Ponsford et al.

Pilot and Feasibility Studies (2018) 4:102

https://doi.org/10.1186/s40814-018-0279-3

STUDY PROTOCOL Open Access

Study protocol for the optimisation,

feasibility testing and pilot cluster
randomised trial of Positive Choices: a
school-based social marketing intervention
to promote sexual health, prevent
unintended teenage pregnancies and
address health inequalities in England
Ruth Ponsford1* , Elizabeth Allen2, Rona Campbell3, Diana Elbourne2, Alison Hadley4, Maria Lohan5,
G. J. Melendez-Torres6, Catherine H. Mercer7, Steve Morris8, Honor Young6 and Chris Bonell1

Abstract
Background: Since the introduction of the Teenage Pregnancy Strategy (TPS), England’s under-18 conception rate
has fallen by 55%, but a continued focus on prevention is needed to maintain and accelerate progress. The teenage
birth rate remains higher in the UK than comparable Western European countries. Previous trials indicate that school-
based social marketing interventions are a promising approach to addressing teenage pregnancy and improving
sexual health. Such interventions are yet to be trialled in the UK. This study aims to optimise and establish the feasibility
and acceptability of one such intervention: Positive Choices.
Methods: Design: Optimisation, feasibility testing and pilot cluster randomised trial.
Interventions: The Positive Choices intervention comprises a student needs survey, a student/staff led School
Health Promotion Council (SHPC), a classroom curriculum for year nine students covering social and emotional
skills and sex education, student-led social marketing activities, parent information and a review of school sexual
health services.
Systematic optimisation of Positive Choices will be carried out with the National Children’s Bureau Sex Education
Forum (NCB SEF), one state secondary school in England and other youth and policy stakeholders.
Feasibility testing will involve the same state secondary school and will assess progression criteria to advance to
the pilot cluster RCT.
Pilot cluster RCT with integral process evaluation will involve six different state secondary schools (four
interventions and two controls) and will assess the feasibility and utility of progressing to a full effectiveness trial.
The following outcome measures will be trialled as part of the pilot:

(Continued on next page)

* Correspondence: [email protected]
1
London School of Hygiene and Tropical Medicine, 15-17 Tavistock Place,
London WC1H 9SH, UK
Full list of author information is available at the end of the article

© The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0
International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and
reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to
the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver
(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
Ponsford et al. Pilot and Feasibility Studies (2018) 4:102 Page 2 of 15

(Continued from previous page)

1. Self-reported pregnancy and unintended pregnancy (initiation of pregnancy for boys) and sexually
transmitted infections,
2. Age of sexual debut, number of sexual partners, use of contraception at first and last sex and non-volitional
sex
3. Educational attainment

The feasibility of linking administrative data on births and termination to self-report survey data to measure our
primary outcome (unintended teenage pregnancy) will also be tested.
Discussion: This will be the first UK-based pilot trial of a school-wide social marketing intervention to reduce
unintended teenage pregnancy and improve sexual health. If this study indicates feasibility and acceptability of
the optimised Positive Choices intervention in English secondary schools, plans will be initiated for a phase III trial
and economic evaluation of the intervention.
Trial registration: ISRCTN registry (ISCTN12524938. Registered 03/07/2017).
Keywords: Teenage pregnancy, Sexual health, SRE, RSE, School intervention, Adolescent, Cluster randomised trial

Background precluded meta-analysis but narrative synthesis con-

Teenage pregnancy and sexual health cluded this was a promising approach to addressing un-
Between 1998 and 2015, a period which included the intended teenage pregnancy.
1999–2010 implementation of the Labour government’s ‘Safer Choices’ and the ‘Children’s AIDS Society (CAS)
Teenage Pregnancy Strategy (TPS), England’s under-18 Carrera’ Program were two effective social marketing in-
conception rate has fallen by 55% [1]. Following the con- terventions identified in the above review. Safer Choices is
clusion of the strategy, a continued focus on prevention a school-based intervention involving a school health pro-
is needed to maintain progress and reduce disparities in motion council coordinating intervention activities, a
conception rates between different parts of the country. classroom-based sexual health curriculum, student-led so-
Data from 2017 indicate that there is a sixfold difference cial marketing campaigns and information for parents. A
in the under-18 conception rate between local author- randomised controlled trial (RCT) of this intervention in
ities (LAs) and 60% of LAs have at least one electoral the US reported reduced unprotected last sex and reduced
ward with a significantly higher rate than the average for numbers of partners with whom unprotected sex occurred
England [2]. The teenage birth rate remains higher in but did not measure effects on pregnancy [15–17]. The
England and Wales than comparable Western European ‘CAS Carrera’ Program is an after-school intervention pro-
countries [3]. Even after controlling for prior disadvan- viding careers, academic, arts, sports and life-skills ses-
tage, teenage pregnancy is associated with adverse med- sions and sexual health services. An RCT of this
ical, social, educational and economic outcomes for both intervention in New York City reported fewer pregnancies
mothers [4–6] and children [7, 8]. Teenage pregnancy is and delayed sexual debut among girls [18]. An attempted
a symptom of and contributes to the maintenance of replication trial in other US locations reported no such re-
health inequalities [9]. In 2006, it was estimated that ductions, reportedly due to poor fidelity [19]. The Gate-
teenage pregnancy cost the NHS £63 million per year house project, although not cited in the above review, is a
[10]. HIV and other sexually transmitted infections (STI) further intervention that adopts social marketing princi-
disproportionally affect young adults and cost the NHS ples and was found to be effective in postponing age of
large sums [11, 12]. However, effective prevention saves sexual debut. The Gatehouse project is a school-based
money. Return on investment analysis of the Teenage intervention which includes a student needs survey and
Pregnancy Strategy, for example, calculated that for classroom-based curriculum addressing social and emo-
every £1 spent, £4 was saved [13]. tional learning. Although primarily addressing mental
health, an RCT in Australian high schools reported partic-
Social marketing interventions to prevent teenage ipants’ increased age of sexual debut, but did not measure
pregnancy impacts on teenage pregnancy explicitly [20].
A recent systematic review of social marketing inter- School-based social marketing initiatives to prevent un-
ventions to reduce teenage pregnancy examined studies intended teenage pregnancy and promote sexual health
of interventions embracing social marketing elements have not been trialled in the UK. Our study aims to opti-
regardless of whether these were explicitly termed ‘so- mise, feasibility test and pilot trial in English secondary
cial marketing’ in their description [14]. Heterogeneity schools a whole-school social marketing intervention
Ponsford et al. Pilot and Feasibility Studies (2018) 4:102 Page 3 of 15

