Managing Risks and Improving VA Health Care
In response to serious and longstanding problems with veterans’ access to care, which were highlighted in a series of congressional hearings in the spring and summer of 2014, Congress enacted the Veterans Access, Choice, and Accountability Act of 2014 (Pub. L. No. 113-146, 128 Stat. 1754), which provides $15 billion in new funding for Department of Veterans Affairs (VA) health care. Generally, this law requires VA to offer veterans the option to receive hospital care and medical services from a non-VA provider when a VA facility cannot provide an appointment within 30 days, or when veterans reside more than 40 miles from the nearest VA facility. Under the law, VA received $10 billion to cover the expected increase in utilization of non-VA providers to deliver health care services to veterans. The $10 billion is available until expended and is meant to supplement VA’s current budgetary resources for medical care. Further, the law appropriated $5 billion to increase veterans’ access to care by expanding VA’s capacity to deliver care to veterans by hiring additional clinicians and improving the physical infrastructure of VA’s facilities. It is therefore critical that VA ensures its resources are being used in a cost-effective manner to improve veterans’ timely access to health care.
VA operates one of the largest health care delivery systems in the nation. As of fiscal year 2014, VA was operating an expansive system of health care facilities, including 150 medical centers and more than 800 community-based outpatient clinics nationwide. In the years since the United States began conducting military operations in Afghanistan and Iraq, enrollment in the VA health care system has increased significantly—from 6.8 million veterans in fiscal year 2002 to 8.9 million veterans in fiscal year 2013. Consequently, VA has faced a growing demand by veterans for its health care services, a trend that is expected to continue. For example, the total number of annual outpatient medical appointments VA provided increased by 39.9 million visits (or about 85 percent) between fiscal years 2002 and 2013. Over that same period, Congress provided steady increases in VA’s annual health care budget, with amounts more than doubling, increasing from $23.0 billion to $55.5 billion between fiscal years 2002 and 2013. Despite these substantial budget increases, for more than a decade there have been numerous reports—by GAO, VA’s Office of the Inspector General, and others—of VA facilities failing to provide timely health care. In some cases, the delays in care or VA’s failure to provide care at all have reportedly resulted in harm to veterans.
While timely and cost-effective access to needed health care services is essential, it also is imperative that VA ensures the quality and safety of the services it provides. With the increased utilization of non-VA providers that is expected to occur as a result of the Veterans Access, Choice, and Accountability Act, veterans may be required to navigate multiple complex health care systems—the VA health care system and those of non-VA providers—to obtain needed health care services. Coordination of care between VA and non-VA providers is critical. Without it, there is increased risk of unfavorable health outcomes for veterans. For example, a lack of care coordination may lead to unnecessary duplication of services, which is not only costly, but may also pose health risks to veterans who may receive care that is not needed. Moreover, the quality of care may be adversely affected if important clinical information is not promptly communicated between VA and non-VA providers. Safeguarding the quality and safety of health care services provided within VA facilities is also essential. A series of infectious disease outbreaks at several VA facilities over the past several years—and allegations that VA officials may have withheld information about the outbreaks from the public—have raised concerns about the effectiveness of patient safety practices at its facilities.
These risks to the timeliness, cost-effectiveness, quality, and safety of veterans’ health care, along with other persistent weaknesses we have identified in recent years, raise serious concerns about VA’s management and oversight of its health care system. Based on these concerns, we have concluded that VA health care is a high-risk area and have added it to the High Risk List in 2015.
We have categorized our concerns about VA’s ability to ensure the timeliness, cost-effectiveness, quality, and safety of the health care the department provides into five broad areas: (1) ambiguous policies and inconsistent processes, (2) inadequate oversight and accountability, (3) information technology challenges, (4) inadequate training for VA staff, and (5) unclear resource needs and allocation priorities. We have made numerous recommendations that aim to address weaknesses in VA’s management of its health care system—more than 100 of which have yet to be fully resolved, including recommendations we made in regard to each of the following findings:
Ambiguous policies and inconsistent processes. Ambiguous VA policies lead to inconsistency in the way VA facilities carry out processes at the local level. In numerous reports, we have found that this ambiguity and inconsistency may pose risks for veterans’ access to VA health care, or for the quality and safety of VA health care they receive.
