Twenty-nine studies included in the review included four randomised controlled trials (RCTs), 19 controlled before and after studies and five interrupted time series. Reported sample sizes ranged from 17 to 4,166. All RCTs were judged to be high quality. Fifteen controlled before and after studies were judged to be good quality. All interrupted time series studies were judged to be poor quality.
Interventions to prevent adolescent pregnancy (seven studies): One high-quality RCT reported that a brief intervention consisting of one session of motivational interview was significantly associated with an increase in the number of participants who used effective contraception compared with controls (64% versus 48%, p=0.03). Controls were associated with a significantly higher risk of alcohol-induced pregnancy as measured by change in drinking habits (OR 2.9, 95% CI 1.49 to 5.45).
One good-quality controlled before and after study reported that using infant simulators to prevent adolescent pregnancy significantly increased participants' awareness that failure to use contraceptives increased risk of unplanned pregnancy (p<0.001). Another poor-quality controlled before and after study did not show a significant difference between the intervention and control groups after several weeks follow-up.
There was inconsistent evidence from three studies about the effectiveness of abstinence-based programmes in preventing teenage pregnancy. Evidence from one good quality controlled before and after study did not support use of a large multicomponent intervention to prevent teenage pregnancy.
Interventions to prevent repeat adolescent pregnancy (six studies): Three good-quality controlled before and after studies and one poor-quality interrupted time series reported that intensive case management interventions conducted by a culturally matched social worker were effective as part of multicomponent interventions (along with other components that included peer education) in preventing repeat pregnancy.
One poor-quality interrupted time series study showed that the daily presence of a public health nurse in school monthly (pregnancy testing, health counselling and referral and health education classes) was effective in preventing repeat pregnancy; the repeat adolescent pregnancy rate declined from 25% at the year before the intervention to a mean of 4.7% over the nine years of intervention duration. The results from one good-quality controlled before and after study were not reported.
Interventions of school-based health centers (six studies): Four studies (two good-quality controlled before and after studies, one poor-quality controlled before and after study and one poor-quality interrupted time series) supported the direct provision of contraceptives dispensed on site from school-based health centers on increasing contraceptive provision. However, two studies (one good-quality controlled before and after study and one poor-quality interrupted time series) showed no improvement in outcomes for school-based health centers that offered only health care assessments or counselling.
Further results were reported on other relevant interventions investigated by a small number of studies.