General Background Information on the Survey of Enrollees

I. General Background Information on the Survey

The VHA Survey of Veteran Enrollees Health and Reliance Upon VA (Survey of Enrollees), developed in 1999, facilitates a better understanding of enrolled Veterans’ health care needs. In addition to collecting basic demographic information, the survey explores insurance coverage, VA and non-VA health care use, pharmaceutical use, attitudes and perceptions about VHA services, perceived health status, and smoking habits of Veterans enrolled in the VA health care system. All of these factors are considered predictive indicators of an enrollee’s future use of VHA services and the information is not available in VHA administrative databases. The survey is conducted nationwide via telephone under the auspices of the Office of the Assistant Deputy Under Secretary for Health For Policy and Planning (ADUSH/PP), with the primary purpose of informing the Enrollee Health Care Projection Model (EHCPM). In addition to supporting the EHCPM, data from the survey are used for planning purposes and policy related analyses by several offices within VA and VHA.

II. Overview of Survey Methodology

The 2011 survey was conducted from March 4, 2011 through May 27, 2011, with 43,633 completed interviews. The average interview length was 15.6 minutes. The number of contacted eligible respondents was 57,820. The overall cooperation rate was 75 percent. This was a marked improvement from previous years, which have hovered around 60 percent for the last two years after dipping down to 50 percent in 2007 from 73 percent in 2005. The cooperation rate is defined by the American Association for Public Opinion Research (AAPOR) as the proportion of completed interviews in the number of contacted eligible respondents.

Part of the drop in 2007 may have been due to Veterans’ concerns after a laptop containing files with Veterans’ personal data was stolen in 2006. These files were not related to the Survey of Enrollees. The improvement in cooperation rates may have been influenced by increased communication efforts to assure that all key offices in VA and that the Veterans Service Organizations were aware when the survey was fielded. In addition, the use of pre-notification letters, which were sent to each Veteran in the sample beginning in 2008, likely had a positive impact on cooperation rates. These letters included Frequently Asked Questions with a toll-free number for both VA for questions Veterans may have, and ICF International if the enrollee was ready to start the interview. The toll-free number was also left on answering machines after three attempts to reach the Veteran.

2011 Survey of Enrollees Key Methodology Data
Enrollees Represented	7,895,108
Weighted Enrolled Population as of	09/30/2012
Eligible Contacts	57,820
Completed Interviews	43,633
AAPOR Cooperation Rate	75%
Survey Interview Timeframe	March – May 2011

Once a Veteran is contacted and agrees to complete the survey, the interview is seen through until completion, even if the target is met. Therefore, 1,633 more Veterans than the 42,000 targeted were surveyed.

The 2011 Survey of Enrollees was an English-only, Computer-Assisted Telephone Interviewing (CATI) survey using a stratified design of enrollees to obtain a target of 42,000 completed interviews. ICF International, provided technical and data collection services for the survey.

Interviewed Veterans belonged to a stratified sample of 420,011 Veterans selected from the VHA Enrollment File. Enrollees were excluded from the sample if either of VA’s key Veteran data files, the Health Eligibility Center (HEC) file or the VA Vital Status File, contained a Date of Death. From this population, 298 strata based on VISN (21), enrollee type (2: pre or post enrollees), priority group (7: 1-6, 7/8) and OEF/OIF/OND (2: Yes/No) were derived. The OEF/OIF/OND stratum was added in 2008 because analysis in 2005 and 2007 showed that this cohort of Veterans was under-represented among respondents. Veterans of these conflicts have been oversampled in the 2008, 2010, and 2011 surveys. Pre-enrollees are defined as those Veterans who used the VA Health Care System during fiscal years 1996, 1997, or 1998 or enrolled during the first six months of enrollment, October 1, 1998 to March 31, 1999. All other enrolled Veterans are considered Post-enrollees. The sample did not include Veterans living outside the United States or Puerto Rico or those Veterans for whom VA did not have a valid address.

In general, the characteristics of people who do not reply to a survey tend to differ from those who do participate in the survey process. Therefore, responses were adjusted statistically to account for what is referred to as “non-response bias” and is associated with non-participation. This weighting adjustment was primarily based on demographics and health care utilization information from administrative records.

