HSR&D Home » Research » CRE 12-038 – HSR&D Study
Impacts of Delivery of Comprehensive Women's Health Care in the VA
Danielle E Rose, PhD MPH
VA Greater Los Angeles Healthcare System, Sepulveda, CA
Funding Period: March 2013 - March 2018
Women Veterans (WVs) are a numerical minority in the VA healthcare system. Their care needs are often made more complex by significant mental health burden, integration of gender-specific services into historically male-only practices, and historical fragmentation of routine women's health services across multiple providers (e.g., primary care providers, gynecologists) and settings within and outside the VA. The VA has therefore faced considerable challenges in delivering comprehensive care to women. In response, VA policymakers have advanced specific care models and services that must be made available to women to ensure equitable access to comprehensive care.
This study was part of the 5-study VA HSR&D-funded Women's Health CREATE initiative to use research to accelerate implementation of comprehensive care for women seeking care in VA healthcare settings. In this component study, we aimed to: (1) assess determinants of variations in the delivery of comprehensive care for women Veterans; and (2) examine the impact of comprehensive care delivery on WVs' quality of and experience with care.
For Aim #1 (assess determinants of variations in delivery of comprehensive care), we developed a multi-module survey approach with four key informants: 1) primary care director (PC); 2) women's health medical director or women's health liaison (WHMD); 3) mental health lead (MH), and; 4) women Veteran program managers (WVPMs). Domains varied across the different survey instruments. The first three modules focused on describing care arrangements, governance, and management challenges in primary care, mental health and women's health, all of which could influence implementation of comprehensive care for women Veterans. The WVPM survey focused on outreach to enrolled and non-enrolled women Veterans, first-contact with VA, and involvement in governance activities (e.g., Women Veterans Health Committee meetings, Primary Care Advisory Board meetings, involvement in developing facility strategic plan). We linked the four surveys to organizational and administrative data (e.g., facility type: VA Medical Center, Healthcare Center or Community Based Outpatient Clinic; workload, e.g., number of unique patients associated with a primary care program) and area-level characteristics (e.g., large metropolitan area v. not; U.S. Census region, primary care healthcare shortage area), to examine prevalence of comprehensive care for women Veterans, as well as multi-level determinants of different levels and types of comprehensive care achievement.
For Aim #2, we linked the combined dataset from Aim #1 with VA quality of care (External Peer Review Program) and patient survey data (Survey of Healthcare Experiences of Patients) to examine the impact of comprehensive care on women Veterans' outcomes. We modeled achievement of performance on quality of care and patient experiences as a function of availability of comprehensive care, controlling for patient, facility, and area-level characteristics.
Population: The sample included key informants at all sites serving 300 or more women Veterans (primary care 326 sites; mental health 267; WVPM (healthcare system level) 142, and Women's Health Medical Director 294).
Statistical Methods: The statistical models used depended on the nature of the outcome. For continuous measures, we use linear regressions; for binary (0/1) measures we used logistic regression, and for ratings of care (Always, Sometimes, Rarely, Never) we used ordinal logistic regression. For linear regressions, if the outcomes were counts, we performed sensitivity analyses with Poisson regressions. For Aim 1, we conducted single-level models for organizational characteristics predicting comprehensive care. For Aim 2, we conducted multi-level models for organizational characteristics predicting patient ratings of care, and/or receipt of guideline concordant care or intermediate outcomes.
We collected at least one survey at 255 sites nationally, covering all VISNs, 147 VA medical centers, 14 healthcare centers and 94 Community Based Outpatient Clinics. The response rates for each key informant varied: 89% for WVPMs, 59% for WH Medical Directors, and 45% for primary care and mental health leads. The linked data sources included SHEP, with 16,703 women Veterans who completed the survey during FY 2017. For EPRP we had quality of care and intermediate outcomes data for 67,533 women.
Aim 1A Availability of comprehensive women's health care in primary care:
We defined comprehensive care as the number of basic women's health services available on site. These include screenings and tests (e.g., cervical cancer screenings or Pap tests), trauma or safety screenings (e.g., Military Sexual Trauma screening), gender-specific care (e.g., diagnosis and treatment or menstrual disorders or menopausal symptom management) and contraception and pregnancy care. For general PC clinics, 79 percent of all sites reported offering 12-14 basic women's health (WH) services at their site (mean number: 12.3, SD: 3.0, range: 0-14). When we looked at the time component of availability for certain WH services (e.g., menstrual disorders evaluation and treatment, menopausal counseling and management, referral for screening mammogram) we found that 77 percent of sites reported offered 7-8 services all weekday hours (mean: 6.6, SD: 2.6, range: 0-8) v. some weekday hours or not available in primary care. Another component of comprehensiveness was the availability of female providers when requested. 84% of PC leads reported that female PC providers were always available when requested, and 81% reported that female PC providers designated to care for women Veterans were always available when requested. In primary care, 66% of sites reported that that MH providers were integrated in their Primary Care - Mental Health Integration (PC-MHI) program; 15% reported PC-MHI providers were co-located, 18% of sites reported MH providers were co-located but operated independently, and 1% reported other arrangements. Almost two-thirds of sites (66%) reported that same-day MH access was available, and 63% of sites reported that warm hand-offs to MH providers were available (when needed).
