Comparison of Three Interview Methods on Response Pattern to Sensitive and Non-Sensitive Questions

BACKGROUND To get more precise responses when gathering information about sensitive topics such as drug use, it is important to use the most optimal method. OBJECTIVES This study was carried out to address the impact of three interview methods (street-based, household, and telephone interviews) on response pattern to sensitive and non-sensitive questions in terms of participation, disclosure and discontinuing rates. PATIENTS AND METHODS We selected three culturally diverse major cities of Iran. Then, we randomly selected 300 subjects, 100 for each type of interview, from each major city (899 in total). For street-based interviews only pedestrians who were walking alone were recruited, for household interviews only one individual from each house participated (3-4 houses in each alley were selected), and for telephone interviews we selected phone numbers using a random number list. We asked five non-sensitive and five sensitive (related to drug use and sexual contact among their personal network) questions. RESULTS For telephone and household interviews, relative to street-based interviews, participants were less likely to disclose alcohol and drug-related behaviors (Adjusted OR = 0.76; 95% CI: 0.60- 0.97) and sexual behaviors among their network (Adjusted OR telephone/street-based = 0.64; 95% CI: 0.39- 1.07 and Adjusted OR household/ street-based = 0.56; 95% CI: 0.33- 0.95). We found that participants who were interviewed via the telephone were more likely (Adjusted OR = 1.24) and those who were interviewed at home were less likely (Adjusted OR = 0.86) to report non-sensitive information compared to participants who were interviewed on the street; however, these findings were not statistically significant. The largest participation rate and the least discontinuation rate were observed for household interviews. CONCLUSIONS It seems that the methods of interview effect response to both sensitive and non-sensitive questions. We believe that for street-based interviews, respondents may disclose more sensitive information than telephone and household interviews.