We thank Kelly Roe for her thoughtful comments about our article, Primary Breast Cancer Decision-making Among Chinese American Women (Lee & Knobf, 2015b). We welcome the opportunity to clarify some points about our work and discuss issues of cultural differences in research about patients' decision-making. Questions were raised about the cultural adaption of the theoretical framework and the validation of the Decision Conflict Scale in Chinese populations.

We agree with Roe's view that cultural relevance of a theoretical framework is imperative in the application to other cultures or ethnic groups. In our work, the Ottawa decision support framework was used as a guide to selecting the key determinants in the treatment decision-making process that would address the effect on decision outcomes. On the basis of the expectancy value model, decisional conflict theory, and social support theory, this framework asserts that the determinants of decisions include knowledge, expectation, value, decisional conflict, perception of the important others, and resources to make the decision (O'Connor et al., 1998). Prior to this published study, Lee interviewed 10 healthy Chinese Americans about what they would prefer to do in the process of selecting a treatment if they had a cancer and who would make the decision (Lee & Dixon, 2003). The cancer treatment decision-making in these participants was characterized as searching for information and help, weighing the outcomes of cancer and treatment based on their beliefs and physician and family being involved. On the basis of the findings from that preliminary work and our review of the literature, we believe that the concepts in the Ottawa decision support framework are applicable for decision-making among Chinese and Chinese American populations.

However, we recognize that decision-making behaviors are embedded in culture and that cultural differences still exist such as decision-making for pain management (Chen, Tang, & Chen, 2012). Cultural variation may be manifest by value orientation or priority, social organization, communication, or environmental factors. The concept of decisional conflict that decision-making was a stressful process related to the uncertainty or ambiguity existing within choices (Janis & Mann, 1977) is relevant for the Chinese. Mann et al. (1998) confirmed that the construct of decision-making patterns or styles commonly used by decision-makers to deal with the stress that was generated by the conflict was similar across six countries (United States, Australia, New Zealand, Japan, Hong Kong, and Taiwan). However, a cross-cultural difference was found as the variation of frequency and strength of certain patterns. Asian subjects tended to score higher on avoidant or panicky patterns of decision-making and were more likely to share with or handover responsibility to significant others as compared with the Western subjects. Likewise, although many Chinese American women believed mastectomy was a safer, preferable treatment (Killoran & Moyer, 2006), Asian American women remained concerned about the body image or sexual consequence of breast surgery (Tam Ashing, Padilla, Tejero, & Kagawa-Singer, 2003). The authority of decision-making within the family and the value of maintaining the good for the whole family are viewed as the principles in determining who makes the decision for certain health-related decisions in Chinese. Nevertheless, studies in Hong Kong also showed that most of the Chinese women wanted shared decision-making and wanted to know their physician's preference of surgery (Lam, Fielding, Chan, Chow, & Ho, 2003). The shared decision-making among Chinese for healthcare may reflect a collaborative process between the patient, the family, and the physician.

In our research on surgical treatment decision-making in Chinese American women, we carefully assessed both specific cultural and universal components for instrument selection. In order to retain cultural sensitivity, we considered the social support and resource in Decisional Conflict Scale (DCS; O'Connor, 2010), the expectation of both physician and spouse or family keys in Stanton's Breast Cancer Decision-Making Questionnaire Post-Decision Version (Stanton et al., 1998). A self-designed profile to measure situational and clinical factors, such as patient-physician communication, immigration, insurance, and English literacy, was added to elicit possible environmental factors. We also conducted semistructured interviews to explore family involvement (Lee & Knobf, 2015a). All standardized instruments, including DCS, Breast Cancer Decision-Making Questionnaire, Decisional Regret Scale, and Decisional Satisfaction Scale were validated via a 4-point scale that was adopted from Lynn's content validity index before translation, and they also showed good convergent validity after being translated in Chinese (Lee, 2003). The content validity index was rated by a bilingual expert panel of two doctorally prepared Chinese nursing researchers and three postgraduate American Chinese healthcare providers with breast cancer to verify cultural relevance and inclusiveness of constructs and item importance.

Demonstration of factor invariance or factor similarity is as important as adopting a psychological or cognitive measurement into a culture that differs from its origin. However, issues of different factors extracted from studies of different cultures are common. It raises questions about whether the difference exists in the nature of the concept, the translation, the sampling and sample size, the decisional task, or the time. In our work, an exploratory factor analysis of this Chinese version DCS in this sample revealed that all 12 items (four subscales of uncertain, uninformed, value unclear, and unsupported) were loaded on one of three factors instead of the original four factors (Lee, 2003); it accounted for a total of 74.02% variance. One factor represented six items that were originally in the subscales of Uninformed and Value Unclear, and the rest of the factors consisted of items originally labeled as either the subscale of unsupported by others or uncertainty. People articulate their values based on their knowledge of options, risks, and benefits. It is not surprising that original items in an uninformed subscale and value-unclear subscales are loaded as correlates or as a latent factor in common-especially in this sample of women with limited English literacy and information. The similar finding was also reported in Lam et al.'s (2015) recent work. However, the unsupported subscale being able to successfully differentiate from other factors in our work, but failing to do so in Lam et al.'s (2015) work, might be an explication of environmental or situational differences related to the healthcare system and the norm of physician-patient relationship in the United States versus Hong Kong. Consistently, the physician views himself/herself as a partner in the shared decision-making for some women with breast cancer in Western society. But a partnership between the patient and the physician during the clinic visit is not common in most Asian societies but may exist for some Asian Americans. Further study in construct validity evaluation for DCS is warranted. Because of the potential differences in the decisional conflict construct for Chinese population, until more research is conducted, it is preferable to use the summated DCS score instead of the subscale score. Again, we are thankful for this opportunity to further clarify our thought. With this kind of dialogue, more perspectives can be recognized and discussed regarding this important topic.

REFERENCES