[Analysis of antibody response to measles vaccine with censored data]

ABSTRACT

Analysis of measles serum antibody responses is an important tool for evaluating the level of immunity in vaccinated people and for determining the factors which influence protective immunity. Usually antibody concentration values which are determined by quantitative assays are left censored, so standard analysis such as simple linear regression model may not be appropriate. Measles antibodies were measured using Enzyme Linked Immunosorbent Assay [ELISA] in 200 person aged 5-25 years before and after vaccination. A censored regression model proposed in order to reduce the effects of censored data in parameter estimating. Twenty two persons [11 percent] of the sample had titer below detection limit which made censored data. 83 percent of children and teenagers [less than 16 years] and 88 percent of participant aged 16 and more were protective against measles disease, but there was no statistically significance between age and immunity [P= 0.4]. Although the number of females who had protective antibodies against measles was less than males, they possessed an average titer 42 percent higher than males [without considering covariates]. This proportion changed into 39% by using the censored regression model and adjusting with respect to [antibody titer before vaccination]. Both ratios showed higher vaccine- induced protection in females [P< 0.05]