Side effects from influenza vaccination: differences between returned and random surveys

Annually vaccinated Air National Guard members were surveyed by return of questionnaire in 1991 (20.5%) and in 1992 by random phone survey (23.5%). By returned questionnaire the sample reported any side effects at the rate of 48.1%; by randomly telephoned sample a significantly lower rate, 37.7% (p < 0.02), was reported. An inverse relationship existed between age and return of the questionnaire, present only in males, despite an increased rate of reporting of side effects in the younger ages. Females reported a significantly higher rate of side effects than males by return of questionnaire in all age groups and for the sample as a whole (p < 0.03). In the randomized survey the age effect disappeared, except for the tendency for the oldest group to report fewer side effects than the other age groups in their aggregate (p < 0.07). The higher reporting rate for females was less apparent by random survey but remained as a trend (p = 0.08; odds ratio 1.398; 95% confidence interval = 0.776-4.049).