called ‘Positive Choices’. Positive Choices is informed by 6. What level of student reach does the intervention
selected components from the Safer Choices, CAS Carrera achieve?
and Gatehouse programmes and ‘whole-school’ ap- 7. What do qualitative data suggest in terms of
proaches to health improvement found to be effective in intervention mechanisms and refinements to
addressing a range of health risk behaviours, including programme theory and theory of change?
those related to teenage pregnancy and sexual health 8. How do contextual factors appear to influence
[21, 22]. The Positive Choices intervention involves implementation, receipt and mechanisms of action?
multiple components comprising a student needs survey, 9. Are any potential harms suggested and how might
a student/staff led School Health Promotion Council, a these be reduced?
classroom curriculum addressing social/emotional skills 10. What sexual health-related activities occur in and
and sex education, student-led social marketing, parent in- around control schools?
formation and a review of school sexual health services. 11. Are methods for economic evaluation in a phase III
The use of a student need survey to inform school coord- trial feasible
ination of the intervention by a student/staff council was
informed by the Gatehouse intervention. The use of a Methods
school health promotion council, classroom curriculum The study is 33 months long involving:
and student-led social marketing was informed by the
Safer Choices intervention. The intervention was in- 1. A facilitated, systematic optimisation of the Positive
formed by the CAS Carrera intervention not in terms of Choices intervention with the National Children’s
specific activities but more in terms of a focus on aiming Bureau Sex Education Forum (NCB SEF), a state
to promote sexual health via a focus on developing posi- secondary school and other youth and policy
tive skills and attitudes. In this study, our aim is not to as- stakeholders––(April 2017–March 2018).
sess the effects of Positive Choices, but to optimise the 2. A formative feasibility assessment of intervention
intervention with the National Children’s Bureau Sex Edu- components in one secondary school and
cation Forum (NCB SEF)––a voluntary sector organisa- subsequent intervention refinement––(September
tion that advocates for and provides resources to support 2017–August 2018).
the delivery of evidence and rights-based relationships 3. An external pilot cluster RCT across six schools
and sex education in England––one state secondary with integral process evaluation and study to assess
school and other stakeholders, feasibility test and refine the feasibility of economic evaluation—(May 2018–
the intervention in the optimisation school, and conduct a July 2019).
pilot RCT and process evaluation in six schools to assess
the feasibility and acceptability of the intervention and our Figure 1 below illustrates the overall timeline for the
trial methods. The study will determine the feasibility and study.
utility of conducting a phase III trial of intervention effect-
iveness and cost effectiveness. Our research questions are Intervention
outlined below. Intervention theory
The Positive Choices programme theory is informed by
social marketing principles, and has been developed with
Research questions experts in this field, addressing the ‘4Ps’ [23–25]. Posi-
tive Choices will ‘sell’ consumers a Product they want
1. Is it possible to optimise Positive Choices in (education on emotions and relationships) in an access-
collaboration with NCB SEF, a secondary school ible Place (school) at a low Price (free to students), with
and other stakeholders? Promotion to peers and parents (campaigns, parent in-
2. Is it feasible and acceptable to implement each formation) addressing competing influences from peers,
component of this intervention in the secondary media etc. [26]. Our survey component enables our
school involved in optimisation and what School Health Promotion Councils (with student in-
refinements are suggested? volvement) to tailor provision in each school to local
3. In light of a pilot RCT across six schools, is consumer priorities. The intervention theory of change
progression to a phase III trial justified in terms of (see Fig. 2) draws on those used in the interventions
pre-specified criteria? informing Positive Choices (Safer Choices, CAS Carrera,
4. Are secondary outcome and covariate measures and the Gatehouse project).
reliable and what refinements are suggested? Our theory of change will be further developed in the
5. With what rates are schools recruited to and optimisation phase informed by models of school
retained in the trial? change, [27] social influence [28] and social cognitive
Ponsford et al. Pilot and Feasibility Studies (2018) 4:102 Page 4 of 15

Fig. 1 Overall study timeline

Fig. 2 Logic model of Positive Choices

Ponsford et al. Pilot and Feasibility Studies (2018) 4:102 Page 5 of 15

theory [29] to address the following factors associated Informed by the needs-assessment data, School
with reduced risk of teenage pregnancy and STIs and Health Promotion Councils will select in what order
improved sexual health: sexual health knowledge, self- to deliver modules; whether to deliver within
efficacy, skills and competence; communication with personal, social and health education (PSHE); tutor
parents; school-wide social norms supporting positive groups or integrated into other lessons (e.g. English)
relationships/sexual health. Informed by the social devel- and whether to use our materials or existing
opment model [30], we anticipate that the intervention materials if these conform to our curriculum.
will also lead to positive aspirations and school engage- 4. Student-led social marketing facilitated by trained
ment, a further determinant of teenage pregnancy [31]. teachers and led by teams of 12–18 students per
The review of sexual health services will also improve school. Campaigns may use social and other media,
access to contraception. Although Positive Choices is a posters and events and will focus on healthy
universal intervention, by addressing determinants of relationships, sexual and human rights, delayed sex
teenage pregnancy, such as lack of positive aspirations and access to local services. Student social
and school engagement, that are socially stratified, the marketers will use data from the student needs
intervention is intended to reduce sexual health survey to segment the student population based on
inequalities. multiple characteristics such as existing knowledge
and attitudes to sexual health as well as cultural
Intervention components styles (e.g. hip-hop, skate) and peer group
Positive Choices is a manualised, whole-school social identifications (e.g. sporty boys, cool girls). The
marketing intervention, delivered for one English aca- student social marketers will use such information to
demic year (September–July) in the feasibility testing and design social marketing campaigns which address the
pilot trial phases. In a full trial, we expect the intervention most important topics among the groups who need
to be delivered over two consecutive academic years. The interventions most.
intervention is intended to build on and augment rather 5. Parent information—three newsletters, two
than entirely replace existing school sex education and homework assignments per year addressing parent-
sexual health provision. Some flexibility is built into the child communication.
programme so that schools can tailor intervention activ- 6. Consultancy on school sexual health services,
ities to the specific needs of their pupils. which will involve an audit of available sexual
The Positive Choices intervention comprises: health services in and around school, their
accessibility for young people and how they are
1. A student needs survey of year 8 students (aged promoted within schools.
12–13) which will be used to enable each
intervention components 3–6 below to be tailored In the feasibility assessment phase, different interven-
to local priorities in each school. tion components will be delivered in different terms of
2. A School Health Promotion Council (SHPC) which one academic year. In the pilot, intervention compo-
will comprise six staff/six students from different nents are implemented at the start of the academic year.
year groups who will review local needs data and In the feasibility testing and pilot trial phases, the cur-
use the data to tailor each intervention component riculum element will be targeted at year 9 pupils (aged
3–6 below to the school and will then coordinate in 13–14). In a full trial, it is envisaged that a classroom
school delivery of the intervention. curriculum would be delivered to both year 9 (aged 13–
3. A classroom curriculum that will address social/ 14) and year 10 (aged 14–15) pupils. Although the cur-
emotional skills (5-h class time per year) and sex riculum is targeted at particular year groups, the inter-
education (5-h class time per year) delivered by vention is a universal ‘whole-school’ intervention and as
school staff. The curriculum will be designed as a such has the potential for greater population-level im-
set of learning modules. Social and emotional skill pacts than targeted interventions [9] while minimising
modules will cover establishing respectful the risk of ‘positive deviancy training’, which can be an
relationships in the classroom and the wider school, issue in targeted interventions that bring together ‘at-
managing emotions, understanding and building risk’ individuals [32].
trusting relationships, exploring others’ needs and To enable the above intervention components, in the
avoiding conflict, and maintaining and repairing optimisation and pilot stage schools are provided with
relationships. Sexual health modules will cover the following inputs:
healthy relationships, negotiation and
communication skills, positive sexual health, sexual 1. A manual guiding each of the intervention
risk reduction, contraception and local services. components
Ponsford et al. Pilot and Feasibility Studies (2018) 4:102 Page 6 of 15