For example, in December 2012, we reported that unclear policies led staff at VA facilities to inaccurately record the required dates for appointments, and to inconsistently track new patients waiting for outpatient medical appointments at VA facilities. These practices may have delayed the scheduling of veterans’ outpatient medical appointments and may have increased veterans’ wait times for accessing care at VA facilities. In some cases, we found that staff members were manipulating medical appointment dates to conform to VA’s timeliness guidelines, which likely contributed further to the inaccuracy of VA’s wait-times data for outpatient medical appointments. Without accurate data, VA lacks assurance that veterans are receiving timely access to needed health care.
In November 2014, we reported that VA policies lacked clear direction for how staff at VA facilities should document information about veteran suicides as part of VA’s behavioral health autopsy program (BHAP). The BHAP is a national initiative to collect demographic, clinical, and other information about veterans who have died by suicide and use it to improve the department’s suicide prevention efforts. In a review of a sample of BHAP records from five VA facilities, we found that more than half of the records had incomplete or inaccurate information. The lack of reliable data limits the department’s opportunities to learn from past veteran suicides and ultimately diminishes VA’s efforts to improve its suicide prevention activities.
We have also identified gaps in VA policies related to facilities’ response to adverse events—clinical incidents that may pose the risk of injury to a patient as the result of a medical intervention or the lack of an appropriate intervention, such as a missed or delayed diagnosis, rather than due to the patient’s underlying medical condition. Specifically, we found that VA policies were unclear as to how focused professional practice evaluations (FPPE) should be documented, particularly what information should be included. An FPPE is a time-limited evaluation during which a VA facility assesses a provider’s professional competence when a question arises regarding the provider’s ability to provide safe, quality patient care. In our December 2013 report, we reported that gaps in VA’s FPPE policy may hinder VA facilities’ ability to appropriately document the evaluation of a provider’s skills, support any actions initiated, and track provider-specific incidents over time.
Inadequate oversight and accountability. We also have found weaknesses in VA’s ability to hold its health care facilities accountable and ensure that identified problems are resolved in a timely and appropriate manner. Specifically, we have found that (1) certain aspects of VA facilities’ implementation of VA policies are not routinely assessed by the department; (2) VA’s oversight activities are not always sufficiently focused on its facilities’ compliance with applicable requirements; and (3) VA’s oversight efforts are often impeded by its reliance on facilities’ self-reported data, which lack independent validation and are often inaccurate or incomplete.
In a July 2013 report, for example, we reported that VA needed to take action to improve the administration of its provider performance pay and award systems. In that report, we found that VA had not reviewed performance goals set by its facilities for providers and, as a result, did not have reasonable assurance that the goals created a clear link between performance pay and providers’ performance in caring for veterans. At four VA facilities included in our review, performance pay goals covered a range of areas, such as clinical competence, research, teaching, patient satisfaction, and administration. Providers who were eligible for performance pay received it at all four of the facilities we reviewed, despite at least one provider in each facility having personnel actions taken against them related to clinical performance in the same year. Such personnel actions resulted from issues including failing to read mammograms and other complex images competently, practicing without a current license, and leaving residents unsupervised during surgery.
