OBJECTIVES
Objectives
Define quality improvement (QI).
Define goals for applying QI to reduce health-care inequities.
Describe how QI can affect health-care inequities
Review the importance of collecting data on race, ethnicity, language, and social and behavioral risk factors.
Review the importance of patient safety for vulnerable patients.
Describe the “Triple Aim” QI framework.
Describe the eight steps of a QI initiative.
Olivia has worked for 3 months at a community health center that serves a vulnerable patient population. The staff is wonderful, with many dedicated people, but some are worn down by chronic staffing shortages and overwhelming patient needs. The patients are very poor with many psychosocial stressors. Some have no telephone or are homeless; others cannot afford bus fare to the clinic; about at third speak little or no English. A small yet significant number of patients use illicit drugs. Olivia has been given 10% “QI time” in her job description to help the clinic tackle long-standing quality of care issues and to start making some operational and clinical improvement. While finding the task somewhat daunting, she is excited to dive in. However, she is not sure how to begin.
INTRODUCTION
Even the most dedicated and idealistic practitioners can be daunted in their quest to care for complicated patients in limited resource settings. Health-care delivery systems can support the individual practitioner in that quest. Those systems that focus on continually improving patient health (outcomes) define improved care (performance) as requiring the constant, combined efforts of everyone in the system. System-based approaches to improving patient outcomes and care can follow formal methods of analysis and strategies for progress.
Quality Improvement (QI) is one system-based approach borrowed from other industries that has been applied to health care. The Agency for Healthcare Research and Quality defines QI as “a multidisciplinary, systems-focused, data-driven method of understanding and improving the efficiency, effectiveness, and reliability of health processes and outcomes of care.”1,2 While interest in QI has exploded in the United States over the past two decades, the extent to which those efforts have improved health care for vulnerable patients is unclear. This chapter addresses the specific challenges of quality improvement in safety-net settings with vulnerable patient populations. We present an eight-step approach to designing, implementing, and evaluating QI efforts, emphasizing data-driven and team-based approaches, adapted from the Institute for Healthcare Improvement. We describe concrete steps that can be used to guide QI efforts, with advice about how to adapt or expand their use in safety-net settings.
QUALITY IMPROVEMENT AND VULNERABLE PATIENT POPULATIONS
Since the publication of the Institute of Medicine (IOM) report “Crossing the Quality Chasm,”3 a report that documented the ways in which the US health-care system not only failed to deliver high quality care but actually harmed patients, the United States has experienced a surge in QI activities. QI efforts have been closely tied to efforts to improve patient safety (see “Patient Safety and Vulnerable Populations”) and to decrease medical errors. While there is debate about how best to improve quality,4,5 QI has succeeded when there is engaged leadership, adequate resources, and experience, and the processes are data-driven, stepwise approaches to improving services. Quality improvement efforts have been promoted by many organizations.
The potential for QI initiatives to decrease health disparities has long been recognized. In the report “Crossing the Quality Chasm,” from the Institute of Medicine (IOM), included improving “equity” and reducing racial and ethnic disparities as one of the six critical pillars of improving health-care quality in the nation.3
The IOM’s 2003 report, “Unequal Treatment,” further focused attention on health-care disparities. However, the traditional, implicit stance of the QI movement has been that, while health-care disparities exist and should be addressed, a well-designed QI initiative will be of benefit to all patient groups. The Institute for Health Care Improvement, for example, in their 22-session QI educational series for health-care trainees, has no module that addresses the relationship of QI and health-care disparities,6 or how to use QI to address health-care disparities. The assumption has been that overall improvement in quality of care will help patients vulnerable to health-care disparities, or that a “rising tide lifts all boats.”
The Health Research and Services Administration’s (HRSA) National Health-Care Disparities initiative, on the other hand, has explicitly attempted to link quality improvement efforts with a reduction in health-care disparities. Nevertheless, the vast majority of QI initiatives have had neither the explicit goal of decreasing health-care disparities nor have they assessed the interventions for their impact on health-care disparities. Failing to explicitly consider the needs of vulnerable patients and neglecting to assess the impact of QI initiatives on patient subgroups may result in actually creating health-care inequities. The QI efforts may not benefit all groups equally or could, in some cases, decrease health-care quality for some patient groups.
As the number of vulnerable patients increases, it is especially critical to realize the potential of QI efforts to reduce health-care disparities. Recognizing the major challenges to ensuring that QI efforts reduce racial, ethnic, and language disparities is vital for success.7 Documenting disparities is never a comfortable undertaking, even when attempting to improve them. Indeed, attempting to diminish the appearance of disparities can create perverse incentives for providers to avoid serving vulnerable patients altogether. Fully engaging vulnerable patients in QI programs may also present difficulties due to language or other barriers. QI projects must also be tailored to address specific needs of vulnerable populations. Furthermore, institutions that provide care to underserved patients may not have the resources to fully implement adequate QI programs.
