Reflecting on Patient Choice: Identifying Best Decisions and Understanding Rationality and Cost-Effectiveness

In the last issue of MHR, Margaret Holmes-Rovner questioned whether informing patients of treatment options supports patient choice. The contributions in this InkLinks continue this discussion.

InkLinks is a regular column in which readers reflect on issues related to the lead article. It is meant to tap the rich intellectual resources that this network provides. We welcome your contribution at leahtamar@comcast.net.

Challenges to Identifying Best Decisions

Patient choice is receiving increased attention as a means of improving quality and outcomes of health care. Patient-centered decision supports are being tested to foster patients becoming well-informed and proactive users of health services. Initial efforts are underway to define the quality criteria to evaluate these tools, such as the IPDAS criteria mentioned by Holmes-Rovner, but these criteria lack “best” choices criteria. Holmes-Rovner identifies several key issues as potential barriers to identifying an individually-best decision: ethical debates about patient autonomy, difficulty in measuring personal values, and effective approaches to providing information with “pathways” to aid patients to make well-informed, best decisions.

Several challenges to identifying best decisions can be highlighted.

One issue is that there is often debate about the extent to which a best choice truly can be identified for an individual based on the gaps in existing research knowledge. This is particularly so for populations in which health interventions have not necessarily been well-researched, including populations for which marked health disparities exist. In this type of situation, it may be premature to recommend a best choice. However, best choices should be identified where they do exist. Where patients, health care providers, and the scientific evidence are strongly in agreement about what would be best to do, choices may be “easier” and not require formal decision support approaches.

A second issue is the urgent need (highlighted in the current evidence-based practice movement) for developing and testing translation models to guide the interpretation of scientific evidence for use in decision-making. The need for good translation, including making the central “take home” points of evidence understandable and usable, can apply to both clinicians and their patients. Holmes-Rovner addresses this issue as the proliferation of information and choices which can overwhelm decision makers and result in sub-optimal choices. Early versions of patient-focused decision supports have tended to be overly-complex, text- and number-dense, and not necessarily well-matched to the information preferences, literacy, cultural, and health constraints of a number of individuals. To make a meaningful impact with the populations that are often in the most need of empowerment in making choices, there is a need to “de-complexify” decision supports as appropriate to better match what individuals can and prefer to handle, to foster greater ease of use with less cognitive burden. Indeed less can be more.

A third issue is that the boundaries between informing and persuading are not always clear-cut. What should be the criteria for identifying a best choice for an individual, especially if the situation appears almost too close to call? Formal decision analysis approaches have often been advocated as a means of sifting through complex decisions to arrive at a rational best choice option. However, decision analysis is only as useful as the data that inform the model, and cannot resolve some of the fundamental ethical questions about the extent to which patient autonomy should determine the choices that are made.

Celia E. Willis, Ph.D., R.N.

College of Nursing

Patient Choice, Rationality, and Cost-Effectiveness

Dr. Holmes-Rovner’s discussion raises important questions about the relationships among empowering patient choice, enhancing patients’ understanding of their options using decision support aids, and encouraging a more cost-effective health care system.

She’s surely right to point out that we can draw no simple line from increased patient understanding, to more rational patient choice, to greater patient autonomy, to lower health care costs. A large part of the reason is that there are different kinds of “rationality” at work in each of these contexts. The rationality of cost-effectiveness aims to maximize aggregate population benefits against aggregate cost. The rationality of individual choice is focused on maximizing individual utility, the net of my gains and losses. A choice may often be rational for me, but at odds with what’s most socially desirable. And the rational autonomous choice is something more than utility maximizing, since it aims to make the choice most consistent with those values and goals I hold dear. Most of us (sociopaths and economists excluded) are not mere utility maximizers. We care about more than our own self-interest, and often make fully autonomous choices that leave us worse off.

She’s also right to be skeptical of the impact that enhanced patient understanding will have on cost-effectiveness. Any reduction in cost from greater patient understanding requires that patients’ goals and values regarding treatment are often different than those that motivate their physicians’ recommendations; and that the patients’ goals and values tend to favor less resource use. There’s not much reason to believe either of these is generally true, which is not to say they’re never true. No doubt they partly explain the particular contexts in which patient involvement in choice has had an impact: hysterectomy; prostate screening and treatment; and the choice between mastectomy and lumpectomy. The impact of these decisions on sexuality and self-image is felt much differently by patients than by their doctors, and prioritized differently in deciding what choice is best.

The great bulk of medical treatment decisions are not like this, however. Most of the time, my priorities and my doctor’s coincide. When they do, I’m not going to make any more cost-effective a choice than my doctor. And when they do I won’t have much interest in exercising my autonomous choice, either. This is not to disparage the importance of respecting patient autonomy; it’s only to focus its relevance. Respecting and enabling autonomy is most important where patient and doctor are most likely to reach different conclusions. The “failures” of patient decision aids may be just as important as their “successes” in mapping the geography of patient autonomy. When we know better where to invest our efforts, and what to pass by, we will be more successful at both reducing costs through enhanced patient understanding and at empowering patient autonomy where it really matters.

Tom Tomlinson, Ph.D.

Center for Ethics and Humanities

in the Life Sciences