ORIGINAL ARTICLE
Rural Women Veterans Demographic Report: Defining VA Users’ Health and Health Care Access in Rural Areas Elizabeth Brooks, PhD;1 Nancy Dailey, MSN, RN-BC;2 Byron Bair, MD;2 & Jay Shore, MD, MPH2,3 1 Department of Public Health, University of Colorado Anschutz Medical Campus, Aurora, Colorado 2 Veterans Rural Health Resource Center–Western Region, Veterans Administration–Office of Rural Health, Salt Lake City, Utah 3 Department of Psychiatry, University of Colorado Anschutz Medical Campus, Aurora, Colorado
Abstract This study was funded by the Veterans Rural Health Resource Center–Western Region, Salt Lake City, Utah–Field Office for the Office of Rural Health. The authors’ opinions expressed do not necessarily reflect those of the Veterans Administration. For further information, contact: Elizabeth Brooks, PhD, Department of Public Health, University of Colorado, Anschutz Medical Campus, Mail Stop F800, Nighthorse Native Campbell Building, 13055 E. 17th Ave., Aurora, CO 80045; e-mail: [email protected]. doi: 10.1111/jrh.12037
Purpose: While many women choose to live in rural areas after retiring from active military duty, a paucity of studies examine rural women veterans’ health care needs. This report is the first of its kind to describe the population demographics and health care utilization of rural female veteran patients enrolled in the Department of Veterans Affairs (VA). Methods: Using the National Patient Care Datasets (n = 327,785), we ran adjusted regression analyses to examine service utilization between (1) urban and rural and (2) urban and highly rural women veterans. Findings: Rural and highly rural women veterans were older and more likely to be married than their urban counterparts. Diagnostic rates were generally similar between groups for several mental health disorders, hypertension, and diabetes, with the exception of nonposttraumatic stress anxiety that was significantly lower for highly rural women veterans. Rural and highly rural women veterans were less likely to present to the VA for women’s specific care than urban women veterans; highly rural women veterans were less likely to present for mental health care compared to urban women veterans. Among the users of primary care, mental health, women’s specific, and all outpatient services, patients’ annual utilization rates were similar. Conclusions: Improved service options for women’s specific care and mental health visits may help rural women veterans access care. Telehealth technologies and increased outreach, perhaps peer-based, should be considered. Other recommendations for VA policy and planning include increasing caregiver support options, providing consistency for mental health services, and revising medical encounter coding procedures. Key words access to care, demography, geography, health services research, utilization of health services.
Rural residents are overrepresented among military recruits1 and account for roughly 41% of the veteran population.2 Health care issues for this population are an important consideration. Rural veterans use fewer health care services and frequently travel long distances to receive care.3,4 While rural and urban groups share many of the same demographic characteristics, remotely based veterans have lower mental health-related quality-of-life scores and experience a higher prevalence of physical
146
illness.3,5 Moreover, rurality is a demonstrated risk factor for suicide among veteran patients, even after controlling for mental health availability.6 Although many rural residents need specialty care, such as psychiatric evaluations and gynecological services, access to such appointments is difficult to obtain in remote areas given the lack of local providers.7,8 As the number of women joining the United States military rises,9 more women veterans are returning to
c 2013 National Rural Health Association The Journal of Rural Health 30 (2014) 146–152
Rural Women Veterans
Brooks et al.
rural communities upon completion of active duty. Yet, very little research has been conducted to understand the health needs and utilization of women veterans living in these areas. This is a critical oversight. Documenting the demographic profile and service use patterns of various patient populations is an important step in understanding and improving health care. Prior studies show that personal factors such as patient gender or geographic location are associated with one’s use of health care services.10-14 Likewise, patient characteristics such as age, comorbidity, and diagnostic severity can influence individual health care needs.15-19 This type of information is critical for informing and developing health care planning for large groups or specific services. When group characteristics are not recognized or well-understood, some cohorts may receive different, inappropriate, or even substandard care, which may ultimately result in health disparities between groups. With regard to geography, it is unclear how personal characteristics, health care needs, and service utilization differs between urban and rural veteran women. Recently, leaders at the 2010 VA Women’s Health Services Research Conference emphasized the lack of data about rural women veterans and designated rural women veterans’ health care as 1 of 6 priority areas needed to further the VA Women’s Health Research Agenda.20 In response to the critical need for information, members of the Veterans Rural Health Resource Center– Western Region (under the VA Office of Rural Health) began a new effort in October 2011 to identify the health care needs and experiences of VA-enrolled rural women veterans. The specific aim was to describe the population demographics and health care utilization among rural female veteran patients. Data were gathered by analyzing a large database of patient medical records for outpatientbased visits. To our knowledge, this effort is the first of its kind to document detailed demographic information and service gaps among the rural women veteran population. The results may be used to inform unique planning efforts and outreach considerations for this particular group.
