Two-year findings of an implementation intention intervention for teenage women show reduced consultations for emergency contraception or pregnancy testing and a trend towards reduced pregnancy rates
- Correspondence to Falko F Sniehotta
Newcastle University, Institute of Health and Society, Baddiley-Clark Building, Newcastle upon Tyne, NE2 4AX, UK;
Implications for practice and research
■ Previous attempts to reduce teenage pregnancy have rarely been theory-based and showed limited success.
■ Simple and inexpensive implementation intention interventions in family planning settings appear to have positive effects on consultations for emergency contraception, pregnancy testing and contraceptive supplies.
■ More research is needed to strengthen the evidence base for sustainable implementation intention effects and how this affects pregnancy rates.
Cost-effective, scalable and evidence-based strategies to reduce teenage pregnancy in the UK are currently not available.1 Family planning clinics support women to set goals for contraception (eg, taking the pill). While goal-setting defines the desired outcome (eg, taking the pill every day), implementation intentions are if-then plans linking suitable situations to specific acts (eg, ‘if I am brushing my teeth in the evening, then I will take the pill’). Implementation intentions are a theory-based behaviour change method with considerable evidence of effectiveness.2
Martin and colleagues conducted a single-centre parallel group randomised controlled trial with 261 teenage women attending a family planning clinic in England. Participants were randomised to a self-completed paper-and-pencil implementation intention intervention or a usual care control group. At the 9-month follow-up, findings showed that prompting young women to form implementation intentions specifying when, where and how to obtain or use their preferred contraceptive method and how to overcome barriers to enacting these plans was effective in reducing emergency contraception provision or pregnancy testing (38% intervention participants vs 55% controls).3 The follow-up period was then extended to 2 years to test the sustainability of intervention effects on pregnancy-risk behaviours (ie, measured as emergency contraception provision, pregnancy testing or a clinically verified pregnancy) and contraceptive supplies provision.
A significant association between group allocation and consultation outcomes over the 2-year follow-up was found: emergency contraception provision or pregnancy testing was lower for intervention participants (46%) than controls (62%), and the provision of contraceptive supplies was higher (54% vs 38%). Pregnancy rates were 7% in the intervention group and 12% in the control group, but this difference was not tested for significance.
This study addresses the pressing need for effective interventions to reduce teenage pregnancies suitable for widespread implementation in the UK. In line with previous reviews and current guidance for intervention development,1 4 this intervention was based on behaviour change theory2 and builds on evidence from behavioural science.
The trial's methodological quality (evaluated using the Cochrane risk of bias tool) was stronger than previous trials in this area,1 using appropriate randomisation methods, group allocation concealment, blinding of clinicians, independent outcome assessment and intention-to-treat analysis. We calculated a 0.3 Cohen's d effect size of the intervention at 2 years, which concurs with previous evidence,2 though is limited by the primary outcome's measurement issues (heterogeneous and not a direct measure of contraception behaviours). Nonetheless, the outcome is an objective measure of pregnancy-risk behaviours thus overcoming problems of social desirability and self-report bias.
The 2-year follow-up was not a prespecified outcome; ethical approval was obtained after identifying initial effectiveness at 9 months, making the long-term effects less conclusive. Although not sufficiently powered to detect differences in pregnancy rates and other individual consultation outcomes, reported trends suggest that a larger cluster-randomised trial is warranted to test effects on prespecified outcomes and generalisability to other clinics.
The findings suggesting 2-year sustainable effects are noteworthy as current evidence for enduring effects of implementation intentions is limited.2 Women were prompted to form action plans and coping plans5 (ie, plans when, where and how to use and/or source contraception and plans how to overcome barriers), therefore the findings are congruent with theory.
This was a relatively small, single-centre trial with a heterogeneous proxy outcome measure and ad hoc secondary outcomes for long-term effects; therefore, further research is needed to confirm effects and obtain conclusive findings for specific outcomes (eg, pregnancy rates). Nevertheless, implementation intentions are theory-based, the cognitive mechanisms through which if-then planning affects behaviour is well understood and there is considerable evidence for effectiveness in changing other health-related behaviours.2 5 The intervention is simple, inexpensive, congruent for use in current family planning clinics and unlikely to pose any risk. Based on these considerations, we recommend adopting implementation intentions into standard practice in family planning clinic settings while further confirmatory research is conducted.