In March 2014, we found that VA lacked sufficient oversight mechanisms to ensure that its facilities were complying with applicable requirements and not inappropriately denying claims for non-VA care. Specifically, the March 2014 report cited noncompliance with applicable requirements for processing a sample of non-VA emergency care claims. The noncompliance caused staff at four VA facilities to inappropriately deny about 20 percent of the claims we reviewed and to fail to notify almost 65 percent of veterans whose claims we reviewed that their claims had been denied. We found VA’s field assistance visits, one of the department’s primary methods for monitoring facilities’ compliance with applicable requirements, to be lacking. In these annual on-site reviews at a sample of VA facilities, VA officials were to examine the financial, clinical, administrative, and organizational functions of staff responsible for processing claims for non-VA care; however, we found that these visits did not examine all practices that could lead VA facilities to inappropriately deny claims. Further, although VA itself recommended that managers at its facilities audit samples of processed claims to determine whether staff processed claims appropriately, the department does not require VA facilities to conduct such audits, and none of the four VA facilities we visited were doing so.
In a September 2014 report and in three previous testimonies for congressional hearings, we identified weaknesses in VA’s oversight of veterans’ access to outpatient specialty care appointments in its facilities. VA officials told us they use data reported by VA facilities to monitor how the facilities are performing in meeting VA’s guideline of completing specialty care consults—requests from VA providers for evaluation or management of a patient for a specific clinical concern, or for a specialty procedure, such as a colonoscopy—within 90 days. We found cases where staff had incorrectly closed a consult even though care had not been provided, and found that VA does not routinely audit consults to assess whether its facilities are appropriately managing them and accurately documenting actions taken to resolve them. Instead, VA relies largely on facilities’ self-certification that they are doing so.
Information technology challenges. In recent reports, we also have identified limitations in the capacity of VA’s existing information technology (IT) systems. Of particular concern is the outdated, inefficient nature of certain systems, along with a lack of system interoperability—the ability to exchange information—which presents risks to the timeliness, quality, and safety of VA health care.
For example, we have reported on VA’s failed attempts to modernize its outpatient appointment scheduling system, which is about 30 years old. Among the problems cited by VA staff responsible for scheduling appointments are that the system requires them to use commands requiring many keystrokes and does not allow them to view multiple screens at once. Schedulers must open and close multiple screens to check a provider’s or a clinic’s full availability when scheduling a medical appointment, which is time-consuming and can lead to errors. VA undertook an initiative to replace its scheduling system in 2000 but terminated the project after spending $127 million over 9 years, due to weaknesses in project management and a lack of effective oversight. The department has since renewed its efforts to replace its appointment scheduling system, including launching a contest for commercial software developers to propose solutions, but VA has not yet purchased or implemented a new system.
In 2014, we reported that interoperability challenges and the inability to electronically share data across facilities led VA to suspend the development of a system that would have allowed it to electronically store and retrieve information about surgical implants (including tissue products) and the veterans who receive them nationwide. Having this capability would be particularly important in the event that a manufacturer or the Food and Drug Administration (FDA) recalled a medical device or tissue product because of safety concerns. In the absence of a centralized system, VA clinicians track information about implanted items using stand-alone systems or spreadsheets that are not shared across VA facilities, which makes it difficult for VA to quickly determine which patients may have received an implant that is subject to a safety recall.
Further, as we have reported for more than a decade, VA and the DOD lack electronic health records systems that permit the efficient electronic exchange of patient health information as military servicemembers transition from DOD to VA health care systems. The two departments have engaged in a series of initiatives intended to achieve electronic health record interoperability, but accomplishment of this goal has been continuously delayed and has yet to be realized. The ongoing lack of electronic health record interoperability limits VA clinicians’ ability to readily access information from DOD records, potentially impeding their ability to make the most informed decisions on treatment options, and possibly putting veterans’ health at risk. One location where the delays in integrating VA’s and DOD’s electronic health records systems have been particularly burdensome for clinicians is at the Captain James A. Lovell Federal Health Care Center (FHCC) in North Chicago, the first planned fully integrated federal health care center for use by both VA and DOD beneficiaries. We found in June 2012 that due to interoperability issues, the FHCC was employing five dedicated, full-time pharmacists and one pharmacy technician to conduct manual checks of patients’ VA and DOD health records to reconcile allergy information and identify possible interactions between drugs prescribed in VA and DOD systems.