To assure that QI initiatives realize their promise to decrease health-care inequities, they must be designed specifically to do so. Health-care disparities should be assessed before, during, and after implementation of the initiative. Careful attention needs to be paid to avoid incentives for providers to avoid higher-risk patients. Interventions also must be fashioned to overcome barriers to implementation. For example, programs need to be applied regardless of patient language and in organizations that care for minority patients or patients with socioeconomic challenges.
PATIENT SAFETY AND VULNERABLE POPULATIONS
Mr. P, a patient with long-standing diabetes on insulin and a major depressive disorder, had a syncopal episode while riding the bus. He was taken to the emergency department, where he was found to be hypoglycemic. He subsequently revealed to his primary care provider that he occasionally skipped meals because of lack of access to food, and was not knowledgeable about how to manage his diabetes medications when he did not eat.
Patient safety, or the prevention of harm to patients, poses particular challenges for vulnerable populations.8 Communication challenges such as limited health literacy or limited English proficiency can make self-management of chronic diseases very difficult for patients and families. Inability to interpret medication labels correctly, for instance, can lead to poor or dangerous medication use.
Poverty impedes the management of chronic conditions in indirect ways, such as not adhering to medications due to cost,9 or direct ways, such as lack of food access leading to hypoglycemia.10 Moreover, economic pressure increases the demand of caregiving, as vulnerable patients and families choose between lost wages and supervision of ill family members.
THE IMPACT OF QI ON HEALTH DISPARITIES: FOUR POSSIBLE SCENARIOS
A review of several different QI projects demonstrates some of the complexities in understanding the effect QI projects may have on health-care inequities (Figure 18-1).
In some cases, QI initiatives have improved the quality of care for all patients and closed the health-care gap for our most vulnerable patients (Figure 18-1A). One large-scale example is HRSA’s Health Disparities Collaborative, where health-care disparities in cancer screening were reduced through a traditional QI approach.11,12,13 Another well-known example is a CMS initiative to improve outcomes in hemodialysis patients that both improved care overall and reduced disparities in care for African-American patients.14
In other cases, the rising tide may actually lift all boats, while disparities remain constant (Figure 18-1B). In such cases, care does improve for vulnerable patients, but no more than it improves for the overall population. Examples of this include QI initiatives in the Veteran’s Administration (VA) and Medicare populations, in which interventions across multiple domains of care such as diabetes and cancer screening improved process and outcome metrics for both white and African-American populations, but because improvement was not differentially greater for African Americans, the disparity remained.15 Another example includes the study by Jha et al of the effects of pay-for-performance on poor performing hospitals, which showed consistent overall improvement but no change in health-care disparities.16
In other scenarios, a QI initiative may not improve quality of care for a population overall—but does improve care for vulnerable patients (Figure 18-1C). A large-scale initiative to improve quality of care for patients with diabetes, for example, failed to demonstrate population-wide improvement, but analysis of a subgroup population of low-literacy patients showed significant improvement.17
Another scenario is possible, however. If incentives are applied incorrectly or the intervention’s impact on vulnerable groups is not continually assessed, QI interventions can worsen health-care disparities (Figure 18-1C). The reasons for this are myriad—physicians and health-care organizations may feel that they have to “cherry-pick” patients to remain competitive, or they may design or implement interventions in ways that do not reach their most vulnerable patients. When New York City instituted its coronary artery bypass report cards, for example, hospitals instituted widespread changes in patient selection for surgery to avoid high-risk patients that would worsen their score. Using methods that included racial profiling, the bypass report cards worsened outcomes for minority patients.18
As the challenge of combining QI and health-care disparities reduction efforts has garnered greater attention, several approaches have emerged. These include the standardized collection of data on race, ethnicity, and English language proficiency for all patients; the creation of “equity reports” for organizations in the era of health-care reform19; and the careful crafting of pay-for-performance programs to encourage the reduction of health-care disparities.20,21,22,23,24
Pay for performance (P4P) refers to the increasingly common practice of offering financial incentives to physicians and other health-care providers to meet defined quality, efficiency, or other targets.25 Although P4P remains a small portion of overall health spending and health-care provider compensation in the United States, P4P often drives quality improvement initiatives.
P4P programs may exacerbate health disparities.20,21,22,23,24 Physicians and health systems caring for low-income populations care for more uninsured or publically insured patients. In turn, these systems lack resources to invest in quality improvement processes and information systems. Thus, systems that care for more advantaged populations may improve, while safety-net health systems, and the vulnerable populations they serve, are left behind. This argument is referred to as the “inverse care law,” or the idea that care is least available to those who need it most.26
Moreover, it is clear that patients’ social context affects their ability to adhere to treatment recommendations; P4P measures that encompass patient adherence to recommended screening or treatment will be influenced by patients’ social conditions. Studies have demonstrated that safety-net health systems, those that disproportionately serve vulnerable populations, do not perform as well on quality reporting.27,28,29 The National Quality Forum has recommended risk adjustment for socioeconomic characteristics in quality measurement and reporting, to account for social differences that impact outcomes.
In addition to risk adjustment, a further recommendation is to reward relative improvements in performance rather than an absolute performance threshold, which may benefit safety-net systems. In one study, financial incentives led to improvements in safety-net health systems, bring them on par with other health systems.16
