Transcript PREPARE Project Cape Town

+
Attendance at, and
interim effects of the
PREPARE
intervention, Cape
Town
Cape Town team
Maastricht team
Leif Aarø
+
Overview

Background

PREPARE aims and objectives

Intervention: After-school program

Study Design

Attendance at intervention: Which participants were
exposed to the intervention?

Effects at first follow-up? ITT analysis

Effects among participants with differing levels of
intervention exposure?
+
+
Why was SATZ intervention not
effective in Cape Town?

SATZ was solely curriculum based, competing against a
good curriculum, and did not change the school environment

SATZ was implemented by teachers: We question whether it
was implemented with fidelity.

SATZ did not provide condoms or increase access to sexual
and reproductive health (SRH services

SATZ did not adequately address sexual violence and IPV;
We know IPV increases risk of HIV incidence through
influencing sexual behaviour.
+
PREPARE aim and objectives
AIM
To develop, implement and evaluate interventions to reduce
sexual risk behavior and intimate partner violence (IPV)
among young adolescents
PRIMARY OBJECTIVES/OUTCOMES

Postponing sexual debut

Condom use
SECONDARY OBJECTIVES/OUTCOMES
.

Number of sexual partners

IPV

Three year incidence of conceptions among female participants
+
PREPARE Intervention
1.
Educational component: After-school program of 21
educational sessions with focus on relationships: HIV, IPV
and sexual violence
2.
School environment component focusing on students,
teachers and parents: School safety program to reduce
acceptability and prevalence of IPV and sexual violence,
and to change norms
3.
School health service to increase access to SRH and
mental health services , designed to be compatible with
current approaches and polices
+
Educational sessions: after-school
“club”

One hour after school for 18 weeks; interactive delivery

Young, dynamic, trained facilitators employed and mentored
and monitored closely by PREPARE

Sessions focus on relationships, developed through formative
research and piloting

Mapped on to the LO curriculum

Certificates, gifts; Loyalty Card; Snacks to motivate

Compatible with “MOD Centre” approach
+
+
School safety program

Aims for an environment free of sexual violence and IPV

Uses principles of community development

We involved school safety teams: principal, safety officer, SGB
rep, parent, police officer

Trained by Centre for Justice and Crime Prevention (17/20
schools participated)


Sexual violence and the law

Participatory safety audits

Participatory safety plans
Involves students and parents in a Photovoice project
+
Photovoice
•
1) Group discussions
•
2) Picture taking (PHOTO)
•
3) Caption writing (VOICE)
•
4) Forum meeting presentation
+
School Health Service

Aims to increase access to sexual, reproductive and mental
health

Package of care modeled on ISHP: screening, health
education, and referral to local PHC services (no condoms,
HIV testing, STI treatment)

Collaboration with City and Provincial DoH, WCED and
Desmond Tutu HIV Foundation

Nurses (and health promoter) visit the school once a week
and provide a “health check” for each PREPARE participant

Trained to detect violence and abuse and respond
+
Study Design

Cluster RCT

42 Western Cape High Schools (20 intervention)

Grade 8 students (75 – 100 per school)

Primary outcomes measured in questionnaires at baseline, 6
and 12 months. Questionnaires were done in school hours.

Data from individual participants is linked over time,

In intervention schools participant data is linked to their
intervention attendance records

Review of cental records of births and terminations for 3 year
incidence of conceptions
+
Students’ responses to the survey
 “Thank
you MRC of South Africa for helping me
be open to myself during this questionnaire.”

“It makes me think twice about the things I done
in my life. It will help me understand life more.”
 “Ewe,
thank you to this project because now I
feel happy when I share everything bad that I
have been through to me.”
+
Follow-up surveys: response rates

1st follow-up: 3314/3455 (96% of learners who were
surveyed at baseline)

2nd follow-up: 3054/3455 (88.4% of learners who were
surveyed at baseline)
.
+
Accessibility of after-school
intervention?
Will students attend after-school SRH intervention (educational
sessions and SHS)?
Can we recruit and retain adolescents who are vulnerable to
adverse SRH outcomes?