2. The resources to carry out the student survey and Pilot cluster randomised controlled trial
compile the student needs data The pilot RCT will involve six state secondary schools
3. Staff and student training in running a School (four interventions and two controls) in the south-east
Health Promotion Council (SHPC), staff training in of England.
delivering the social/emotional skills and sex
education curriculum, and in running social Study settings and population
marketing campaigns, all provided by NCB SEF Positive Choices will be delivered in one English state
4. Curriculum materials in social/emotional skills and secondary school in the feasibility phase of the study. In
sex education provided by NCB SEF the pilot, six English state secondary schools (including
5. Consultation on school sexual health services free schools and academies) will be included. Positive
provided by NCB SEF Choices is a universal intervention aimed at 11–16-year-
olds in participating secondary schools in England.
Intervention optimisation While the intervention will have effects at the whole-
The optimisation of the Positive Choices intervention will school level, the study population in the pilot will be stu-
be led by the research team and staff from NCB SEF as dents nearing the end of year 8 (age 12–13 years) at
well as the staff and students of one secondary school plus baseline who will receive the year 9 classroom curricu-
other youth and policy stakeholders. Optimisation of each lum. These pupils will be at the end of year 9 (age 13–
of the six intervention components (outlined above) will 14 years) at follow-up 12 months later (see below). In a
occur through a systematic process as follows: phase III trial, the intervention would target students in
years 9 and 10. The targeting of year 9 students in the
1. Review by researchers and NCB SEF staff of feasibility assessment and pilot RCT phases reflects the
existing systematic reviews and the evaluations of truncated timescales of these phases. Years 9 and 10
and, where appropriate, intervention materials from would be targeted in a phase III RCT because proximal
the Safer Choices, CAS Carrera and Gatehouse risk factors are manifesting [33], prevention is not too
interventions late and sex education is acceptable [34, 35]. Prior stake-
2. Drafting of intervention materials by NCB SEF and holder engagement suggests provision to year 11 stu-
the research team dents is unfeasible because of GCSE exam preparation.
3. Consultation with staff and students from the
optimisation school, as well as the other Inclusion and exclusion criteria
stakeholders We will collaborate with one secondary school in the
4. Refinement of intervention approaches and optimisation and feasibility assessment phases. This
materials school will be purposively selected based on location in
south-east England and having a higher than median
Feasibility assessment and intervention refinement local index of multiple deprivation and value-added
The intervention components will then be implemented GCSE attainment to reflect high need but high capacity
and assessed for feasibility and acceptability in the school to participate in optimisation and refinement.
involved in optimisation (see ‘Process evaluation’ below). For the pilot, all state secondary schools (including free
This will occur over one school year in phases as follows: schools and academies) in south-east England will be eli-
gible. Private schools, pupil referral units (PRUs) or
1. Term 1 (September–December 2017): schools for those with special educational learning needs
implementation of student needs survey, staff or disabilities and boys’ (but not girls’) schools will be ex-
training and School Health Promotion Council cluded from the pilot and full trial since our primary out-
2. Term 2 (January 2018–March 2018): come focuses on unintended pregnancies among girls.
implementation of student curriculum No students in participating schools will be excluded
3. Term 3 (April 2018–July 2018): implementation of from our study. Those with mild learning difficulties or
student-led social marketing and consultancy re- poor English will be supported to by fieldworkers to partici-
garding school sexual health services pate in baseline and follow-up surveys (described below).

Intervention components will be assessed by the re- Recruitment

search team as they are implemented in order to inform The optimisation and feasibility assessment phase of the
further refinements of the intervention components project will involve one purposively sampled secondary
led by NCB SEF (Fig. 3 provides a timeline of the school (see above for criteria) with no random allocation.
feasibility assessment and intervention refinement This school will be recruited via our existing contacts to
phase of the study). ensure the school has the capacity to participate.
Ponsford et al. Pilot and Feasibility Studies (2018) 4:102 Page 7 of 15

Fig. 3 Timeline for feasibility assessment and intervention refinement phase

In the pilot RCT phase, six schools across south-east 1. Materials for the training, school health promotion
England will be recruited (purposively varying by local council, social marketing meetings, student
deprivation and school-level GCSE attainment). Schools curriculum and consultancy on school sexual health
will be recruited to the pilot RCT by a combination of services are optimised in line with the theory of
mail outs, phone calls and prior networks including the change and to the satisfaction expressed in writing
UCL Partners School Health and Wellbeing Research of the research team, NCB SEF, the participating
Network. Response rates will be recorded, as will any secondary school and the study steering committee.
stated reasons for non-participation. 2. According to audio-recordings, provider diaries and
researcher observations, the training, school health
promotion council, social marketing meetings,
Randomisation student curriculum and consultancy on school
For the pilot phase, following the baseline survey with sexual health service components are implemented
students at the end of year 8 (approximately 180 per with 70% + fidelity in the participating school.
school), schools will be randomly allocated to interven- 3. Interviews with students and staff conducted as part
tion/control groups remotely by the Clinical Trials Unit of the process evaluation (described below) indicate
(CTU) at the London School of Hygiene and Tropical that the intervention is acceptable to at least 70% of
Medicine (LSHTM). In the pilot, allocation will be 2:1 students and staff involved in implementation.
favouring the intervention.
Assessment of optimised materials will be based on
Comparators whether stakeholders agree that intervention materials
In the pilot RCT phase, two schools will be randomised are consistent with the theory of change, meet the speci-
to the control group, will not receive the intervention fication laid out in the protocol and are regarded as con-
but will continue with any existing sexual health-related textually appropriate for piloting in English secondary
provision, which will be examined in our process evalu- schools. Materials will be assessed by the research team
ation, as will sexual health services in the surrounding and discussed in a minuted investigator meeting. The
area. Retention of control schools will be maximised via views of the school on the materials will be assessed as
£500 payment and feedback of survey data after trial part of the process evaluation via interviews with four
analysis. school staff and eight year 9 students. The materials, the
minutes of the investigator meeting and the findings
from the above interviews will then be discussed within
Endpoints of the study a minuted SSC meeting which will make an overall rec-
Optimisation and feasibility assessment phases ommendation about whether interventions are accept-
Outcomes for the optimisation and feasibility assessment able or not and any amendments that are required.
phases will be meeting criteria for progression to a pilot Fidelity will be assessed quantitatively against tick-box
RCT. Progression criteria comprise the following: quality metrics which will form an integral part of each
Ponsford et al. Pilot and Feasibility Studies (2018) 4:102 Page 8 of 15

