Scaling-up Teen Pregnancy Prevention Programming: Examining Changes in Knowledge, Skills, and Beliefs

Introduction: Although US teen pregnancy rates have declined over the previous decade, rates in Tennessee consistently remain above the national average. Many of the current teen pregnancy evidence-based programs (EBPs) were originally tested in limited populations. The current study examined how an EBP first developed for inner-city, minority youth performed when scaled-up across rural, suburban, and urban Tennessee communities (n = 1,970 youth, 74.6% White). The goals of the study were to: 1) Examine outcomes related to sexual education knowledge, risk avoidance skills, and beliefs related to risk and protective factors; and 2) Identify characteristics of participants that are predictors of change in those domains.

Methods: Paired t-tests were used to assess whether changes in items measuring knowledge, skills, and beliefs were statistically significant. Pre- to posttest item change scores were calculated and added together to create a composite score for each domain. Multiple regression models were estimated to examine whether grade level, sex, race, and urbanicity were predictors of change.

Results: Pre- to posttest changes in items measuring knowledge were statistically significant (p <.001, d ranging from .37-1.06), with two items measuring knowledge of HIV transmission having particularly large effect sizes (d = .82 and 1.06). All items measuring skills were also statistically significant (p <.001), yet effect sizes were small to moderate (d = .24-.63). All but one of the items measuring beliefs were statistically significant (p <.001) but with small effect sizes (d = .27-.29). Middle school adolescents had significantly higher changes in knowledge (95% CI .79, 1.56), skills (95% CI .29, .76), and beliefs (95% CI .07, .44) compared to high school adolescents. White adolescents had significantly higher changes (95% CI .37, 1.66) in knowledge compared to Black adolescents. Adolescents residing in rural communities had significantly lower changes in knowledge (95% CI -1.52, -.36) compared to those in urban communities.

Conclusions: When scaling up an EBP and assessing fit, it is important to assess its outcomes and examine population predictors of change. While participants showed significant changes knowledge, skills, and beliefs, effect sizes indicate that participants benefited most from its content on HIV transmission. Middle school adolescents had the highest changes across all three domains, which points to the EBP being most appropriate for that age group. Identifying as White was a predictor of higher changes in knowledge, which suggests a promising fit with that population. However, further research is needed to assess whether the EBP is a proper fit for youth residing in rural communities due to its association with lower changes in knowledge.