Methods Database Review Dataset, File Build, and Missing Data. The information for this report utilizes data from VA National Patient Care Datasets (NPCD) and the VSSC Operation Enduring Freedom/Operation Iraqi Freedom (OEF/OIF) Roster File. The NPCD contains background and medical visit information for patients seen at VA health care clinics, such as demographic data, diagnoses, and service dates. It does not contain visit information about con-
c 2013 National Rural Health Association The Journal of Rural Health 30 (2014) 146–152
tracted care services. Research has shown the NPCD to be a valuable resource for health services exploration.21 The VSSC OEF/OIF Roster File identifies women whose period of service includes a tour of duty in either Iraq or Afghanistan. For analytic purposes, we matched together the NPCD and VSSC Roster File using a scrambled social security number common to both datasets. We extracted information for women veterans who visited the VA during Fiscal Year 2011 (October 1, 2010September 30, 2011). Our primary interest was to compare differences between urban and rural patients. While there are many definitions of rural, we used the classification system originated by the VA Office of Policy and Planning based on our assumption that the report findings would be most applicable to the forecasting and development needs of the VA. Accordingly, 3 broad categories of geographic populations were assigned: urban, rural, and highly rural. Urban areas included any enrollee located in a Census-defined urbanized area; rural areas included any enrollee who was not classified as urban; and highly rural areas were rural areas with counties having fewer than 7 people per square mile. Using the VA ZIP Code Crosswalk developed by West,22 we matched patients’ home ZIP codes to each geographic designation. We dropped patient data for 1,354 cases whose ZIP codes did not have a match in the Crosswalk (less than 1% of the data, most from US territories). The NPCD contains health information for both veterans and nonveteran patients. Prior research shows that nonveteran data accounts for about 50% of female users but only 3.0% of male users.23 Therefore, we removed nonusers (n = 187,096) from the dataset using a patient eligibility code variable. After these exclusions, the final dataset contained 327,785 unique patients.
Model Variables The NPCD contains information about patient encounters for each visit called “clinic stops.” We examined clinic stops using all of the available data to account for women who access multiple services in one visit. This may be particularly relevant for rural patients who often consolidate visits to reduce travel. We excluded duplicate clinic stop codes occurring on the same day at the same location. We identified primary care and mental health encounters according to the clinic stop identifiers used in the Sourcebook: Women Veterans in the Veterans Health Administration.24 Primary care visits included services such as, but not limited to, general, geriatric, and nursing care. Mental health visits included services such as individual counseling, posttraumatic stress disorder (PTSD) group therapy, and social work. Women’s specific encounters were identified by the authors using clinic stops that clearly pertained
147
Rural Women Veterans
to care that only (or mostly) women veterans would obtain (eg, mammograms, pap smears, and female genderspecific cancer screenings). We included active duty and nonactive duty sexual trauma in this category, although we realize that some male veterans also access this type of clinical care. The examination of women’s specific services should be used only as a guide because facilities often vary in their coding for such visits. Moreover, some types of women’s specific procedures are available only on a contractual basis with external providers, and therefore, not included in the NCPD datasets. Having any visit, as well as the total number of visits, for each service type (all outpatient, primary care, mental health care, and women’s specific care) served as our dependent variables for the regression analyses (see below). Throughout this report, patient visits represent “service use” in our discussion. We included the following covariates in our examination: rural category (urban, rural, and highly rural), age category (18-44, 45-64, 65, or older), marital status (married/unmarried), service era (OEF/OIF and “other”), percent of service-connected disability status (0%-49%, 50%-99%, and 100%), and military sexual trauma (MST) history (yes, no). We also examined the presence of several mental and physical health diagnoses (addiction disorders [includes alcohol and/or substance abuse], PTSD, other anxiety disorders, mood disorders, eating disorders, hypertension, and diabetes). As indicated above, we examined service-connected disability status categorically. Disability ratings indicate an injury or illness that was incurred or was aggravated while serving, and compensation benefits are wholly or partly based on this score.
