Interventions to reduce the prevalence
of female genital mutilation/cutting in
Kunnskapsesenterets
African countries
nye PPT-mal
29.-31. May 2012
Rigmor C Berg, Ph.D., CHES
BACKGROUND
 Female genital mutilation / cutting (FGM/C):
”the partial or total removal of the female external
genitalia or other injury to the female genital organs
for cultural or other non-therapeutic reasons” (WHO, 1997)
 4 classifications / types: (WHO, 2008)
– Clitoridectomy
– Excision
– Infibulations
– Other
April 8, 2015
2
BACKGROUND - Prevalence
 About 100 – 130 million worldwide
 About 3 million at risk every year
 Primarily in 28 countries in Africa
– Some countries in the Middle East and Asia
– Immigrant communities in Western countries
April 8, 2015
Somalia
April 8, 2015
FGM/C prevalence among women aged 15-49
Source: Female genital mutilation/cutting : a statistical exploration. New York, NY, UNICEF; 2005.
BACKGROUND - Concerns
 Violates a series of well established human
rights principles, norms and standards, e.g.:
– Universal Declaration of Human Rights, 1948
– International Covenant on Civil and Political Rights, 1966
– Convention on the Elimination of all Forms of
Discrimination against Women, 1979
– Convention on the Rights of the Child, 1989
 No known health benefits
April 8, 2015
BACKGROUND - Concerns
 Almost all cut girls/women experience health problems:
– pain, chronic infections, difficulty in passing urine and
faeces; obstetrical complications (WHO 2000, 2006, 2008)
– systematic review on physical health complications following
FGM/C underway at NOKC
 Little or no change in prevalence over last decades
 Usually carried out on girls under the age of 15  trend
towards lowering of age
 Usually performed by traditional practitioners trend
towards “medicalization”
April 8, 2015
BACKGROUND – Our previous SRs re FGM/C
 3 systematic reviews
Reasons
April 8, 2015
Consequences
Effectiveness
7
BACKGROUND – Our previous work re FGM/C
 Reasons for and against FGM/C:
April 8, 2015
8
BACKGROUND – Our previous work re FGM/C
 Consequences
– Psychological:
– may be more likely to experience psychological disturbances
(have a psychiatric diagnosis, suffer from anxiety,
somatisation, phobia, and low self-esteem)
– Sexual:
– more likely to experience pain during intercourse
– more likely not to experience sexual desire
– lower sexual satisfaction
April 8, 2015
9
BACKGROUND – Our previous work re FGM/C
 Effectiveness of interventions
– Included 6 studies of low methodological quality
– Uncertainties regarding relevance of the interventions
(e.g. regarding objectives, intervention targets,
activities); reasons for limited effectiveness
April 8, 2015
10
OBJECTIVE
1. What is the effectiveness of interventions designed
to reduce the prevalence of FGM/C compared to no
or other active intervention?
2. How do factors related to the continuance and
discontinuance of FGM/C help explain the
effectiveness of interventions designed to reduce
the prevalence of FGM/C?
April 8, 2015
11
METHODS
 Systematic review (transparent, reproducible)
 Search: 13 e-databases, organizations’ websites,
reference lists, experts
 Independent and paired screening, appraisal of
methodological quality, data extraction
 Data analysis
April 8, 2015
12
METHODS
Research Questions:
1. What is the effectiveness of interventions designed to reduce the prevalence of FGM/C compared to no or other
active interventions?
2. How do factors related to the continuance and dicontinuance of FGM/C help explain thye effectiveness of
interventions designed to reduce the prevanelce of FGM/C?
Literature search:
One comprehensive search for empirical studies that address the topic of FGM/C
Screening 1:
Sorting of publications about the effectiveness of
intervention programs to reduce the prevalence of
FGM/C. Application of inclusion criteria.
Synthesis 1: Effectiveness studies
-Quality assessment
-Description, in text and tables, of the programs
-Extraction of effect estimates
Screening 2:
Sorting of publications about factors related to the
continuance and discontinuance of FGM/C.
Application of inclusion criteria
Synthesis 2:
Quantitative studies
-Quality assessment
-Extraction of quant. data
-Synthesis of quant. data
Synthesis 3:
Qualitative studies
-Quality assessment
-Extraction of qual. data
-Synthesis of qual. data
Synthesis 4: Quant-Qual Integrative
Quantitative and qualitative data synthesis of factors
related to the continuance and discontinuance of FGM/C)
Synthesis 5: Realist synthesis approach
Synthesis of results from synthesis 1 (the effectiveness of interventins) and
synthesis 4 (factors related to the continuance and discontinuance of FGM/C)
April 8, 2015
13
METHODS – Realist synthesis
 Realist synthesis attempts to explain how outcomes (efficacy) of an
intervention varies depending on the particular configuration of its
constituent mechanisms and contexts
– The approach is hypothesis generating, the result of which leads to tentative
recommendations meant to influence the design of new programs

