Differences by Survey Language and Mode Among Chinese Respondents to a CAHPS Health Plan Survey

As efforts to measure, compare, and report patients' health care experiences expand in scope and importance, corresponding efforts have been underway to expand the reach of the underlying survey instruments to patients who prefer languages other than English. One challenge in the expansion of these surveys to such populations, when comparing to English-preferring populations, is that there may be differences in (a) actual health care experiences; (b) expectations, cultural norms, and preferences regarding health care; and (c) survey response tendencies that are a function of cultural differences and/or the language of survey administration. We consider these issues in a quantitative case study of a single Medicare-managed care plan with predominantly Chinese enrollees. Of 961 sampled members, 481 responded to bilingual mailings (323 responding in English, 158 in Chinese) and 170 responded to bilingual telephone follow-up (151 in Chinese). Unadjusted scores showed notable differences by language; adjusted analyses showed few language/mode differences for "objective" items, but dramatic differences remained for subjective rating items, casting doubt on their appropriateness for cross-cultural comparisons. We provide some evidence that previously observed general population mode effects for this survey (telephone positivity) may not exist in the Chinese language. Additionally, we find some evidence that previously observed negative response tendencies among Asians in assessments of health care relative to non-Hispanic whites may not be consistent across languages of response, with more positive Chinese than English subjective ratings.