Analysis We compared differences in patient background characteristics and diagnoses between geographic cohorts: (1) urban versus rural and (2) urban versus highly rural. Logistic regression examined differences in the proportion of urban, rural, or highly rural women veterans seeking outpatient visits for primary care, mental health care, and women’s specific services. Each examination was separated by the type of outpatient care obtained. Among users of each service type, we employed negative binomial regression to examine differences between geographic cohorts, with urban patients serving as the reference group. The regression analysis adjusted for differences in patients’ demographic characteristics. Due to the large number of cases, nearly every comparison resulted in statistical significance using a P value
Rural Women Veterans Demographic Report: Defining VA Users’ Health and Health Care Access in Rural Areas Elizabeth Brooks, PhD;1 Nancy Dailey, MSN, RN-BC;2 Byron Bair, MD;2 & Jay Shore, MD, MPH2,3 1 Department of Public Health, University of Colorado Anschutz Medical Campus, Aurora, Colorado 2 Veterans Rural Health Resource Center–Western Region, Veterans Administration–Office of Rural Health, Salt Lake City, Utah 3 Department of Psychiatry, University of Colorado Anschutz Medical Campus, Aurora, Colorado
Abstract This study was funded by the Veterans Rural Health Resource Center–Western Region, Salt Lake City, Utah–Field Office for the Office of Rural Health. The authors’ opinions expressed do not necessarily reflect those of the Veterans Administration. For further information, contact: Elizabeth Brooks, PhD, Department of Public Health, University of Colorado, Anschutz Medical Campus, Mail Stop F800, Nighthorse Native Campbell Building, 13055 E. 17th Ave., Aurora, CO 80045; e-mail: [email protected]. doi: 10.1111/jrh.12037
Purpose: While many women choose to live in rural areas after retiring from active military duty, a paucity of studies examine rural women veterans’ health care needs. This report is the first of its kind to describe the population demographics and health care utilization of rural female veteran patients enrolled in the Department of Veterans Affairs (VA). Methods: Using the National Patient Care Datasets (n = 327,785), we ran adjusted regression analyses to examine service utilization between (1) urban and rural and (2) urban and highly rural women veterans. Findings: Rural and highly rural women veterans were older and more likely to be married than their urban counterparts. Diagnostic rates were generally similar between groups for several mental health disorders, hypertension, and diabetes, with the exception of nonposttraumatic stress anxiety that was significantly lower for highly rural women veterans. Rural and highly rural women veterans were less likely to present to the VA for women’s specific care than urban women veterans; highly rural women veterans were less likely to present for mental health care compared to urban women veterans. Among the users of primary care, mental health, women’s specific, and all outpatient services, patients’ annual utilization rates were similar. Conclusions: Improved service options for women’s specific care and mental health visits may help rural women veterans access care. Telehealth technologies and increased outreach, perhaps peer-based, should be considered. Other recommendations for VA policy and planning include increasing caregiver support options, providing consistency for mental health services, and revising medical encounter coding procedures. Key words access to care, demography, geography, health services research, utilization of health services.
Rural residents are overrepresented among military recruits1 and account for roughly 41% of the veteran population.2 Health care issues for this population are an important consideration. Rural veterans use fewer health care services and frequently travel long distances to receive care.3,4 While rural and urban groups share many of the same demographic characteristics, remotely based veterans have lower mental health-related quality-of-life scores and experience a higher prevalence of physical
146
illness.3,5 Moreover, rurality is a demonstrated risk factor for suicide among veteran patients, even after controlling for mental health availability.6 Although many rural residents need specialty care, such as psychiatric evaluations and gynecological services, access to such appointments is difficult to obtain in remote areas given the lack of local providers.7,8 As the number of women joining the United States military rises,9 more women veterans are returning to
c 2013 National Rural Health Association The Journal of Rural Health 30 (2014) 146–152
Rural Women Veterans
Brooks et al.
rural communities upon completion of active duty. Yet, very little research has been conducted to understand the health needs and utilization of women veterans living in these areas. This is a critical oversight. Documenting the demographic profile and service use patterns of various patient populations is an important step in understanding and improving health care. Prior studies show that personal factors such as patient gender or geographic location are associated with one’s use of health care services.10-14 Likewise, patient characteristics such as age, comorbidity, and diagnostic severity can influence individual health care needs.15-19 This type of information is critical for informing and developing health care planning for large groups or specific services. When group characteristics are not recognized or well-understood, some cohorts may receive different, inappropriate, or even substandard care, which may ultimately result in health disparities between groups. With regard to geography, it is unclear how personal characteristics, health care needs, and service utilization differs between urban and rural veteran women. Recently, leaders at the 2010 VA Women’s Health Services Research Conference emphasized the lack of data about rural women veterans and designated rural women veterans’ health care as 1 of 6 priority areas needed to further the VA Women’s Health Research Agenda.20 In response to the critical need for information, members of the Veterans Rural Health Resource Center– Western Region (under the VA Office of Rural Health) began a new effort in October 2011 to identify the health care needs and experiences of VA-enrolled rural women veterans. The specific aim was to describe the population demographics and health care utilization among rural female veteran patients. Data were gathered by analyzing a large database of patient medical records for outpatientbased visits. To our knowledge, this effort is the first of its kind to document detailed demographic information and service gaps among the rural women veteran population. The results may be used to inform unique planning efforts and outreach considerations for this particular group.
