Measurement of influenza vaccination status of the elderly by mailed questionnaire: response rate, validity and cost.

Response rate, validity and cost of measurement of influenza vaccination status by mailed questionnaire were studied among seniors enrolled in a group family practice. The response rate to the questionnaire was 73%, 85% and 90% after one, two and three mailings, respectively. Questionnaire responses regarding influenza vaccination were compared with vaccination status as determined by review of office records. For the immediately preceding vaccination period, agreement beyond chance (Kappa statistic) between questionnaire responses and office records was .88. When questionnaire responses were assessed against a "gold standard" of office records supplemented by telephone interview of apparent false positives, the questionnaire performed with a sensitivity of .93 (95% CI .87-.96) and specificity of .98 (95% CI .96-.99). As assessed by records review, questionnaire respondents were four times more likely to have received influenza vaccine than non-respondents (p less than .003, 2-tailed).