intervention component. For example, each training and and Share trials [34, 35, 37] will be assessed via self-
curriculum session will be assessed against session- complete paper and pen questionnaire:
specific quality metrics relating to the topics covered,
the exercises used and opportunities for discussion; 1. Self-reported pregnancy and unintended pregnancy
meetings will be assessed against meeting-specific qual- (initiation of pregnancy for boys) and STIs
ity metrics relating to the agenda items covered, oppor- 2. Age of sexual debut, number of sexual partners, use
tunities for discussion and the actions agreed; and of contraception at first and last sex and non-
consultancy on school sexual health services will be volitional sex
assessed against quality metrics concerning the review of 3. Educational attainment (which is a plausible and,
existing services and action taken to enhance these. The for scale up, critical outcome of our intervention).
investigators will agree a set of metrics to assess fidelity
and acceptability with the Study Steering Committee The full trial will conduct exploratory analyses to
(SSC) in the early stages of the project, prior to carrying examine how effects on the above outcomes are moder-
out fieldwork. ated by SES, gender, ethnicity and baseline risk to assess
intervention impact on health inequalities. These ana-
Pilot RCT lyses will be tested as part of the pilot.
As already outlined, the pilot RCT will not aim to assess Informed by our theory of change [27–30], we will also
intervention effects, but to assess the feasibility and ac- conduct exploratory analyses to examine the following
ceptability of implementing the intervention in state sec- mediators [38] using existing measures [34, 35, 39–41]:
ondary schools in England. Pilot primary outcomes will
be meeting criteria for progression to a phase III trial 1. School-level social norms supportive of positive
comprising: relationships and sexual health
2. Individual-level sexual health knowledge and skills,
1. The intervention is implemented with 70% fidelity contraceptive skills and access, self-efficacy, sexual
in ≥ 3 of four intervention schools. competence, communication with parents, school
2. Process evaluation indicates that the intervention is engagement and career/educational aspirations
acceptable to 70% of students and staff involved in
implementation. All of the above measures will be assessed for reliability
3. Randomisation occurs and ≥ 5 of six schools accept in our pilot. We will assess reliability by reporting an intra-
randomisation and continue within the study. cluster correlation coefficients (ICC) to examine the
4. Student questionnaire follow-up rates are ≥ 80% in consistency of measures from baseline to follow-up among
≥ 5 of SIX schools. the control group (this is a different ICC to that measuring
5. Linkage of self-report and routine administrative clustering within schools) and Cronbach’s alpha statistics
data on pregnancies is feasible. at baseline and follow-up for scaled outcomes. See Eco-
nomic evaluation section below for economic outcomes.
Measures to be tested as part of the pilot trial
Outcome measures and analyses that would be used in a Assessment and follow-up
phase III RCT will be tested in the pilot. Our current Feasibility assessment phase
intention is that the primary outcome for a phase III A baseline needs survey of students approaching the end of
trial will examine unintended teenage pregnancies via year 8 will be undertaken. Other data collected in the feasi-
routine data on births and terminations assessed at bility phase are described below under ‘Process evaluation’.
48 months (age 16/17). Around half of under-18 concep-
tions in England and Wales end in abortion [1]. Yet, while Pilot RCT
recognising that a proportion of teenage pregnancies will In the pilot RCT, baseline surveys will be carried out be-
be intended [36], we anticipate that an outcome measure fore randomisation as students near the end of year 8
that takes into account both conceptions and terminations (age 12/13) in June 2018 and will collect data on pre-
will provide a better indication of unintended pregnancies hypothesised outcome variables, covariates and modera-
than a measure purely focused on terminations. This is tors, drawing on existing survey items as outlined above.
because many unintended pregnancies will not result in Paper questionnaires will be completed confidentially in
termination, and rates of termination will be strongly af- classrooms supervised by fieldworkers, with teachers
fected by the availability of local services. remaining at the front of the class to maintain quiet and
We will also assess secondary outcomes which would order, but unable to see student responses. We will sur-
be assessed in a phase III RCT at 24 months. The follow- vey absent students by leaving questionnaires and
ing secondary outcome measures drawn from the Ripple stamped addressed envelopes with schools.
Ponsford et al. Pilot and Feasibility Studies (2018) 4:102 Page 9 of 15

We will resurvey students at 12 months (June 2019) as findings. Qualitative research will involve purposive sam-
students near the end of year 9 (age 13/14) and will col- ples aiming to encompass diversity on key criteria.
lect self-report data on experiences of the intervention,
outcomes and pre-hypothesised potential mediators. Protection against bias
Fieldworkers will be blind to allocation. Based on past Although the aim of this study is to optimise the inter-
experience, [37, 42] in the pilot, we expect 95% baseline vention, assess feasibility and then pilot outcome mea-
survey participation and 90% at follow-up. In the pilot, sures and analyses rather than estimate intervention
data on terminations and births at 18 months will be ob- effects, we will pilot methods aimed at minimising bias.
tained in collaboration with the Office for National Sta- The investigator team and the intervention delivery team
tistics (ONS) and the Department of Health (DH) by will be separately managed. In the pilot RCT, outcome
linking data on female trial participants via the national data will be collected and analysed blind to allocation,
pupil database and other identifiers to routine ONS data and we will examine effects adjusting for potential base-
on registration of births and statutory termination notifi- line confounders (age, gender, SES and ethnicity). We
cations, by staff blind to allocation. Linkage of birth and will aim to maximise response rates at each pilot RCT
termination data has been previously conducted for ob- site at baseline and follow-up to minimise non-response
servational studies [33] but has not involved linkage to and attrition bias, for example, following up those indi-
survey data, and hence has not examined participant viduals not present during survey sessions. Response
consent rates, which will be a key focus for our research. rates and qualitative data will be analysed to refine data
Initial discussion with ONS has established that data collection methods prior to a phase III trial examining
linkage is feasible despite the limited identifiers attached effectiveness. Blinding of participants to allocation is not
to termination records and is consistent with DH guid- possible.
ance and data protection law.
Process evaluation
Analytic sample and proposed sample size Integral process evaluation informed by existing frame-
Feasibility assessment will be carried out in one London works [44–46] has three purposes:
school. Involving only one school at this stage enables
intensive engagement while assessing that intervention 1. To examine intervention feasibility, fidelity, reach
activities are in principle feasible. The survey will involve and acceptability in the feasibility and pilot RCT
approximately 180 year-9 students but note that this is phases
undertaken to assess the feasibility of this as an inter- 2. To assess provision in control schools and potential
vention component only. Evaluation activities (see below contamination in the pilot RCT
for details) in this feasibility phase will involve small 3. To explore context and potential mechanisms of
samples intended to contribute to assessing feasibility in action in the pilot RCT phase, including potential
principle rather than providing statistically representa- unintended effects, in order to refine the
tive findings. Qualitative research will involve purposive intervention theory of change and design
samples aiming to encompass diversity on key criteria.
Six schools varying by local level of deprivation and Feasibility assessment phase
school-level GCSE attainment will be recruited to the The feasibility assessment phase will assess the ‘progres-
pilot. No power calculation for this phase has been con- sion criteria’ to advance to the pilot RCT phase (outlined
ducted since the aim of this phase is to assess progres- above). Data will be collected via audio-recording of
sion criteria in a pragmatic, relatively small but NCB SEF training for school staff; surveys of school staff
purposively diverse sample of schools prior to a future trained by NCB SEF; diaries (including time logbooks) of
phase III RCT which would involve a larger sample in school staff implementing School Health Promotion
order to examine intervention effects informed by a for- Councils, curriculum and social marketing meetings;
mal power calculation. Approximately 1080 students will structured observations of two sessions of School Health
be surveyed in the pilot at the end of year 8 (age 12/13) Promotion Councils, curriculum lessons and social mar-
at baseline and followed-up 12 months later. While no keting meetings; and individual or group interviews with
power calculation for this phase has been undertaken, four NCB SEF staff and four school staff (purposive by
previous similar studies suggest this sample will be suffi- role/seniority) and eight year 9 students (purposive by
cient in order to assess student response rates, the reliabil- gender and SES).
ity of measures and intervention reach [42, 43]. Evaluation
activities (see below for details) in this feasibility phase will Pilot RCT phase
involve samples intended to contribute to assessing local In addition to assessing the ‘progression criteria’ relating to
feasibility rather than providing statistically representative intervention feasibility and acceptability (outlined above),
Ponsford et al. Pilot and Feasibility Studies (2018) 4:102 Page 10 of 15