Inadequate training for VA staff. In a number of reports, we have identified gaps in VA training that could put the quality and safety of veterans’ health at risk. In other cases, we have found that VA’s training requirements can be particularly burdensome to complete, particularly for VA staff who are involved in direct patient care.
In a November 2014 report that examined VA’s monitoring of veterans with major depressive disorder (MDD) and whether those who are prescribed an antidepressant receive recommended care, we determined that VA data may underestimate the prevalence of major depressive disorder among veterans and that a lack of training for VA clinicians on diagnostic coding may contribute to the problem. In a review of medical record documentation for a sample of veterans, we found that VA clinicians had not always appropriately coded encounters with veterans they diagnosed as having MDD, instead using a less specific diagnostic code for “depression not otherwise specified.” VA’s data on the number of veterans with MDD are based on the diagnostic codes associated with patient encounters; therefore, coding accuracy is critical to assessing VA’s performance in ensuring that veterans with MDD receive recommended treatments, as well as measuring health outcomes for these veterans.
In a May 2011 review, we found that training for staff responsible for cleaning and reprocessing reusable medical equipment (RME), such as endoscopes and some surgical instruments, was lacking. Specifically, VA had not specified the types of RME for which training was required; in addition, VA provided conflicting guidance to facilities on how to develop this training. Without appropriate training on reprocessing, we found that VA staff may not be reprocessing RME correctly, posing patient safety risks.
In our October 2014 report on VA’s implementation of a new, nationally standardized nurse staffing methodology, staff from selected VA facilities responsible for developing nurse staffing plans reported that VA’s individual, computer-based training on the methodology was time-consuming to complete and difficult to understand. These staff members said they had difficulty finding the time to complete it while also carrying out their patient care responsibilities. Many suggested that their understanding of the material would have been greatly improved with an instructor-led, group training course where they would have an opportunity to ask questions.
Unclear resource needs and allocation priorities. In many of our reports, we have found gaps in the availability of data required by VA to efficiently identify resource needs and to ensure that resources are effectively allocated across the VA health care system.
For example, in October 2014, we reported that VA facilities lacked adequate data for developing and executing nurse staffing plans at their facilities. Staffing plans are intended to help VA facilities identify appropriate nurse staffing levels and skill mixes needed to support high-quality patient care in the different care settings throughout each VA facility, and are used to determine whether their existing nurse workforce sufficiently meets the clinical needs of each unit, or whether they need to hire additional staff. At selected VA facilities, staff responsible for developing and executing the nurse staffing plans told us that they needed to use multiple sources to collect and compile the data—in some cases manually. They described the process as time-consuming, potentially error-prone, and requiring data expertise they did not always have.
In a May 2013 report, we reported that VA lacked critical data needed to compare the cost-effectiveness of non-VA medical care to that of care delivered at VA facilities. Specifically, VA lacks a data system to group medical care delivered by non-VA providers by episode of care—all care provided to a veteran during a single office visit or inpatient stay. As a result, VA cannot efficiently assess whether utilizing non-VA providers is more cost-effective than augmenting its own capacity in areas with high non-VA health care utilization.
In a September 2014 report, we identified concerns with VA’s management of its pilot dialysis program, which had been implemented in four VA-operated clinics. Specifically, we found that, five years into the pilot, VA had not set a timetable for the completion of its dialysis pilot or documented how it would determine whether the pilot was successful, including improving the quality of care and achieving cost savings. We also found that VA data on the quality of care and treatment costs were limited due to the delayed opening of two of the four pilot locations. Veterans who receive dialysis are one of VA’s most costly populations to serve, but VA has limited capacity to deliver dialysis in its own facilities, and instead refers most veterans to non-VA providers for this treatment. VA began developing its dialysis pilot program in 2009 to address the increasing number of veterans needing dialysis and the rising costs of providing this care through non-VA providers.
