Delivering Quality to Patients

In this issue of JAMA, Liu and colleagues¹ report that for several surgical procedures, a disproportionately small number of ethnic minorities and poorly insured patients receive care in high-volume hospitals, where quality of care is assumed to be superior. The authors suggest that there is a need for explicit measures to address this disparity. Although intuitively appealing, the authors' observations and suggestions implicitly embrace 2 assumptions that deserve closer scrutiny: (1) ethnic minority and poorly insured patients would want to go to high-volume hospitals if they knew the benefits and could overcome barriers to access, and (2) volume-based referral policies are a good way to improve surgical quality.

The easiest explanations for why ethnic minority and poorly insured patients are less likely to use high-volume hospitals are that they cannot (because of barriers to access) or that they may not be aware of other options (because of lack of information). These are real problems that society needs to address, but there is another possible explanation for why some patients do not go to high-volume hospitals—they do not want to. Just as shoppers do not all choose to shop at the stores with the best values, some patients may not choose to receive care at hospitals that health care experts believe deliver the highest-quality care. To some, the desirability of a hospital may be driven more by familiarity, ethnic or cultural makeup of hospital staff and patients, or other factors related to social comfort. Because of racial discrimination, ethnic minorities may be more inclined to seek care at hospitals where they already feel comfortable. Among the poor who live on the social margins, there may be a greater desire to avoid moving from a familiar hospital to an unfamiliar one. The popular appeal of what is referred to as “high-quality” care is probably grossly overestimated, while the strength of patients' preferences for “lower-quality” hospitals where they may feel socially more at ease is probably underestimated. Quality to a patient is a lot more than a lower surgical mortality rate, especially when the patient does not expect to die.

Given a choice, many patients do not want to change hospitals, even knowing that results might be better elsewhere. This is a lesson our research group was surprised to learn 10 years ago, when we systematically studied the preferences of 100 rural veterans with respect to the location of their surgical care.² Using a standardized survey, the amount of additional operative risk these veterans would be willing to accept to keep their surgical care local rather than travel to a distant hospital with better average outcomes was quantified. The majority of these patients were willing to accept at least some additional risk, and about one fifth of them were willing to accept very high levels of risk to keep their surgical care local. It is true that the preferences of this cohort may not be generalizable to the broader population. For example, in the study by Liu et al,¹ the high-volume hospitals that were not used were often nearby. However, our findings among veterans demonstrated that there is a difference between what physicians see as quality, and what patients see as desirable.

If patients see the desirability of their hospitals in terms broader than surgical mortality rates, then volume-based referral policies may displease many of the very patients they are designed to benefit, especially ethnic minorities and the poor for whom disruptions of familiar patterns of care may be particularly socially discomfiting. Some might discount this concern, ascribing greater importance to traditional health outcomes than to intangible social values, but it is important for policy makers to address both interests.

In addition to the apparent disregard of volume-based referral policies for social values, there are many other reasons to question whether such policies are a good way to improve surgical quality. A host of practical concerns have been voiced, including the logistical challenge of moving patients, the scarcity of high-volume hospitals in many health care markets,³ and the adverse effect such policies might have on lower-volume hospitals.⁴ An additional criticism is the relatively small effect such policies would have on surgical care as a whole. While nearly all agree that the volume-mortality correlation is real, for the vast majority of surgical procedures the magnitude of the effect is not large, nor is the correlation perfect. The procedures targeted for volume-based referral are therefore appropriately limited to a few high-risk operations, and thus the great preponderance of surgical care is left “unimproved.” Liu et al have now added one more reason to be concerned about volume-based referral policies: if such policies are payer-driven, the disparity between those who are insured and those who are uninsured would likely worsen.

At a fundamental level, volume-based referral is a lazy approach to quality improvement. In fact, volume-based referral has very little to do with quality improvement at all. When a payer mandates that patients be sent to high-volume hospitals exclusively, the average outcome for that payer's patients may be better, but the quality of care delivered by individual hospitals remains unchanged. In a sense, volume-based referral policies are an “end run” around the issue of quality: they neither require that the essential components of quality be identified (they just measure volume as a proxy for quality), nor do they require that quality is improved at hospitals (they simply direct the patient to specific centers that already meet that standard). The central focus of quality improvement should be the task of delivering quality care to patients, not the other way around. Granted, there probably is a role for volume-based referral for the few procedures for which the volume-outcome association is particularly strong (eg, pancreatic surgery). However, an approach that simultaneously sidesteps the task of improving quality and ignores the vast majority of surgical procedures should not be the crown jewel of the surgical quality movement.

Despite unprecedented activity in the area of health care quality research, there is still much to learn about what quality is and how to deliver it. Providing quality equitably in a setting of diverse values and preferences, racial and social barriers, and differences in ability to pay is a particularly challenging task. Of course, the only true remedy for ethnic and economic disparities in health care is to eradicate the racial discrimination and economic injustice of the society in which the US health care system functions. Until that ideal is achieved, physicians must remain vigilant.