we will also examine reach via qualitative research as well staff time, training events/workshops and consumables.
as questionnaire survey items at follow-up. The informa- Measures will include standardised sessional checklists
tion collected on socio-demographic, educational and to monitor and document attendance, preparation and
neighbourhood characteristics in the student surveys will delivery time for key training events, School Health Pro-
also allow us to examine reach according to these mea- motion Councils, student-led social marketing meetings
sures and how this varies by institutional setting. We will and the review of school sexual health services; the com-
also assess the fidelity, reach and perceived impacts of staff pletion of surveys and diaries by school staff charged
training activities. Data will be collected via audio- with intervention delivery, assessing time spent on tasks
recording of NCB SEF training for school staff; surveys of relating to intervention, staff travel and other expenses
school staff trained by NCB SEF; diaries (including time relating to the intervention charged to a specific project
logbooks) of school staff implementing School Health Pro- grant code.
motion Councils, curriculum and social marketing meet- The Child Health Utility (CHU) 9D measure [47] will
ings; and structured observations of randomly selected be used to assess student’s health-related quality of life
session per school of School Health Promotion Councils, as part of the economic evaluation. The CHU-9 is a vali-
curriculum lessons and social marketing meetings. Individ- dated age-appropriate measure that was explicitly devel-
ual or group interviews with two trainers, four staff per oped using children’s input and has been suggested to be
intervention school (purposive by seniority/activity in- more appropriate and function better than other health
volved), and 8 × year 9 students per intervention school, utility measures for children and adolescents. Student
(purposive by involvement, risk status and gender), will utility values will be collected (at baseline and at follow-
also be conducted. up surveys at 24 and 36 months in a full RCT) using the
CHU-9D and by converting the SF-12 questionnaires,
Provision in control schools and potential contamination respectively. It is anticipated that these measures would
We will examine sexual health provision in and around be used in a phase III trial to measure short-term impact
control schools in order to describe our comparator. We on health-related quality of life.
will examine the potential for contamination across arms
to ensure this is not a threat to internal validity in a Analyses
phase III trial. Data will be collected via student surveys, Feasibility assessment
interviews with two staff per control school (purposive Our analysis in this phase will determine whether the
by seniority) and 4 × year-9 students (purposive by gen- study should proceed to the pilot RCT phase. Descrip-
der and SES) per control school. tive statistics on fidelity will draw on audio-recordings of
training, diaries of providers and structured observations
Context and mechanisms of action of intervention activities. Analysis of acceptability will
In addition to piloting intermediate outcome variables re- draw on interviews with staff and students. Findings will
quired for mediator analyses in a subsequent phase III be fed back to NCB SEF staff who will be responsible for
RCT, we will use rich, contextual qualitative data and ana- refining the intervention ready for implementation in
lyse these data in order to explore potential mechanisms of the pilot RCT.
action and thus refine our theory of change. These qualita-
tive analyses will also examine how mechanisms may vary Pilot RCT
with context, students’ socio-demographic characteristics Our main analyses will determine whether criteria for
and/or other factors, in order to refine and optimise the in- progression to a phase III trial are met. Descriptive sta-
tervention’s theory of change. We will also analyse qualita- tistics on fidelity will draw on audio-recordings of train-
tive data to explore any mechanisms that might give rise to ing, diaries of providers and structured observations of
unintended, potentially harmful consequences. intervention activities. Statistics on acceptability will
draw on surveys of students and trained staff and inter-
Economic evaluation views with staff and students. School randomisation and
The pilot RCT will examine whether it is feasible to as- retention, and student follow-up will be described using
sess cost-effectiveness using a cost-consequence analysis a CONSORT diagram [48]. We will assess the precision
within a phase III trial. Within the pilot, study methods of data linkage in association with ONS researchers.
to measure the incremental cost of the intervention in a Additional analyses will address our other research
phase III trial study will be developed and piloted. With questions. Descriptive summaries of baseline and follow-
the use of a broad public and third sector perspective, up data by arm will be tabulated. We will assess the reli-
resources to be measured will include resources used by ability of secondary outcome measures by reporting
NCB SEF, schools and the NHS. Within this, key inter- intra-cluster correlation coefficients (ICC) to examine
ventional resources will include NCB SEF and school the consistency of measures from baseline to follow-up
Ponsford et al. Pilot and Feasibility Studies (2018) 4:102 Page 11 of 15

(this is different from the use of ICC to measure cluster- well as the circumstances in which we would need to
ing of variables within schools) and Cronbach’s alpha breach confidentiality (see ‘Safeguarding’ below).
statistics at baseline and follow-up for scaled outcomes. The research will also involve the piloting of the linkage
We will pilot intention-to-treat analyses of outcomes of student survey data to administrative data on births
[44] and moderator analyses (how effects vary by SES, and terminations by the Office for National Statistics. Sur-
gender, ethnicity and baseline risk). vey participants will be informed of this process as part of
Qualitative data will be subject to thematic content ana- consent procedures and their consent to it sought.
lysis (in vivo/axial codes; constant comparison [49] in- In addition, students’ parents will be contacted by let-
formed by realist approaches to evaluation [50] and May’s ter 1 week prior to any specific research fieldwork
implementation theory [46] to examine potential mecha- informing them about this and providing them with the
nisms of action and of harm, determine how contextual option of withdrawing (opting out) their child by con-
factors influence implementation and mechanisms, de- tacting the school or the research team. As is normal
scribe relevant activities in and around intervention and within public health and educational research involving
control schools and refine our programme theory and the- secondary school students in the UK, we will not seek
ory of change. opt-in consent from student participants’ parents.
Our economic feasibility study will pilot collection of
quality of life and assess the feasibility of methods to be Safeguarding
used within a full trial, which in line with NICE guid- We will develop and maintain standard operating proce-
ance, would involve a wider cost consequence analysis, dures for dealing with safeguarding concerns and report-
comparing intervention costs with the full range of study ing serious adverse events. In collaboration with the
outcomes. National Children’s Bureau, we will develop a priori cat-
egories of abuse reported through the research that ne-
Ethical issues cessitate our breaching confidentiality to ensure
Ethical approval for the study was obtained from the individuals are offered care and protection. These cri-
London School of Hygiene and Tropical Medicine Ethics teria will be established so that we balance our ethical
Committee on 21 March 2017 (Ref. 11927). duty of promoting participant autonomy by respecting
confidentiality and promoting participant wellbeing
Informed consent when we determine that we need to breach confidential-
Head teachers as gatekeepers will be asked for informed ity to address abuse that appears to be serious and on-
consent for intervention and random allocation, as is going. Where such abuse is reported through a
standard practice in cluster randomised trials in schools questionnaire, we will contact the safeguarding lead in
[42]. As is normal within public health and educational re- the school. Where the report occurs directly to research
search in secondary schools in the UK (e.g. [34, 35, 37]) staff, we will first discuss the need for a response with
informed written opt-in consent will be sought from all the research participant prior to contacting the school
research participants, including students, judged compe- safeguarding lead.
tent to provide this (a model information sheet and con- Qualitative research (interviews, focus groups, obser-
sent form is provided at Additional file 1). In all cases of vations) will not ask staff or students about their experi-
data collection including surveys, interviews and focus ence of sex. However, if participants nonetheless
groups, observations and audio-recordings, except where describe any sexual abuse, or otherwise become upset in
practically impossible, participants will be given an infor- any way, our researchers will be trained in how to re-
mation sheet several days before data collection. Just be- spond. In the case of focus groups, researchers will be
fore data collection participants will also receive an oral trained to ensure that discussions do not move in the
description of the study and have the opportunity to ask direction of personal disclosures of sexual behaviour since
questions. Participants will then be advised that participa- this is not the purpose of the groups and it would be very
tion is voluntary, and they may withdraw at any point. All difficult to ensure that all focus group participants did not
participants will be advised that they are free to withhold talk about such disclosures outside the group. Staff will be
consent, and this matter will not be fed back to teachers trained to identify the potential for such disclosures, work
or, in the case of staff participants, their managers. Stu- to avoid them but then to approach participants immedi-
dents opting not to participate in surveys will be offered ately after the focus group to offer support and to assess
alternative activities in the classroom. Those opting out of whether any other response is needed, using the same
other data collection will be free to continue with their procedures as described above.
normal activities. All participants, including students, will Any member of the research/fieldwork team visiting a
be informed in consent materials of the confidentiality school will be required to have a full Disclosure and Bar-
with which the information they provide will be treated as ring Services (DBS) check. All work will be carried out
Ponsford et al. Pilot and Feasibility Studies (2018) 4:102 Page 12 of 15