Methodological Experiments: From previous Notice of Actions, Term of Clearance for the Survey of Enrollees encourages VA to conduct a series of methodological experiments to continuously improve the conclusions drawn from the Survey of Enrollees. These experiments include the use of Pre-Notification Letters, additional call attempts, and attempts to contact Veterans without telephone numbers in the VHA records. Actions taken based on these experiments are:

In 2007 a small sample of enrollees received a pre-notification letter before the survey. Based on improved response rates as a result of this experiment, beginning in 2008, all enrollees in the selected sample received a pre-notification letter.

The number of call attempts was modified in an effort to increase response rates. The increase of call attempts from six to ten resulted in a less than 2 percent increase in the response rate in 2007. Therefore, beginning in 2008, VHA agreed to seven call attempts in an effort to increase response rates, but not overly burden Veterans.

In 2010, a reverse look-up service provided valid numbers for 5,731 of the 6,870 enrollees without valid telephone numbers. This subsequently resulted in the completion of 870 additional interviews. In 2011, the organization providing the reverse look-up service was unable to meet the VA security requirements. Therefore, no reverse look-ups have been performed since then. Instead, 15,339 enrollees without valid telephone numbers received a modified pre-notification letter explaining that while VA did not have a valid telephone number, the enrollee’s view was important to VA, and the enrollee was encouraged to call in to a provided contact number. This resulted in 244 (1.5%) additional completed interviews. Although, this did not yield the response rate that the reverse look-up service did in 2010 (15%), it was a low-cost, but useful method for reaching enrollees without valid telephone numbers.

III. Highlights from the 2011 SoE data collection

Results of the survey are weighted to represent the population of 7,895,108 Veteran enrollees at the time of the survey weighting (September 30, 2010). Information includes socioeconomic characteristics of the enrollee population, public and private insurance coverage, health status measures, and future use of VA health care services by the Veteran enrollee population.

Demographic and Socioeconomic Characteristics of the Enrollee Population: The average age of enrollees was 62 years. The majority of enrollees were male, married, and white; However, Ethnicity and Race identifications tended to be more diverse among younger enrollee age groups. Women represented 6 percent of the total enrollee population and 12 percent of enrollees with OEF/OIF/OND status. The median reported annual household income for all enrollees remained steady at $35,000. Most enrollees (60%) were not in the labor force; the unemployment rate for those who were in the labor force rose was 22 percent in 2011. The largest percentage of enrollees (41%) served during the Vietnam era.

Health Status , Key Drivers of Enrollees ’ Health Care Decision Making, and Planned Future Use of VHA Health Care Services: Overall, 65 percent of enrollees considered their health status to be “Excellent/Very Good/Good.” Enrollees in the 45 – 64 age group have consistently rated their perceived health status lower than either the less than 45 age group or the 65+ age group in the years that the survey has been conducted. Perhaps because they tend to be younger, OEF/OIF/OND enrollees were more likely to report a favorable health status (77%) than all other enrollees (64%). Most enrollees (75%) reported that they use VA for at least some of their health care needs. Seventeen questions that in previous years have proved to be significant predictors of VHA utilization (“key drivers”) were included in the survey; the highest level of agreement (i.e., respondents “agree” or “completely agree”) was found for questions related to “Quality,” followed by “Cost,” and then by “Availability and Accessibility of Services.” Overall, the “Quality” statement “VHA health care providers treat their patients with respect” received the highest level of concurrence (84%). Users of VA care are more likely to concur with “key driver” statements than enrollees who do not use VA care. Enrollees with OEF/OIF/OND status tended to concur less with “key driver” statements than did their counterparts. The most common response to how enrollees planned to use VA services was for primary care (48%).

Public and Private Health Insurance Coverage: Most enrollees (77%) reported some type of public or private health insurance coverage in addition to their VA health care benefits. Uninsurance rates decrease as age or as annual household income increases. Female enrollees had a higher uninsurance rate (30%) than male enrollees (23%). Enrollees with Medicare represent 51 percent of the total population.