From WHMDs, we learned that 65% of sites reported having some sort of women's health clinics at VA sites. Among those sites, 74% reported a comprehensive women's health clinic was available (v. separate space without comprehensive services, or designated space or clinic within an existing primary care program. For availability of PC-MHI in these comprehensive women's health clinics, 24% of sites reported that MH providers were integrated in the women's health team, 26% of WHMD reported MH providers were co-located, but operated independently, 24% of sites had PC-MHI and long-term MH services available, and the remaining sites reported other arrangements (7%) or no mental health providers available in women's health clinic (19%). Of the sites reporting MH available in women's health clinics, 54% reported same day MH access in WH clinics, and 60% reported warm handoffs to MH providers were available when needed. For specialty mental health services, we also looked at the number of mental health services available as a measure of comprehensive care (e.g., mild-to-moderate MH conditions, severe depression and substance use disorders. 80% of sites reported having all services, average was 8.4, SD: 1.8, range 0-9). Another component of comprehensive care was outreach activities to non-enrolled and enrolled women Veterans. for non-enrolled women Veterans, mean: 5.3, SD: 1.5, (range 2-8), 7% reported doing all activities. For outreach to VA enrollees, the mean: 6.7, SD: 2.0 range (0-9), 19% reported doing all 9 activities listed in the past year.
Aim 1B - Predictors of availability of comprehensive women's health care in primary care:
Statistically significant predictors of availability of basic women's health services in general primary care clinics were impacted by WVPM and WHMD activities. For example, sites where WVPM reported greater involvement in environment of care improvement had more basic women's health services available, and more women's health services available all weekday hours. Sites where WVPM and WHMD reported working on the healthcare system quality improvement and the outreach committee also reported greater number of women's health services available. Similarly, sites where WHMD reported involvement in the facility strategic planning process had greater number of service available, and available all normal weekday hours. Sites that offered educational lectures on women's health issues also appeared to offer more services and had more services available all weekday hours. Surprisingly, sites where WVPMs served on medical executive committees reported fewer services available on site, or available during all weekday hours. For a number of outreach activities, we found that sites where there are women's health primary care providers to whom women Veterans are designated, and sites where WVPMs are involved with healthcare system quality improvement committees also reported greater number of outreach activities. Sites where WHMDs are on the OEF/OIF committee reported fewer outreach activities.
For Aim 2, we explored if the availability of elements of comprehensive care (e.g., availability of basic WH services, availability of integrated PC-MHI in primary care or women's health clinics, outreach activities to non-enrolled and enrolled women Veterans) was associated with women Veterans' quality of care (e.g., processes of care and intermediate health outcomes) (2A), and patient ratings of care (2B). We found that sites with greater number of basic women's health services also had higher rates of influenza immunizations among women Veterans. Sites with comprehensive women's health clinics reported higher likelihood of MOVE referrals among eligible populations of women Veterans, and CVD risk management and renal screening for women Veterans with diabetes. However, these same sites also reported lower likelihood of guideline concordant treatment for newly diagnosed depression, of well-controlled hypertension among diabetics, and of retinal exams among diabetics.
Sites reporting same-day mental health access in primary care reported greater likelihood of meeting guidelines for treating acute and continuous phases of depression. Sites where PC leads reported that warm handoffs to MH providers were available when necessary, also had greater likelihood of poor glycemic control among diabetes. In women's clinics offering MH care and/or warm handoffs, there was a higher likelihood of CVD screening for women Veterans with diabetes, but lower likelihood of guideline-concordant cervical cancer screening. Women Veterans with diabetes seeking care at sites with integrated providers in PC-MHI enjoyed greater likelihood of guideline-concordant retinal screens. Overall, there were mixed relationships between the availability of elements of comprehensive care and quality of care. While in some instances, women Veterans seeking sites with comprehensive care elements may be benefitting from better care (e.g., CVD risk management), there still appears to be challenges with some critical screenings (e.g., Pap tests), as well as follow-up care for patients with depression and diabetes.
For the patient ratings of health care experience measures, in general, we did not find any relationships between comprehensive care and women Veterans' ratings of care. The one exception is that women Veterans receiving care at sites with greater number of specialty MH services available, gave lower overall ratings of their primary care providers.
VA has made considerable strides in enhancing availability of comprehensive care to women Veterans, nonetheless, there are aspects of women's health in need of improvement. Women Veterans seeking care in VA primary care clinics should be able to obtain basic women's health services during normal weekday clinic hours. The vast majority (80%) of sites reported these services were available, however, one in five clinics do not.
We identified a number of strategies and practices that could be important in improving availability of comprehensive care; it was interesting to note that unlike findings of earlier studies, it was not structures (e.g., women's health clinics) that facilitated availability of women's health services, but rather WVPM and WHMD representation on key committees that made predicted service availability. We lack the data here to fully understand why, but qualitative interviews with women's health leaders may provide greater insight into how participation in certain committees may positively influence VA sites' ability to offer basic women's health services during weekday hours.
For women Veterans' quality of care, organizational factors appeared to impact quality of care, more than patients' experience of care. From the mixed results of organizational factors (some positive, some negative), we saw that structures that may benefit women Veterans' quality of care (e.g., women's health clinics), may also deplete resources needed for optimal follow-up care for chronic conditions (depression, diabetes). Surveying multiple informants appears to help assess when potential resource constraints may result in sites making trade-offs between the provision of primary care, women's health and mental health care. More research is needed to understand how and why these tradeoffs operate in practice to ensure delivery of comprehensive care for women Veterans more uniformly.
External Links for this Project
NIH ReporterGrant Number: I01HX000995-01
Dimensions for VADimensions for VA is a web-based tool available to VA staff that enables detailed searches of published research and research projects.
If you have VA-Intranet access, click here for more information vaww.hsrd.research.va.gov/dimensions/
VA staff not currently on the VA network can access Dimensions by registering for an account using their VA email address. Search Dimensions for this project
DRA: Health Systems
MeSH Terms: none