in accordance with guidelines laid down by the Eco- We also consulted with five members of the ALPHA
nomic and Social Research Council (ESRC), the Data (Advice Leading to Public Health Advancement) youth
Protection Act 1998 and the latest Directive on GCP group based at the DECIPHer Centre, Cardiff University,
(2005/28/EC). on 29 October 2014. Participants were enthusiastic
The trial steering group (which because this is a pilot about the intervention, supported this starting in year 9,
not a phase III RCT will undertake data monitoring and very supportive of school-based sexual health services
ethics duties) and the London School of Hygiene and and felt that targeting would be problematic. Although
Tropical Medicine ethics committee will be provided with some components are already being delivered in some
anonymised reports of all disclosures of serious abuse and schools, none use a coherent programme informed by
any other serious adverse events. These will categorised by social marketing principles.
type, circumstances and the extent of any possible con- Ongoing consultation with NCB SEF and one secondary
nection with intervention or research activities. school in London is inbuilt into the optimisation phase of
In each school and within NCB a senior member of the study. Policy stakeholder’s events and consultation
staff will be identified who is not directly involved with with the ALPHA youth advisory group will continue to
the intervention and whom staff or students may go to if take place throughout the duration of the project.
they have complaints about any elements of the research
study. This will be communicated to students outside of Expected output of research/impact
the research process to increase trust that this is truly As well as reporting in the NIHR Public Health Research
independent. journal, we would submit two open-access papers to high-
Quantitative and qualitative data will be managed by impact journals reporting our key findings regarding (1)
project staff using secure data management systems and process evaluation of integrated social marketing strategy
stored anonymously using participant identification and (2) student/staff experiences of the intervention. We
numbers. Quantitative data will be managed by London will present our findings at two international conferences
School of Hygiene and Tropical Medicine accredited (Society of Prevention Research and International Associ-
clinical trials unit (CTU). Where collected, participant ation for Adolescent Health) in 2019, as well as national
identification numbers and corresponding participant conferences. We will disseminate the results to participat-
names will be held in separate files; these files will be ing schools, to the ALPHA youth group based at DE-
password-protected folders. The names used in qualita- CIPHer and to schools in the Institute of Education/
tive data will be replaced with pseudonyms in interview/ UCLPartners School Health and Wellbeing Research Net-
focus group transcripts. In reporting the results of the work and Healthy Schools London network, both of which
process evaluation, care will be taken to use quotations we are already heavily involved in. We will draft an article
which do not reveal the identity of respondents. for the Times Education Supplement about the research.
In line with MRC guidance on personal information in The research team will also use blog-posts and Twitter to
medical research, we will retain all research data for increase public awareness of the study. Knowledge ex-
20 years after the end of the study. This is to allow sec- change is built into the proposed work from the outset via
ondary analyses and further research to take place and the stakeholder group. We will present emerging findings
to allow any queries or concerns about the conduct of at two meetings with policy stakeholders, including policy
the study to be addressed. In order to maintain the ac- officials and public health commissioners in the UK na-
cessibility of the data the files will be refreshed annually tions. Two policy and practice dissemination events will be
and upgraded if required. held: one seminar in partnership with Public Health Eng-
land and one at the Association for Young People’s Health.
Public and patient involvement (PPI) The most important scientific outputs generated by
In preparing this protocol, we have collaborated with this project will be increased knowledge about the feasi-
staff and students from five schools involved in UCL bility and acceptability of delivering and trialling an
Partners Schools Health Research Network (co-directed intervention which uses social marketing strategies and
by CB) via consultations in September–October 2014. is informed by existing effective interventions to prevent
These informed our decisions to focus our year 9 cur- unintended teenage pregnancies. This will inform the
riculum on social/emotional skills and our year 10 cur- development of a subsequent proposal for a phase III ef-
riculum on sexual health and contraception/protection, fectiveness trial.
include a focus on school sexual health services, ensure
student-led social marketing embraces social media, use Research governance
interviews where appropriate in our process evaluation Trial registration and conduct
and interview students as well as staff in control schools The trial is registered with www.controlled-trials.com
to assess usual provision. (ISCTN 12524938). As the trial does not take place
Ponsford et al. Pilot and Feasibility Studies (2018) 4:102 Page 13 of 15

