It Is Possible

Palliative care is specialized medical care for people with serious illnesses. It is focused on giving patients relief from the symptoms, pain, stress, and treatments of a serious illness—whatever the diagnosis. The goal is to provide the best possible quality of life for the patient and the family.¹ Although 66% of large hospitals have a palliative care program,² there is significant variation in the types of programs and services offered, giving rise to concern regarding variation in the quality of care provided by these programs.

The article by Walling and colleagues³ in this issue of JAMA Internal Medicine has important implications for the field of palliative medicine that reach beyond the specific findings of the study. Veterans Affairs (VA) has exemplified the type of quality improvement initiatives that need to occur across a broad range of health care settings and populations to expand access to supportive cancer care. The study also found that despite system-wide efforts to improve access and quality, a significant amount of work remains to elevate the quality of supportive cancer care provided in the VA, which underscores the need for quality initiatives to be iterative and ongoing.

One of the most significant findings of the study is that 86.4% of the veterans who died an expected death during the study period were referred to either palliative or hospice care, reflecting their deep penetration within the VA system. Although the study evaluated only a cohort of veterans with common solid tumors (as opposed to other noncancer diagnoses), the cohort had varying prognostic and clinical features. The Veterans’ Health Care Eligibility Reform Act of 1996 (Title 38 Code of Federal Regulations §17.38) standardized the provision of hospice and palliative care to eligible veterans who need these services. It established that hospice and palliative care are covered services, having equal priority with any other medical care service provided by the VA. This standardization of access systemwide created a culture in the VA where palliative and hospice care are integrated into the continuum of cancer care. The results of this study support the use of new standards for access to quality hospice and palliative care as a potential policy lever for the Centers for Medicare & Medicaid Services and accreditation bodies to scale palliative care delivery across health care settings outside the VA.

A second important implication of this study for the field of palliative medicine is that it serves as a proof of concept that quality of care can be measured in an older and medically complex patient population. This is a pivotal finding, since discussions regarding improving the quality of palliative care are premised on the notion that we can systematically measure quality. This study demonstrates that such measurement is possible and should be replicated across systems and settings. The study used the Cancer Quality–Assessing Symptoms and Side Effects of Supportive Treatment (ASSIST) measures, which span multiple domains of care from symptom management to spiritual support. The VA has demonstrated that quality can be measured in the veteran population using this tool. The study also highlights, however, that the assessment of meaningful and person-centric measures is costly. Specifically, data abstraction for the Cancer Quality–ASSIST measures required 3 days of training for experienced oncologic nurses, access to a real-time consultation with a senior nurse reviewer, and the use of the VA’s information technology infrastructure, which is more comprehensive than that of most other integrated systems or health care providers. On average, complete medical record abstraction required more than 2 hours for each patient. As such, a significant barrier to other health care systems or providers in carrying out similar quality improvement initiatives is the potentially high cost of the measurement process.

The study by Walling et al³ also demonstrates the significant gap between the evidence base for supportive care processes that improve quality of life for patients with cancer and the variable implementation of such care processes in standard oncologic practice. This implementation “gap” represents a significant challenge for the field of palliative medicine and is the focus of 2 recent articles^4,5 regarding the need for investigators to write about their implementation experiences within a scientific framework and with a common language to maximize learning from current and past initiatives. Although we know what works for improving care in a population of older veterans, we need more information regarding how and why some processes of care are more consistently and successfully implemented than others. Why was the VA more successful in improving quality indicators in the inpatient setting? A greater focus on implementation research would enable innovative approaches to overcoming barriers in adopting evidence-based interventions to be more widely understood and disseminated.

Last, the study by Walling et al³ highlights the unique aspects of the VA system that both facilitate the assessment of quality and may have led to the relatively high quality scores in some domains. Despite the fact that, on average, veterans received only about half of recommended care, significant variation occurred across measures and settings, and there is reason to believe that the VA may be on the higher end of success in terms of these outcomes compared with other settings.^6–8 As an integrated health care system functioning under a global budget with a fully employed clinical staff, the incentives promoting supportive care services are likely stronger than in nonintegrated systems in which the quality improvements and efficiencies generated from greater supportive care services may not be enjoyed or even recognized by the provider of such services. The extent to which those system design factors and incentives influenced the care provided in the VA is unclear but worthy of further study. Similarly, the extent to which the VA links clinician incentives, training, and electronic medical record processes to make quality measurement and assessment more visible and actionable at the point of care is unknown. Does the electronic medical record prompt clinicians to assess shortness of breath? Does the VA reward or penalize movement on the Cancer Quality–ASSIST measures? What were the specific actionable steps taken by the VA to address the results of this study? More information about these issues would facilitate widespread adoption of these care processes and support efforts for ongoing improvement in the quality of palliative care in the VA and nationally.