The purpose of this study was to examine whether there were racial differences in the designation of decision-makers for elderly people who were asked to consider how their medical care should be handled if they become incompetent. The sample included patients enrolled in a program to provide community-centered services to frail elderly people across the country. The dates of enrollment were not reported.

A substantial number of patients were eliminated from the analyses. Of 1193 patients eligible, 801 had complete data in their medical records (67%). Thus, selective losses and generalizability must be questioned.

First, there were significant racial differences in whether or not healthcare choices were noted in the medical records. Hispanics and Asians were less likely than non-Hispanic whites to have such notes. The difference between blacks and non-Hispanic whites was not statistically significant.

Second, there were significant racial differences in whether or not the patients expressed their own health care choices, as opposed to a surrogate: 91.8% of non-Hispanic whites, 86.3% of Hispanics, 82.9% of Asians, but only 66.7% of blacks expressed their own health care wishes. After controlling for age, education, gender, marital status, having children and mental status, whites were twice as likely as blacks and Asians were three times more likely than blacks to express their own health care choices. Additionally, whites were eight times more likely, Asians 12 times more likely, and blacks four times more likely than Hispanics to express their own health care needs

Finally, there were significant racial differences in who expressed the health care choices for patients. After adjusting for other variables, blacks were less likely to have a spouse as a surrogate but more likely to have a child, particularly a daughter, as a surrogate compared with whites. The same was true for Hispanic patients. In contrast, Asian patients were more likely to have a son as their surrogate. Only among blacks, and particularly among black men, was a relative other than a spouse or a child a likely surrogate.

The authors conclude that the findings that ethnic minorities are less likely to express their own health care choices indicate that this group may have a greater chance of having treatment choices selected that are at variance with those that would have been selected by the patient. The authors explain the ethnic differences as possibly reflecting cultural factors. "The method of decision-making, particularly the importance of shared decision-making, may be different between ethnic groups, based on the structure of the family and traditional gender roles and responsibilities." Religious beliefs might also contribute to this pattern. Finally, the structure of the family is a likely influence on who was chosen as a surrogate (daughters for Hispanics and black women, and sons for Asians), 'whereas black men, often absent from the nuclear family, may have to rely on other extended family members to be surrogate decision-makers."