey execution and reporting analysis.

Project execution

The project was conducted over a period of 28 calendar days from contract start to end of data collection. This project was executed within Corybant’s formal project management framework that included creation of formal communication plan, regularly scheduled review meetings, and managed, informal dialog with HCPF team. During this period Corybant implemented the survey instrument within the IVR modality, reviewed this instrument with the SF-12v2® authors and gained authorization to use this instrument from HCPF team.

As a part of its goal to support HCPF Corybant undertook the additional task, initially planned to be conducted by HCPF, of developing the methodology for, and processing of, the target survey participants. This effort included assigning unique survey numbers, and creating the sample set for the project that was delivered to HCPF for printing and mail merge.

An automatic interface for scoring the surveys using the web service available from the SF-12v2® authors was implemented and certified.

Sample selection

This survey was conducted via telephone using an IVR call processing system that guided callers through the SF-12v2® survey. The SF-12v2® survey is a 12-question subset of the SF-36v2® survey that measures eight domains of health. It is a brief, measure of overall health status, used in large population health surveys.

The sample set is a random selection of 118,839 records from 283,687 Medicaid client records provided by HCPF, which included Adult and EBD records.

Two data sets containing a total of 283,687 records were provided to Corybant. One dataset included 19,041 records for Elderly, Blind, and Disabled and a second dataset included 264,646 records from adult Medicaid beneficiaries. Corybant used these records to create the survey sample set that included 100% of the EBD records and approximately 100,000 records from the second data set, while maintaining the distributions for benefit groups, gender, and language. Corybant assigned unique identification number to each record to be used as a login credential (pass code) for each client taking the survey. Corybant delivered the working sample set containing the selected records and the survey codes to HCPF.

Notification

In a mail merge process, HCPF merged the identification codes into personalized letters that described the objective of the survey, instructions for taking the survey and the incentive program. These letters were printed and mailed in 3 bulk mailings over 3 days by the State of Colorado printing facilities. The instructions in the letter included the start and end date/times for the survey, the toll-free phone number to call, and instructions for entering the pass code. Each letter was printed in English on one side with a Spanish translation on the opposite side. Letters were sent in both English and Spanish. Participants were also informed that the survey was limited to the first 5,000 participants and those participants who completed a survey would be entered into a drawing for a grocery gift card, the drawing to take place on or about June 30, 2011. The bulk mailing of letters began on July 6, 2011 and concluded on June 8, 2011. The IVR Survey system was activated on the morning of June 8, 2011 and terminated at midnight on June 14, 2011.

Data collection and Scoring

The IVR system validated each client that called using the pass code, and collected client responses for each of the 12 questions in the SF-12v2® survey. Client responses were logged in a database. Anonymous survey results for each participating client were scored and correlated to categories including Gender, Language, Age Group, Eligibility Type, County, and Adult or EBD populations. The resulting scores were recorded in the database. This report includes a summary of the data collected with consolidated scores for the Physical (PCS) and Mental (MCS) components of the survey. The raw data was delivered to HCPF in machine-readable form on Thursday June 17, 2011 -- less than 35 days after the contract was awarded.

After final reveiw, HCPF submitted a formal report prepared by Corybant to the US Center for Medicare and Medicaid.