In these analyses, vulnerable adolescents were :



those who at baseline had had sexual debut
those who had been exposed to IPV and sexual violence
We also investigated whether mental health (as measured
with SDQ), and school climate were associated with
attendance
+
Methods

Facilitators kept attendance registers at every after-school
session

Nurses kept records of every participant who visited the SHS

We linked participant baseline survey data with their
attendance at education sessions and SHS

Bivariate and multiple Poisson regression analyses were
used to assess which factors were associated with attendance
rate at educational sessions

Bivariate and multiple logistic regression analyses were used
to assess which factors were associated with having visited
SHS
+
Attendance at 21 educational
sessions
+
Participant attendance of PREPARE
educational sessions

In 18/20 intervention schools we implemented all 21 afterschool educational sessions. 1576 participants were in the 18
schools and are included in these analyses

Mean attendance at after-school sessions: 8.0 sessions
Females
Males
Mean attendance
8.8 sessions
6.9 sessions
>/= 5 sess.
59.2%
48.8%
>/= 10 sess.
43.6%
32.9%
>/= 15 sess.
30.6%
21.9%
+
Attendance of educational sessions
by vulnerable adolescents
Attendance rates
Even had sex*
Sexual violence vict*
IPV victim*
IPV perpetrator*
Yes
6.3
No
8.5
Yes
7.0
No
8.6
Yes
6.5
No
8.2
Yes
5.2
No
8.3
+ Factors associated with rate of attendance of
educational sessions: multivariate analysis (Table 2)
Higher rate of attendance

Gender: being female

Being bullied in past 6 months

Scoring higher on the school connectedness scale

Scoring higher on the SDQ prosocial scale
Lower rate of attendance:

Age: older students

Having ever had sex

School performance: those who had ever repeated a year

Delinquency: those who reported more frequent fighting, stealing or vandalism in past
three months

Being a perpetrator of IPV
+
Attendance at the school health
service (SHS)
+
Attendance of school health
service

17.3% of participants in 18 schools visited the school nurse
(14.9% of males and 18.7% of females)
Attended SHS (%)
Had had sex
Sexual violence vict
IPV victim
IPV perpetrator
Yes
14.3%
No
18.2%
Yes
16.9%
No
17.3%
Yes
19.0%
No
17.1%
Yes
15.4%
No
17.7%
+
Attendance at school health service:
multivariate analysis (Table 3)
Factors associated with attendance:

Gender: being female

Having been bullied in past 6 months

Scoring higher on the condom knowledge scale
+
Summary: attendance at afterschool intervention

Adolescent girls had a higher rate of attendance at
educational sessions and SHS

Adolescents who had had “early” sexual debut had a lower
rate of attendance at education sessions, but they were no
less likely to visit the SHS

Adolescents with higher prosocial scores had a higher rate of
attendance of education sessions but were no more/less
likely to visit the SHS.

Adolescents who felt more “school connectedness” had a
higher rate of attendance at educational sessions but were
no more or less likely to visit the SHS.

Magnitude of rate differences: not very large
+ Reasons for non-attendance of after-school
sessions? Our observations

Transport and ensuring safe means of travel

Social pressures: if friend group not attending

Uncool to be enthusiastic: need visible encouragement

Competing activities including homework

Change over of facilitators

(Dis) Organisational environment of the school

School ownership of and commitment to intervention

Facilitator even 1 minute late: don’t trust it will happen
+ Reasons for attendance in participants’
words

I wanted to be protected if something happened

My parents said I had to go/ My mother forced me

People told us it was interesting

I wanted to share my feelings/It helps you to talk and feel free all
the time/Helped me to feel relieved and tell the truth/ Makes me
feel better

I wanted to be honest: no-one here to open up to

I want the people to learn more about me

Just to have fun

It tells me things/ You learn more about the future
+
Reasons for non-attendance:
participants’ words