within clinical settings nor using clinical samples nor effective, Positive Choices has the potential to make sig-
using a medicinal product, there is no requirement to nificant population-level health improvements.
comply with the ‘The Medicines for Human Use (Clinical
Trials) Regulations 2004’. We will follow the UK Medical Trial status
Research Council (MRC) guidelines on good clinical prac- At the time of submission (November 2017), the opti-
tice for clinical trials. misation school has been recruited, the needs survey has
The London School of Hygiene & Tropical Medicine been carried out in this school, the first set of interven-
will act as the main sponsor for this study. Delegated re- tion materials have been optimised and implementation
sponsibilities will be assigned locally. of the intervention for the feasibility assessment phase
The principal investigator (CB) will have overall re- has begun. The research team is in the process of
sponsibility for the conduct of the study. The day-to-day recruiting schools for the pilot trial.
management of the trial will be coordinated by the trial
manager (RP) based at LSHTM. The following govern- Additional file
ance structures will be instituted:
Additional file 1: Consent form for ‘Positive Choices’ student
1. Trial executive group (TEG): the PI (CB) will chair questionnaire. (PDF 140 kb)
weekly TEG meetings with the trial manager (RP),
statistician (EA) and, where appropriate, NCB SEF, Funding
CTU and fieldwork staff. Funding for this trial has been provided by the UK National Institute for
Health Research (NIHR) Public Health Research programme (grant number
2. Trial investigators’ group (TIG): CB will also chair a 14/184/02). The views expressed in this paper are those of the authors and
TIG which will include all co-investigators and not necessarily those of the NHS, the NIHR, MRC, CCF, NETSCC, the Public
members of the TEG; the TIG will meet monthly Health Research programme or the Department of Health. This funding
source had no role in the design of this study and will not have any role
during the early stages of the research (months during its execution, analyses, interpretation of the data or decision to
1–6) and then every 3 months thereafter. submit results.
3. Study steering committee (SSC): an independent
SSC will be established and meet three times Authors’ contributions
CB conceived of the study and led study design, the overall analysis plan
throughout the life of the project to advise on the and funding applications. RP drafted the paper and will be responsible for
conduct and progress of the trial, and relevant management of the trial and process evaluation. EA, DE and CM contributed
practice and policy issues. Because this is a pilot to study design and led on the plan for the management and analysis of
quantitative data. HY contributed to overall study design and advised on
not a phase III RCT, the SSC will undertake data public engagement with young people. AF and GJMT led on the design
monitoring and ethics duties and be informed of of the process evaluation and advised on public engagement with young
any serious adverse events as described under people. SM designed the economic evaluation. AH contributed to study
design and advised on public engagement with policy stakeholders. ML
‘ethics’ above. and RC contributed to intervention and study design. All authors read and
approved the final manuscript.
The project will employ standardised research proto-
cols and pre-specified progression criteria, which will be Ethics approval and consent to participate
This study was approved by the London School of Hygiene and Tropical
agreed and monitored by the TIG and SSC. Medicine (LSHTM) on 21 March 2017, Ref. 11927. Informed written opt-in
consent will be sought from all research participants, including students,
judged competent to provide this as per the process outlined under ‘In-
formed consent’.
Discussion
This will be the first UK-based pilot trial of a school-wide Consent for publication
social marketing intervention to reduce unintended teen- All authors read and approved the final manuscript for publication.
age pregnancy and improve sexual health. The study takes
Competing interests
an innovative approach, encompassing the systematic op- The authors declare that they have no competing interests.
timisation of the Positive Choices intervention, feasibility
testing and refinement in one school followed by a pilot
Publisher’s Note
trial in six schools and accompanied with integral process Springer Nature remains neutral with regard to jurisdictional claims in
evaluation that draws on principles of realist evaluation to published maps and institutional affiliations.
understand processes and mechanisms of action. If the
Author details
pilot trial proves feasibility and acceptability of the opti- 1
London School of Hygiene and Tropical Medicine, 15-17 Tavistock Place,
mised Positive Choices intervention in English secondary London WC1H 9SH, UK. 2London School of Hygiene and Tropical Medicine,
schools, plans will be initiated for a phase III trial of inter- Keppel Street, London WC1E 7HT, UK. 3University of Bristol, 39 Whatley Road,
Bristol BS8 2PS, UK. 4University of Bedfordshire, University Square, Luton LU1
vention and cost-effectiveness. As a universal whole- 3JU, UK. 5Queens University Belfast, University Road, Belfast BT7 1NN, UK.
school social marketing intervention, if found to be 6
Cardiff University, 1-3 Museum Place, Cardiff CF10 3BD, UK. 7University
Ponsford et al. Pilot and Feasibility Studies (2018) 4:102 Page 14 of 15

