Abstract
Decision making for incompetent patients is a much-discussed topic in bioethics. According to one influential decision making standard, the substituted judgment standard, a surrogate decision maker ought to make the decision that the incompetent patient would have made, had he or she been competent. Empirical research has been conducted in order to find out whether surrogate decision makers are sufficiently good at doing their job, as this is defined by the substituted judgment standard. This research investigates to what extent surrogates are able to predict what the patient would have preferred in the relevant circumstances. In this paper we address a methodological shortcoming evident in a significant number of studies. The mistake consists in categorizing responses that only express uncertainty as predictions that the patient would be positive to treatment, on the grounds that the clinical default is to provide treatment unless it is refused. We argue that this practice is based on confusion and that it risks damaging the research on surrogate accuracy.
Similar content being viewed by others
Notes
This is not to say that incompetent patients are, by definition, incapable of choosing among alternatives, but only that they lack the ability to make sufficiently rational and informed decisions. What the latter exactly amounts to is controversial, but there is no shortage of uncontroversial examples. Patients who are in a state of coma or who suffer from severe dementia, for example, clearly lack the relevant ability.
This assumption could be questioned since, among other things, it presupposes extensive self-knowledge.
The studies differ from one another in many respects, such as with respect to the chosen patient populations, how the surrogates are identified, the treatment options used, and more. These differences, however, will be of no importance in the following.
This is not the only questionable way to make use of the “unsure” responses. Another has to do with assigning every response—predictions as well as non-predictions—a value, for example to discern the percentage of people who prefer treatment. See Pruchno et al. [10].
The rationale behind the clinical default may be found in the highly influential President’s Commission for the Study of Ethical Problems in Medicine and Biomedical and Behavioral Research [13]. There it is suggested that a best interest standard (i.e., promoting the patient’s well-being) ought to be adopted when a patient’s decision is likely to be unknown.
Of the 16 studies reviewed by Shalowitz and colleagues [4], at least seven dichotomized the responses in the manner that is criticized in this paper. This is no doubt a substantial part of the total amount of the studies on surrogate accuracy, but it is far from all. It shows that researchers do not normally pay any significant attention to the fact that different dichotomizing strategies are used in the research.
Again, this is on the assumption that surrogates’ “unsure” responses are meant to convey surrogate uncertainty, or unwillingness to make a prediction, but this dichotomization strategy obviously does not fare better on the much less plausible assumption that these responses are predictions of patient uncertainty or indifference. Treating such predictions as predictions of positive attitudes is equally misleading, and ought to be considered inaccurate whenever the patient is positive (or, of course, negative) towards treatment.
Another statistical issue concerns the effect on mere chance accuracy. For example, by treating two of three choice options as relevantly the same, the mere chance accuracy becomes higher in relation to the case where “unsure” responses are excluded from the analysis.
We owe this way of framing the issue to an anonymous referee.
Given other interpretations of the non-predictive responses, we might have reasons to view case two and case three differently. And when it is unclear how to interpret “unsure,” they ought to be excluded from the analysis, while they do neither indicate surrogate accuracy nor surrogate inaccuracy. Notice, that there is no reasonable interpretation that justifies the standard practice of treating these responses as accurate predictions.
References
Broström, L., and M. Johansson. 2007. Surrogates have not been shown to make inaccurate substituted judgments. Unpublished manuscript.
Ditto, P.H., et al. 2001. Advance directives as acts of communication: a randomized controlled trial. Archives of Internal Medicine 161, no. 3: 421–30.
Layde, P.M., et al. 1995. Surrogates’ predictions of seriously ill patients’ resuscitation preferences. Archives of Family Medicine 4, no. 6: 518–523.
Shalowitz, D.I., E. Garrett-Mayer, and D. Wendler. 2006. The accuracy of surrogate decision makers: a systematic review. Archives of Internal Medicine 166, no. 5: 493–497.
Uhlman, R.F., R.A. Pearlman, and K.C. Cain. 1989. Understanding of elderly patients’ resuscitation preferences by physicians and nurses. The Western Journal of Medicine 150, no.6: 705–707.
Schneiderman, L.J., et al. 1993. Do physicians’ own preferences for life-sustaining treatment influence their perceptions of patients’ preferences? The Journal of Clinical Ethics 4, no. 1: 28–33.
Seckler, A.B., et al. 1991. Substituted judgment: how accurate are proxy predictions? Annals of Internal Medicine 115, no. 2: 92–98.
Houts, R.M., et al. 2002. Predicting elderly outpatients’ life-sustaining treatment preferences over time: the majority rules. Medical Decision Making 22, no. 1: 39–52.
Gerety, M.B., et al. 1993. Medical treatment preferences of nursing home residents: relationship to function and concordance with surrogate decision-makers. Journal of American Geriatrics Society 41, no.9: 953–960.
Pruchno, R.A., et al. 2005. Spouse as health care proxy for dialysis patients: whose preferences matter? Gerontologist 45, no. 6: 812–819.
Sulmasy, D.P., et al. 1998. The accuracy of substituted judgments in patients with terminal diagnoses. Annals of Internal Medicine 128, no. 8: 621–629.
Fagerlin, A., et al. 2001. Projection in surrogate decisions about life-sustaining medical treatments. Health Psychology 20, no. 3: 166–175.
President’s Commission for the Study of Ethical Problems in Medicine, Biomedical and Behavioral Research. 1983. Deciding to forego life-sustaining treatment. Washington D.C.: U.S. Government Printing Office.
Acknowledgments
We gratefully acknowledge support from The Vårdal Foundation and The Vårdal Institute, The Swedish Institute for Health Sciences. We would also like to thank two anonymous referees for several helpful comments.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Johansson, M., Broström, L. Turning failures into successes: a methodological shortcoming in empirical research on surrogate accuracy. Theor Med Bioeth 29, 17–26 (2008). https://doi.org/10.1007/s11017-008-9059-z
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11017-008-9059-z