I have to get home and I don’t like it

Going after my friends

Not enough time

I had school work to do

Scary to walk back alone after school/ the bus would leave me

I was just shy

It was too long for me

The learners were badly behaved/ The people in school like to
bully

I just did not feel so
+
IMPACT AT FIRST FOLLOW-UP
(6- Months post-baseline)
+
Interim evaluation: Objectives and
methods
Are there changes in theoretical mediators and behaviours 6
months post-baseline: intention to treat analysis (ITT)
Is there a dose-response? We compared control group to:
“Exposed”, ie attended at least 1 session
“Highly exposed”, ie attended >10 sessions)
Intervention effects were assessed with linear and logistic
regressions (adjusted for clustering) with the outcome as
dependent variable, study arm as factor, and baseline score,
gender and age as covariates.
+
See Results Table
+
Theoretical Mediators
ITT
>/=1
>10
Knowledge condoms
+
+
+
Knowledge HIV/AIDS
+
+
+
Attitudes condoms (pros)
ns
ns
ns
Attitudes condoms (cons)
ns
ns
+
Attitudes delay sex (pros)
ns
+
+
Attitudes delay sex (cons)
ns
+
+
Social norms condoms
ns
ns
ns
Social norms delaying sex
ns
ns
ns
Self efficacy condoms
ns
ns
ns
Self efficacy delaying sex
ns
ns
ns
Action planning condoms
ns
ns
+
Action planning delaying sex
ns
ns
ns
Intentions to have sex
ns
ns
ns
Intentions to use condoms
ns
ns
ns
+ Summary: theoretical mediators for sexual
risk behaviour

Substantial improvements in knowledge (ITT)


Eg Condom knowledge: Control 28% to 29%; Intervention 29% to
37% (43% in highly exposed group)
Improvements in attitudes and action planning in exposed
groups vs control

Eg Negative attitudes to delaying sex: Control means 2.56 to 2.38;
Intervention: 2.51 to 2.28 (2.08 in highly exposed group)

No significant differences in self-efficacy and social norms

All differences in theoretical mediators favoured the
intervention group.
+
Sex behaviour outcomes
ITT
>/=1
>10
Ever had sex (vaginal, anal)
ns
ns
ns
Sexual debut during study (vag, anal)
ns
ns
ns
Ever had sex (vag, anal, oral)
ns
ns
-
Sexual debut (vag, anal, oral) in study
ns
-
-
Vaginal sex frequency
ns
ns
ns
Anal sex frequency
ns
ns
ns
Casual sex (with non-partner)
ns
ns
ns
Condom use last sex (in sub sample)
ns
ns
ns
Condom use ever (in sub-sample)
ns
ns
ns
Condom use frequency (in subsample)
ns
ns
ns
Ever carried a condom
ns
ns
ns
+
Summary: Sexual risk behaviours

No significant differences in sexual debut (vaginal and anal
sex) during 6 months follow-up


When sexual debut includes oral sex, the exposed groups
were more likely to have sexual debut during study than
control


Control 10.5%; Intervention 12.7%
Control 11.2%; Intervention 14.3%; (15.0% and 15.8% in exposed
and highly exposed groups)
No significant differences in frequency of sex, condom use
and carrying condoms
+
IPV theoretical mediators
ITT
>/=1
>10
Attitudes coerced sex
ns
ns
+
Attitudes physical IPV A (pros)
ns
ns
+
Attitudes physical IPV B (pros)
ns
ns
ns
Attitudes physical IPV C (cons)
ns
ns
ns
Social norms coerced sex
ns
ns
ns
Social norms physical IPV
ns
ns
ns
Self efficacy coerced sex
ns
ns
+
Self-efficacy physical IPV
ns
ns
ns
Action planning coerced sex
ns
ns
ns
Action Plan physical IPV perpetration
ns
ns
ns
Action Plan physical IPV vict
ns
ns
+
+
IPV and bullying behavioural
outcomes
ITT
>/=1
>10
IPV victimisation
ns
ns
ns
IPV perpetration
ns
ns
ns
Bullying victimisation
+
+
+
Bullying perpetration
ns
ns
ns
+
Summary: IPV theoretical
mediators and behaviours

ITT analysis: no significant impact on mediators. But, in the
highly exposed group, we had a significant impact on 4/11
mediators

Ignoring statistical significance, for most mediators, there
was a favorable difference between arms, with some
indication of dose response

No evidence of an impact on IPV behaviours

Bullying victimisation was lower in intervention arm: chance
finding?
+
Effects at 6 months: Conclusions

Intervention exposure was not intense enough to change
most mediators and behaviours in a ITT analysis.

Will programme effects have dissipated or intensified over
the last 6 months follow-up?

How likely was it that the intervention to change the school
environment was sustained?

Stratification for effects by sex?

Publication of interim findings?
+
What did participants say about
the intervention?

It made me feel good about myself. Come again!

Prepare helped a lot of people in personal life

Thank you for questions not normally asked

You helped me find solution to my private problems

Made me feel release about my secrets.

I feel better about myself and more confident

Thank you for getting rid of some of my problems

This is the best study of our lives