Using Evidence-Based Epidemiologic Indicators to Evaluate Impact of Abstinence Education Programs: For Example, Number Needed To Teach (NNTT)

Most evaluations about the effect of abstinence and sex education school-based programs report results in terms of statistically significant mean differences between treatment and control groups derived from Likert-type questionnaires. Recently a number of studies use indicators such as Cohen's D to examine effect magnitude. Proposed is the use evidence-based epidemiologic measures common in randomized clinical trials.

Pre-post surveys were administered to over 500 students and over 100 controls to examine the immediate impact of an abstinence until marriage 6—8 hour instructional high school program. Self-report questions were about knowledge, attitudes, behavioral intentions and sexual activity. Total and by race/ethnicity, age, and gender mean differences were examined with univariate and multivariate statistics. Additional measures were: Cohen's D, pre-post event rates, relative risk, relative risk reduction, attributable risk reduction, number needed to teach (NNTT)/prevent(NNTP)and relative benefit increase and relative risk increase.

Results were observed with respect to statistically significant pre-post mean differences for attitudes, knowledge, intentions and behaviors. Small (<10%) and some medium (>10% - <20%) effect sizes were noted. NNTT to change one additional student to choose to be sexually abstinent ranged from 7 to 25, dependent upon student demographics.

Epidemiologic indicators may be better suited to address the fundamental objectives of abstinence and sex education programs – specifically, demonstrate relative increase in the occurrence of primary and secondary abstinence from sexual activity and relative decrease in the occurrence of teen pregnancy and relative decrease in sexually transmitted infections.