Patient-doctor communication in cross-national perspective. A study in Mexico.

The authors assessed the cross-national replicability previously used to study medical encounters in the United States by adapting them to Mexico. The main research questions focused on information-giving, gender and social class differences in communication, and attention to socioemotional concerns in primary encounters. Sixty-two primary care encounters were audiotaped. Questionnaires were translated into Spanish, then translated back into English. Coding and transcription techniques were taught to Spanish-speaking researchers. Measures of communication were treated as dependent variables and were related by nonparametric statistical analyses to characteristics of physicians, patients, and clinical settings. Doctors in Mexico spent an average of 2.1 minutes (+/- 1.7 standard deviation [SD]), or 16.7% (+/- 10.7 SD) of total interaction time, in information-giving. Mexican doctors asked an average of 27.3 questions per encounter (+/- 18.0 SD), whereas patients asked an average of 1.5 questions (+/- 2.0 SD). Substantial interphysician variability was observed in total time of interaction (Kruskal-Wallis analysis of variance, chi-square = 27.2, P = 0.000), physician time in information giving (chi-square = 16.4, P = 0.022), and physician questions (chi-square = 36.7, P = 0.000). Patient characteristics associated with physician information-giving included male gender (chi-square = 4.1, P = 0.04) and age (Kendall's tau-b = .17, P = 0.05) but not education (tau-b = .08, P = 0.41). Information-giving in public clinics did not differ from that in private practices (chi-square = 0.0, P = 0.91). A bootstrap approach to multiple nonlinear regression permitted additional analysis of physicians', patients', and situational characteristics in explaining measures of patient-doctor communication; this analysis further demonstrated the importance of interphysician variability in communicative behavior. Previous methods for studying patient-doctor communication can be adapted and replicated in a non-English-speaking society. With certain exceptions, findings from Mexico were similar to those obtained in the United States.