“interventions offer resources which trigger choice mechanisms (M) which
are taken up selectively according to the characteristics and circumstances of
subjects (C), resulting in a varied pattern of impact (O)” (Pawson, 2006 p25)
 Mechanisms are the engine behind behaviour (what is on offer in the
program that may persuade participants to change)
 Context is important because the action of mechanisms to some
extent depends on the realities of the context in which they are used
(Pawson, 2006; Pawson et al., 2005)
April 8, 2015
14
Screening
Identification
RESULTS
5,450 records after duplicates removed
Included
Eligibility
5,450 records screened
April 8, 2015
472 records identified through
other sources
6,323 records identified through
database searching
105 full texts assessed for eligibility
5,344 records excluded
1 study not obtained in full text
63 full texts excluded:
-7 effectiveness studies
-56 context studies
35 studies included
-8 effectiveness studies (12 publications)
-27 context studies (30 publications)
15
RESULTS - EFFECTIVENESS
Mali
 8 studies
 Weak methodological
quality
 Controlled before- Senegal
and-after design
Egypt
Ethiopia
Kenya
Burkina
Faso
Nigeria
 7 countries
 N=7,042
April 8, 2015
16
RESULTS
Author
Population
Intervention
Comparison
Outcomes
Diop 1998
N=108. Mali. Health personnel
Training, supervision
No intervention
Beliefs, attitudes, knowledge
Mounir 2003
N=682. Egypt. Female university
students
Education
No intervention
Knowledge
Babalola 2006
N=957. Nigeria.
Community members
Multimedia
communication
No intervention
Intentions, beliefs, attitudes, knowledge
Chege 2004a
N=1,440. Somali refugees in Kenya
Outreach, advocacy
Education
Intentions, beliefs, attitudes, knowledge
Chege 2004b
N=819. Ethiopia. Community
members
Outreach, advocacy
No intervention
Intentions, beliefs, attitudes, knowledge
Easton 2002
N=239. Mali. Community members
Tostan educ. program
No intervention
Beliefs
Diop 2004
N=1,332. Senegal. Community
members
Tostan educ. program
No intervention
Prevalence, intentions, attitudes, beliefs,
knowledge
Ouoba 2004
N=1,465. Burkina Faso. Community Tostan educ. program
members
No intervention
Prevalence, behaviors, intentions, attitudes,
beliefs, knowledge
 1997 – 2004; duration 2 weeks – 18 months
April 8, 2015
17
RESULTS – Study level
 49 study level outcomes
 19 of 49 (39%) of outcomes for which there was
baseline similarity showed significant differences
between the groups
– Most of these (74%) were for the secondary outcomes
attitudes/beliefs and knowledge regarding FGM/C in
the community-based interventions
April 8, 2015
18
RESULTS - Pooled
Figure 3. Forest plot, belief that FGM/C compromise human rights of women
 Belief that FGM/C compromised the human rights of women
Intervention
Study or Subgroup
Events
Comparison
Total Events
Risk Ratio
Total Weight
M-H, Random, 95% CI
Chege 2004a
223
720
288
720
50.4%
0.77 [0.67, 0.89]
Chege 2004b
168
400
76
400
49.6%
2.21 [1.75, 2.79]
1120 100.0%
1.30 [0.46, 3.66]
Total (95% CI)
Total events
1120
391
Risk Ratio
M-H, Random, 95% CI
364
Heterogeneity: Tau² = 0.55; Chi² = 57.60, df = 1 (P < 0.00001); I² = 98%
0.01
0.1
1
10
100
Favours intervention Favours comparison
Test for overall effect: Z = 0.50 (P = 0.62)
 Prevalence of FGM/C among girls 0-10 years
Intervention
Study or Subgroup
Diop 2004
Ouoba 2004
Events
Total Events
Risk Ratio
Total Weight
358
103
199
95.0%
0.77 [0.64, 0.93]
16
519
9
217
5.0%
0.74 [0.33, 1.66]
416 100.0%
0.77 [0.64, 0.92]
877
159
Test for overall effect: Z = 2.84 (P = 0.004)
Risk Ratio
M-H, Random, 95% CI
112
Heterogeneity: Tau² = 0.00; Chi² = 0.01, df = 1 (P = 0.93); I² = 0%
April 8, 2015
M-H, Random, 95% CI
143
Total (95% CI)
Total events
Comparison
0.1 0.2
0.5
1
2
5 10
Favours intervention Favours comparison
19
RESULTS - Pooled
Figure 3. Forest plot, belief that FGM/C compromise human rights of women
 Knowledge of harmful consequences of FGM/C (women)
Intervention
Study or Subgroup
Events
Comparison
Total Events
Risk Ratio
Total Weight
M-H, Random, 95% CI
Diop 2004
243
333
50
200
49.4%
2.92 [2.28, 3.74]
Ouoba 2004
497
578
166
228
50.6%
1.18 [1.08, 1.29]
428 100.0%
1.85 [0.65, 5.22]
Total (95% CI)
Total events
911
740
Risk Ratio
M-H, Random, 95% CI
216
Heterogeneity: Tau² = 0.55; Chi² = 62.35, df = 1 (P < 0.00001); I² = 98%
0.1 0.2
0.5
1
2
5 10
Favours intervention Favours comparison
Test for overall effect: Z = 1.16 (P = 0.25)
 Knowledge of harmful consequences of FGM/C (men)
Intervention
Study or Subgroup
Diop 2004
Ouoba 2004
Events
Total Events
Risk Ratio
Total Weight
M-H, Random, 95% CI
54
82
42
198
48.1%
3.10 [2.28, 4.23]
394
448
137
229
51.9%
1.47 [1.31, 1.64]
427 100.0%
2.11 [1.00, 4.42]
Total (95% CI)
Total events
Comparison
530
448
April 8, 2015
M-H, Random, 95% CI
179
Heterogeneity: Tau² = 0.27; Chi² = 20.27, df = 1 (P < 0.00001); I² = 95%
Test for overall effect: Z = 1.97 (P = 0.05)
Risk Ratio
0.1 0.2
0.5
1
2
5 10
Favours intervention Favours comparison
20
RESULTS – CONTEXT DATA
 27 studies (1 qual)
 Methodological
quality= 9 high, 12
moderate, 6 low
 N= 67 to 15,573
(median= 1,020)
April 8, 2015
Mali
k=1
Senegal
k=1
Burkina
Faso
k=2
Nigeria
k=13
Egypt
k=9
Ethiopia
k=0
Kenya
(Somalis)
k=1
21
Training of health personnel (Mali)
 Pro: custom (61%), good tradition (28%), religious necessity (13%)
 Con: medical complications (45%), bad tradition (30%), prevents
sexual satisfaction (13%), painful experience (13%)
 Improvements not triggered by the intervention
 Not clear extent to which contextual factors embedded in program
 Intervention seems to be fitting response:
– Program embedded in local public health services
– Aimed at improving health providers’ involvement with FGM/C
– Medical complications the most frequently voiced reason for opposing
the practice among Malians thinking FGM/C should be stopped
April 8, 2015
22
Education of female students (Egypt)
 Pro: custom (45%), sexual morals (30%), reduce sexual
desires/preserve virginity (16%)
 Con: complications (22%), sexual problems (16%), no benefit or
value (14%)
 Increase in knowledge of dangers of FGM/C
 Not clear extent to which contextual factors embedded in the
curriculum
 Benefits of placing FGM/C in a reproductive health context
– Egypt DHS data showed few women recognized the potential adverse
physical consequences of the practice for women.
April 8, 2015
23
Communication program (Nigeria)
 Pro: custom (61%), reduce/control female sexual desire (37%),
religion (19%)
 Con: medical complications (38%), bad tradition (49%),
unnecessesary (19%)

