RECRUITMENT AND RETENTION IN A CLINICAL-TRIAL FOR LOW-BIRTH-WEIGHT, PREMATURE-INFANTS

We report on recruiting and retaining a sample of low birth weight, premature infants for a clinical trial as well as results of tests evaluating sampling and retention biases. A total of 4551 infants were screened, and 1302 were found eligible. Consent was obtained for 1028 infants. After randomization and the presentation of group assignment, the number of infants enrolled was 985 (75.7% of those eligible). Of these, 92.7% completed the 3-year study. Tests to evaluate recruitment bias revealed significant relationships between nonenrollment and site, maternal race, and infant birth weight. Tests to evaluate retention bias revealed a significant relationship between dropout and maternal education. Additionally, infant birth weight and maternal age interacted with treatment in predicting dropout. Despite these statistically significant recruitment and retention biases, there was no evidence of problems with sample representativeness to the population of interest or of treatment group differences on study-relevant background variables.