Although VA has taken actions to address recommendations we have made related to VA health care, there are currently more than 100 that have yet to be fully resolved, including recommendations related to the five broad areas of concern highlighted above. For example, to ensure that its facilities are carrying out processes at the local level more consistently—such as scheduling veterans’ medical appointments and collecting data on veteran suicides—VA needs to clarify its existing policies. VA also needs to strengthen oversight and accountability across its facilities by conducting more systematic, independent assessments of processes that are carried out at the local level, including how VA facilities are resolving specialty care consults, processing claims for non-VA care, and establishing performance pay goals for their providers. We also have recommended that VA work with DOD to address the administrative burdens created by the lack of interoperability between their two IT systems. A number of our recommendations aim to improve training for staff at VA facilities, to address issues such as how staff are cleaning, disinfecting, and sterilizing reusable medical equipment, and to more clearly align training on VA’s new nurse staffing methodology with the needs of staff responsible for developing nurse staffing plans. Finally, we have recommended that VA improve its methods for identifying VA facilities’ resource needs and for analyzing the cost-effectiveness of VA health care.
The recently enacted Veterans Access, Choice, and Accountability Act included a number of provisions intended to help VA address systemic weaknesses. For example, the law requires VA to contract with an independent entity to (1) assess its capacity to meet the current and projected demographics and needs of veterans who use the VA health care system, (2) examine VA’s clinical staffing levels and productivity, and (3) review VA’s IT strategies and business processes, among other things. The new law also establishes a 15-member commission, to be appointed primarily by bipartisan congressional leadership, which will examine how best to organize the VA health care system, locate health care resources, and deliver health care to veterans. It is critical for VA leaders to act on the findings of this independent contractor and congressional commission, as well as on those of VA’s Office of the Inspector General, GAO, and others, and to fully commit themselves to developing long-term solutions that mitigate risks to the timeliness, cost-effectiveness, quality, and safety of the VA health care system.
It is also critical that Congress maintain its focus on oversight of VA health care. In the spring and summer of 2014, congressional committees held more than 20 hearings to address identified weaknesses in the VA health care system. Sustained congressional attention to these issues will help ensure that VA continues to make progress in improving the delivery of health care services to veterans.
We plan to continue monitoring VA’s efforts to improve the timeliness, cost-effectiveness, quality, and safety of veterans’ health care. To this end, we have ongoing work focusing on topics such as veterans’ access to primary care and mental health services; primary care productivity; nurse recruitment and retention; monitoring and oversight of VA spending on training programs for health care professionals; mechanisms VA uses to monitor quality of care; and VA and DOD investments in Centers of Excellence—which are intended to produce better health outcomes for veterans and service members.
GAO-15-55: Published: Nov 12, 2014. Publicly Released: Dec 12, 2014.
GAO-15-61: Published: Oct 16, 2014. Publicly Released: Nov 17, 2014.
GAO-14-808: Published: Sep 30, 2014. Publicly Released: Oct 30, 2014.
GAO-14-646: Published: Sep 2, 2014. Publicly Released: Sep 2, 2014.
GAO-14-463T: Published: Apr 2, 2014. Publicly Released: Apr 2, 2014.
GAO-14-175: Published: Mar 6, 2014. Publicly Released: Mar 6, 2014.
GAO-14-302: Published: Feb 27, 2014. Publicly Released: Feb 27, 2014.
GAO-14-146: Published: Jan 13, 2014. Publicly Released: Jan 13, 2014.
GAO-14-55: Published: Dec 3, 2013. Publicly Released: Dec 3, 2013.
GAO-13-536: Published: Jul 24, 2013. Publicly Released: Aug 23, 2013.
GAO-13-441: Published: May 31, 2013. Publicly Released: May 31, 2013.
GAO-13-130: Published: Dec 21, 2012. Publicly Released: Jan 18, 2013.
GAO-12-669: Published: Jun 26, 2012. Publicly Released: Jun 26, 2012.
GAO-11-391: Published: May 3, 2011. Publicly Released: May 3, 2011.