College London, Gower Street, London WC1E 6BT, UK. 8University College 20. Patton G, Bond L, Carlin JB, Thomas L, Butler H, Glover S, Catalano R, Bowes
London, 1-19 Torrington Place, London WC1E 7HB, UK. G. Promoting social inclusion in schools: group-randomized trial of effects
on student health risk behaviour and well-being. Am J Public Health. 2006;
Received: 3 January 2018 Accepted: 26 April 2018 96(9):1582–7.
21. Shackleton N, Jamal F, Viner RM, Dickson K, Patton G, Bonell C. School-level
interventions to promote adolescent health: systematic review of reviews.
J Adolesc Health. 2016;58(4):382–96.
References 22. Langford R, Bonell CP, Jones HE, Pouliou T, Murphy SM, Waters E, Komro KA,
1. Conception Statistics, England and Wales. 2015. Office for National Statistics. Gibbs LF, Magnus D, Campbell R. The WHO Health Promoting School
2017. Available at: https://www.ons.gov.uk/peoplepopulationandcommunity framework for improving the health and well-being of students and their
/birthsdeathsandmarriages/conceptionandfertilityrates/bulletins/conception academic achievement. Cochrane Database Syst Rev. 2014;(4):Art. No.:
statistics/2015. Accessed 1 Nov 2017. CD008958. https://doi.org/10.1002/14651858.CD008958.
2. Public Health England, Pregnacny and birth - Public Helath Profiles.
23. Andreasen AR. Marketing social marketing in the social change marketplace.
Available at: https://fingertips.phe.org.uk/profile-group/child-health/profile/
J Public Policy and Mark. 2002;21(1):3–13.
child-health-pregnancy. Accessed 1 Nov 2017.
24. Hastings G, McDermott L. Putting social marketing into practice. Br Med J.
3. ONS. Live births to women aged under-18 in EU-28 countries: 2005, 2014 &
2006;332(7551):1210–2.
2015. Office for National Statistics; 2017. Available at: https://www.ons.gov.
25. Hastings G, Stead M. Social marketing. In: Macdowall M, Bonell C, Davies M,
uk/file?uri=/peoplepopulationandcommunity/birthsdeathsandmarriages/
editors. Health promotion practice. Maidenhead: Open University Press; 2006.
livebirths/adhocs/006816livebirthstowomenagedunder18andunder20per
26. Fletcher A, Harden, A; Brunton, G; Oakley, A; Bonell, C. Interventions
1000womenaged15to17and15to19ineu28countries20052014and2015/live
addressing the social determinants of teenage pregnancy. Health Educ.
birthratesper100womenaged15to17and15to19ineu28countries20052014
2008;108(1):29–39.
and2015.xls. Accessed 1 Nov 2017.
27. Marsh D, Brown E, Crocker P, Lewis, H. Building effective middle schools: a
4. Ashcraft A, Fernández-Val I, Lang K. The consequences of teenage
study of middle school implementation in California schools. Los Angeles:
childbearing: consistent estimates when abortion makes miscarriage non-
University of Southern California; 1988.
random. Econ J. 2013;123(571):875–905.
28. Fisher JD. Possible effects of reference group-based social influence on
5. Fletcher J, Wolfe B. Education and labor market consequences of teenage
AIDS-risk behaviors and AIDS. Am Psychol. 1988;43:914–20.
childbearing: evidence using the timing of pregnancy outcomes and
community fixed effects. J Hum Resour. 2009;44:303–25. 29. Bandura A. Social foundations of thought and action: a social cognitive
6. Ermisch J. Does a ‘teen-birth’ have longer-term impacts on the mother? theory. Englewood Cliffs: Prentice Hall; 1986.
Suggestive evidence from the British Household Panel Survey. Colchester: 30. Hawkins JD, Weiss JG. The social development model: an integrated
Institute for Social and Economic Research; 2003. approach to delinquency prevention. J Prim Prev. 1985;6:73–97.
7. Berthoud, R., Ermisch, J., Francesconi, M., Liao, T., Pevalin, D.J., Robson, K., 31. Gavin LE, et al. A review of positive youth development programs that
Long-term consequences of teenage births or parents and their children. promote adolescent sexual and reproductive health. J Adolesc Health. 2010;
Teenage Pregnancy Research Programme Research Briefing, No. 1. 2004, 46(3 Suppl):S75–91.
London: Department of Health. 32. Dishion TJ, McCord J, Poulin F. When interventions harm. Am Psychol. 1999;
8. Francesconi M. Adult outcomes for children of teenage mothers. Scand J 54(9):755–64.
Econ. 2008;110:93–117. 33. Crawford C, Cribb J, Kelly E. Teenage pregnancy in England CAYT impact
9. Kneale D, Fletcher A, Wiggins R, Bonell C. Distribution and determinants of study: report no. 6. London: Institute for Fiscal Studies; 2013.
risk of teenage-motherhood in three British longitudinal studies: implications 34. Stephenson J, Strange V, Allen E, Copas A, Johnson A, Bonell C, Babiker A,
for targeted prevention interventions. J Epidemiol Community Health. 2013; Oakley A. The long-term effects of a peer-led sex education programme
67(1):48–55. (RIPPLE): a cluster randomised trial in schools in England. PLoS Med. 2008;
10. Department for Education and Skills. Teenage pregnancy next steps: 5(11):e224.
guidance for local authorities and primary care trusts on effective delivery 35. Henderson M, Wight D, Raab G, Abraham C, Parkes A, Scott S, Hart G.
of local strategies. London: DfES; 2006. Impact of a theoretically based sex education programme (SHARE)
11. Health Protection Agency. Focus on prevention: HIV and other sexually delivered by teachers on NHS registered conceptions and terminations: final
transmitted infections in the United Kingdom 2003. Annual report 2004. results of cluster randomised trial. BMJ. 2007;334(7585):133. Epub
London: HPA; 2004. 36. Wellings K, Jones K, Mercer C, Tanton C, Clifton S, Datta J, Copas A, Erens B,
12. Independent Advisory Group on Teenage Pregnancy. Independent advisory Gibson L, Macdowall W, Sonnenberg P, Phelps A, Johnson A. The
group response to government white paper choosing health. London: IAG; prevalence of unplanned pregnancy and associated factors in Britain:
2004. findings from the third National Survey of Sexual Attitudes and Lifestyles
13. Skinner R, Marina J. England’s teenage pregnancy strategy: a hard-won (Natsal-3). Lancet. 2013;382:1807–16.
success. Lancet. 2016;338(10044):538–40. 37. Stephenson JM, Strange V, Forrest S, Oakley A, Copas A, Allen E, Babiker A,
14. Wakhisi AS, Allotney P, Dhillon N, Reidpath DD. The effectiveness of social Black S, Ali M, Monteiro H, Johnson AM; RIPPLE study team. Pupil-led sex
marketing in reduction of teenage pregnancies: a review of studies in education in England (RIPPLE study): cluster-randomised intervention trial.
developed countries. Soc Mark Q. 2011;17(1):56–90. Lancet. 2004;364(9431):338–46.
15. Basen-Engquist K, Coyle K, Parcel GS, Kirby D, Banspach SW, Carvajal SC, 38. Baron RM, Kenny DA. The moderator-mediator variable distinction in social
Baumler E. School wide effects of a multicomponent HIV, STD and psychological research: conceptual, strategic and statistical considerations. J
pregnancy prevention program for high school students. Health Educ Pers Soc Psychol. 1986;51:1173–82.
Behav. 2009;28(2):166–85. 39. Wellings K, Nanchahal K, Macdowall W, McManus S, Erens B, Mercer CH,
16. Coyle K, Basen-Engquist K, Kirby D, Parcel G, Banspach S, Collins J, Baumler Johnson AM, Copas AJ, Korovessis C, Fenton KA, Field J. Sexual behaviour in
E, Carvajal S, Harrist R. Safer choices: reducing teen pregnancy, HIV, and Britain: early heterosexual experience. Lancet. 2001;358(9296):1843–50.
STDs. Public Health Rep. 2001;116(Suppl 1):82–93. 40. Durlak JA, Weissberg RP. The impact of after-school programs that promote
17. Coyle K, Basen-Engquist K, Kirby D, Parcel G, Banspach S, Harrist R, Baumler personal and social skills. Chicago: Collaborative for Academic, Social, and
E, Weil M. Short-term impact of safer choices: a multicomponent, school- Emotional Learning; 2007.
based HIV, other STD, and pregnancy prevention program. J Sch Health. 41. Brooks F, Magnusson J, Klemera E, Spencer N, Morgan A. HBSC England
1999;69(5):181–8. National Report: Health Behaviour in School-aged Children (HBSC): World
18. Philliber S, Kaye JW, Herrling S, West E. Preventing pregnancy and improving Health Organization Collaborative Cross National Study. Herfordshire:
health care access among teenagers: an evaluation of the Children’s Aid University of Hertfordshire; 2011. http://researchprofiles.herts.ac.uk/portal/
Society-Carrera Program. Perspect Sex Reprod Health. 2002;34(5):244–51. en/publications/hbsc-england-national-report-health-behaviour-in-school
19. Kirby DB, Rhodes T, Campe S. Implementation of multi-component youth aged-children-hbsc(0e67185e-1179-49cd-88a3-b382222ee8d7)/export.html.
programs to prevent teen pregnancy modelled after the Children's AID 42. Bonell C, Fletcher A, Fitzgerald-Yau N, Hale D, Allen E, Elbourne D, Jones R,
Society - Carrera Program. Scotts Valley: ETR Associates; 2005. Bond L, Wiggins M, Miner A, Legood R, Scott S, Christie D, Viner R. A pilot
Ponsford et al. Pilot and Feasibility Studies (2018) 4:102 Page 15 of 15

randomised controlled trial of the INCLUSIVE intervention for initiating

change locally in bullying and aggression through the school environment:
final report. Health Technol Assess. 2015;19(53):1-110.
43. White J, Hawkins J, Madden K, Grant A, Er V, Angel L, Pickles T, Kelson M,
Fletcher A, Murphy S, Midgley L, Eccles G, Cox G, Hollingworth W, Campbell
R, Hickman M, Bonell C, Moore L. Adapting the ASSIST model of informal
peer-led intervention delivery to the Talk to FRANK drug prevention
programme in UK secondary schools (ASSIST + FRANK): intervention
development, refinement and a pilot cluster randomised controlled trial.
Public Health Res. 2017;5(7):1–97.
44. Moore G, Audrey S, Barker A, Bond L, Bonell C, Hardeman W, Moore L,
O’Cathain A, Tinati T, Wight D, Baird J. Process evaluation of complex
interventions UK Medical Research Council (MRC) guidance (draft).
London: Medical Research Council; 2013. Available at: https://www.mrc.
ac.uk/documents/pdf/mrc-phsrn-process-evaluation-guidance-final/.
Accessed 1 Nov 2017
45. Linnan L, Steckler A. Process evaluation for public health interventions and
research. San Francisco: Wiley; 2002.
46. May C. Towards a general theory of implementation. Implement Sci. 2013;8:18.
47. Stevens K. Assessing the performance of a new generic measure of health-
related quality of life for children and refining it for use in health state
valuation. Appl Health Econ Policy. 2011;9(3):157–69.
48. Campbell MK, et al. Consort 2010 statement: extension to cluster
randomised trials. Br Med J. 2012;345:e5661.
49. Green J, Thorogood N. Qualitative methods for health research. London:
Sage; 2004.
50. Pawson R, Tilley N. Realistic evaluation. London: Sage; 1997.