Pharmaceutical Use: Although 77 percent of enrollees reported some form of non-VA health care insurance, only 39 percent reported some form of non-VA prescription drug coverage. Forty percent of all enrollees do not consider or do not know that their VA enrollment provided pharmaceutical benefits. Thirty-four percent of enrollees who reported taking any prescription over a 30 day period did not use VA to fill those prescriptions. When combined with “Don’t Know” or “Refused” responses, fewer than 40 percent of enrollees do not consider that their VA enrollment provides drug coverage. Enrollees under the age of 45 are less likely to be aware of their VA Drug coverage (44%) than those age 45-64 (63%) or 65 and older (63%). Thirty-six percent of enrollees with Medicare have opted to purchase Medicare Part D (prescription coverage); the VHA Office of the ADUSH/P&P has begun a study to better understand why enrollees would choose Medicare Part D over VA prescription coverage.

Reliance: The 2011 survey asked enrollees two questions on their overall use of medical services in 2010. Enrollees were first asked to recall non-VA outpatient visits or trips within either the year 2010 or the last 3 months of 2010, and then asked to recall VA outpatient visits or trips within the same time period, excluding dental, mental health, substance abuse, and/or trips to a pharmacy. The dependence upon enrollees’ recall decreases the overall reliability of the data when compared to actual VA healthcare utilization. However, the ability to profile VA reliance based on an enrollee perspective is worthwhile in analyzing actual versus perceived reliance in further studies.

VA reliance is setting-specific (e.g., outpatient) and is defined as the number of visits or trips in a VA setting reported by an enrollee divided by the sum of all visits in both VA and non-VA settings. For example, if an enrollee’s outpatient care was provided entirely by VA, then his/her reported outpatient VA reliance is one (1.0), meaning 100 percent of the enrollee’s self-reported outpatient care was provided by VA. If an enrollee’s outpatient care was provided entirely outside VA, the reported outpatient VA reliance is zero (0.0); meaning VA provided 0 percent of the enrollee’s self-reported outpatient care. If an enrollee reported no visits or trips at all, neither VA nor non-VA, then his/her outpatient reliance is undefined.

Thus, only enrollees who utilize some outpatient care in either a VA and/or non-VA setting have a defined reliance factor for that setting. Therefore, the data is reported as percentages (i.e., the percentage of outpatient visits provided by VA within the reference months/year).

Nationwide, the average reliance on VHA for outpatient care was 47 percent. A little more than half of enrollees who use health care are either not at all reliant (27%) or 100 percent reliant (25%). Economic factors impacted reliance with enrollees who were unemployed, uninsured, and/or earning less than $20,000 as they were more likely to rely on VHA for their health care. Enrollees less than 30 years of age and enrollees from 50 to 64 years of age had higher than average VA reliance (54% and 56% respectively). Enrollees who reported being married (40%) or widowed (46%) had lower VA reliance than enrollees who were divorced (62%), separated (63%), or single (62%). Enrollees who identified themselves as Asian were least reliant (42%), while enrollees who identified as Black or African-American were most reliant (57%) on VA for their health care.

Cigarette Smoking Status: Current smokers (1.6 million) made up 20 percent of the entire enrollee population, only slightly higher than the Center for Disease Control (CDC) estimates of 19.3 percent of U.S. adults who are current smokers. A large percentage of enrollees (68%) reported being a smoker at one time; of these, 71 percent are former smokers. Of the current smoker population, 27 percent reported combat exposure, 20 percent were female, and 35 percent were unemployed.

IV. Proposed Changes to the 2012 SoE

Historically, VHA has utilized the services of contractors to conduct telephone surveys. VHA provides a stratified random sample and phone numbers to the contractor. The telephone survey is administered using Computer Assisted Telephone Information (CATI) with responses entered directly into an electronic database, making the collection of data very efficient and reliable. This reduction of respondent burden through reliance on CATI meets the spirit of the Government Paperwork Elimination Act (GPEA). However, advances in technology and changes in how people communicate have challenged telephone and uni-mode response bias. The 2011 Survey of Enrollees sample included Veterans without telephone numbers, however, these Veterans were asked to respond to a tailored letter and call the contractor to complete the survey. Survey response from these Veterans was low. To capture information from these Veterans and, therefore, to reduce survey response bias, VHA introduced a methodological experiment into the 2012 fielding that offers a small number of enrolled Veterans the opportunity to participate via mail survey or web-based survey. VA introduced this experiment at the suggestion of previous discussions with OMB, as well as OMB’s suggestion to speak with the Project Manager for the 2010 Survey of Veterans.