Methods Database Review Dataset, File Build, and Missing Data. The information for this report utilizes data from VA National Patient Care Datasets (NPCD) and the VSSC Operation Enduring Freedom/Operation Iraqi Freedom (OEF/OIF) Roster File. The NPCD contains background and medical visit information for patients seen at VA health care clinics, such as demographic data, diagnoses, and service dates. It does not contain visit information about con-
c 2013 National Rural Health Association The Journal of Rural Health 30 (2014) 146–152
tracted care services. Research has shown the NPCD to be a valuable resource for health services exploration.21 The VSSC OEF/OIF Roster File identifies women whose period of service includes a tour of duty in either Iraq or Afghanistan. For analytic purposes, we matched together the NPCD and VSSC Roster File using a scrambled social security number common to both datasets. We extracted information for women veterans who visited the VA during Fiscal Year 2011 (October 1, 2010September 30, 2011). Our primary interest was to compare differences between urban and rural patients. While there are many definitions of rural, we used the classification system originated by the VA Office of Policy and Planning based on our assumption that the report findings would be most applicable to the forecasting and development needs of the VA. Accordingly, 3 broad categories of geographic populations were assigned: urban, rural, and highly rural. Urban areas included any enrollee located in a Census-defined urbanized area; rural areas included any enrollee who was not classified as urban; and highly rural areas were rural areas with counties having fewer than 7 people per square mile. Using the VA ZIP Code Crosswalk developed by West,22 we matched patients’ home ZIP codes to each geographic designation. We dropped patient data for 1,354 cases whose ZIP codes did not have a match in the Crosswalk (less than 1% of the data, most from US territories). The NPCD contains health information for both veterans and nonveteran patients. Prior research shows that nonveteran data accounts for about 50% of female users but only 3.0% of male users.23 Therefore, we removed nonusers (n = 187,096) from the dataset using a patient eligibility code variable. After these exclusions, the final dataset contained 327,785 unique patients.
Model Variables The NPCD contains information about patient encounters for each visit called “clinic stops.” We examined clinic stops using all of the available data to account for women who access multiple services in one visit. This may be particularly relevant for rural patients who often consolidate visits to reduce travel. We excluded duplicate clinic stop codes occurring on the same day at the same location. We identified primary care and mental health encounters according to the clinic stop identifiers used in the Sourcebook: Women Veterans in the Veterans Health Administration.24 Primary care visits included services such as, but not limited to, general, geriatric, and nursing care. Mental health visits included services such as individual counseling, posttraumatic stress disorder (PTSD) group therapy, and social work. Women’s specific encounters were identified by the authors using clinic stops that clearly pertained
147
Rural Women Veterans
to care that only (or mostly) women veterans would obtain (eg, mammograms, pap smears, and female genderspecific cancer screenings). We included active duty and nonactive duty sexual trauma in this category, although we realize that some male veterans also access this type of clinical care. The examination of women’s specific services should be used only as a guide because facilities often vary in their coding for such visits. Moreover, some types of women’s specific procedures are available only on a contractual basis with external providers, and therefore, not included in the NCPD datasets. Having any visit, as well as the total number of visits, for each service type (all outpatient, primary care, mental health care, and women’s specific care) served as our dependent variables for the regression analyses (see below). Throughout this report, patient visits represent “service use” in our discussion. We included the following covariates in our examination: rural category (urban, rural, and highly rural), age category (18-44, 45-64, 65, or older), marital status (married/unmarried), service era (OEF/OIF and “other”), percent of service-connected disability status (0%-49%, 50%-99%, and 100%), and military sexual trauma (MST) history (yes, no). We also examined the presence of several mental and physical health diagnoses (addiction disorders [includes alcohol and/or substance abuse], PTSD, other anxiety disorders, mood disorders, eating disorders, hypertension, and diabetes). As indicated above, we examined service-connected disability status categorically. Disability ratings indicate an injury or illness that was incurred or was aggravated while serving, and compensation benefits are wholly or partly based on this score.
Analysis We compared differences in patient background characteristics and diagnoses between geographic cohorts: (1) urban versus rural and (2) urban versus highly rural. Logistic regression examined differences in the proportion of urban, rural, or highly rural women veterans seeking outpatient visits for primary care, mental health care, and women’s specific services. Each examination was separated by the type of outpatient care obtained. Among users of each service type, we employed negative binomial regression to examine differences between geographic cohorts, with urban patients serving as the reference group. The regression analysis adjusted for differences in patients’ demographic characteristics. Due to the large number of cases, nearly every comparison resulted in statistical significance using a P value