VA applied all relevant regulations and statutes regarding privacy and security to these modalities. This experiment will allow VA to compare survey response rates and completion rates across modality. Based on the results of this methodological experiment (i.e., should response rates markedly increase), VA may re-visit survey administration techniques. Successful implementation of a multi-mode survey will also respond to the desire expressed by many Veterans that they receive the survey either through mail or to have the opportunity to complete it online.

The survey questions for the 2012 survey are basically the same as those asked in 2011 with the addition of the Long Term Care Question Module, similar to questions appearing on the 2004 National Long Term Care Questionnaire. These new questions will assist in an effort to enhance the institutional full demand component of the Long Term Care Projection Model. Currently, the full demand projection is based on a subset of the VA-enrolled population. For short stay services, this is the portion of the population that is enrolled in standard Medicare. For long stay services, it is the portion of the population enrolled in Medicaid and present on the Medicaid match data set. The results for these population subsets are then extrapolated to the entire VA-enrolled population to approximate full demand.

The survey questions will enhance the full demand projection in two ways. First, by combining the survey data with enrollment and workload data, the results can be used to validate the assumptions regarding the Medicare/Medicaid sub-populations and their relevance to the entire VA population. Secondly, it will allow VA to understand the prevalence of non-governmental payment sources for long term care services, such as self-pay and private long term care insurance policies. Currently, VA has no data regarding the prevalence of these non-governmental payment sources.

This module was added to the survey without significant changes, the updated form substitutes the previously cleared form for data collection. All of the revisions, updates, and results will be reported during the renewal process. The request for this module was submitted as a revision to a previously cleared collection.

V. Primary Uses of the Survey Data

VHA relies on annual projections of enrollment, utilization, and expenditures from the Enrollee Health Care Projection Model (EHCPM) and the EHCPM is based on information obtained from the Survey of Enrollees. In addition to the enrollment decision, EHCPM projections derived from the Survey of Enrollees, support approximately 95 percent of VHA’s medical care budget estimates every budget cycle. Information collected by the Survey of Enrollees also supports VHA’s preparation for the impact of the Patient Protection and Affordable Care Act (ACA) of 2010. While the 8.2 million individuals who are currently enrolled in VA will be considered to have the individual coverage mandate in accordance with the ACA, most Veterans will have new choices for health care coverage once the act is fully implemented. In addition, a Veteran’s family’s new health care options may further impact their health care utilization choices. In order to appropriately serve Veterans, it is critical that VA is proactive in its preparation for any VHA utilization shifts resulting from ACA.

Information obtained from the survey also helps VHA to understand Veteran family composition and caregiver characteristics. This enables VA to meet the requirements of the Caregivers and Veterans Omnibus Health Services Act of 2010. In implementing this legislation, VA has initiated many new programs and policies aimed at eligible post-9/11 Veterans and their families who elect to receive their care in a home setting from a primary family caregiver. While these are the prime examples of the survey uses in policy and budget decision making, the information gathered from the Survey of Enrollees increasingly is used as a foundation for VA policy and planning. Making sure VHA is able to provide the “right care at the right time” is dependent on a reliable understanding of the Veterans that we serve. Data from the survey is used for planning purposes and policy related analyses by several offices within VA and VHA.

The information that is available in VA administrative databases does not reflect all enrollees’ health status, income, or their reliance upon the VA system; all elements which are critical to making valid projections of demand, supporting the Secretary’s enrollment level decision, and making Veteran-centric policy. Basic demographic and socioeconomic data in VA administrative files are generally not complete or valid for all segments of the VHA enrollee population to be of sufficient utility for enrollment based policy and budget development.

The Veteran population is becoming increasingly diverse with wide ranging needs and expectations for a health care provider. Rapid changes going on in the world around them, such as troop draw downs from recent conflicts, changes to the national health care landscape, adjustments to Medicare/Medicaid, the enactment of the ACA, and shifts in the state of our nation’s economy, will affect Veterans’ opinions about whether or not VA is best able to meet those needs and expectations. The need for real time information on “hot topics” or specific areas of interest necessitates the annual Survey of Enrollees to capture this critical information for input into data-driven policy and budgetary analyses.