Some positive effects

Not clear extent to which identified cultural factors were embedded in the
communication intervention

Sound fit between the program theory of change and program components

With convention theory as a driver of change, dosage of program messages
important (advantage of exposure to a combination of activities and mass
media)
April 8, 2015
24
Outreach and advocacy (Kenya & Ethiopia)
 97% of Somalis in favour of FGM/C: custom license for marriage
(84%), religious obligation (70%), protection of virginity (27%)
 Pre intervention research, embedded in program
 In Kenya, change in comparison group
 In Ethiopia, some positive effects in intervention group
 Embedded in existing reproductive health projects

Critical factors:
– religious leaders
– program exposure
April 8, 2015
25
Tostan educ. prog. (Mali, Senegal, Burkina Faso)
 Mali:
– Pro: custom (61%), good tradition (28%), religious necessity (13%)
– Con: medical complications (45%), bad tradition (30%), prevents sexual
satisfaction (13%), painful experience (13%)
 Senegal:
– Pro: respect tradition (94%), obey religious demand (39%), guarantee
women’s cleanliness (52%), initiate girls (53%), for women to get
married (22%), men prefer cut women (21%)
 Burkina Faso:
— Pro: custom (77%), hygiene (15%), avoid immoral behaviour/preserve
virginity (15%)Con: medical complications (59%), prohibited by law
(35%)
April 8, 2015
26
Tostan educ. prog. (Mali, Senegal, Burkina Faso)
 Unclear whether pre-implementation research
 Issue of FGM/C integrated within a larger project curriculum
 Mali: Marginal effects
 Senegal: Several positive effects
 Burkina Faso: Several positive effects
 Role of religion addressed? Religious leaders’ engagement and
commitment sought?
 Major implementation problems
April 8, 2015
27
SUMMARY
 Some positive developments as a result of
interventions, but:
– low quality of the body of evidence affects the
interpretation of results and draws the validity of the
findings into doubt
– none of the studies randomised, most contained
prognostically dissimilar intervention and comparison
groups, contamination of the intervention seems to
have occurred in four sites
April 8, 2015
28
SUMMARY
 Extent to which can conclude regarding how factors
related to the continuance and discontinuance of FGM/C
help explain the effectiveness of interventions is limited,
because:
– difficult judging match between the interventions’ content
components and factors related to FGM/C’s continuation,
because effectiveness reports lacked descriptions on intervention
content
– studies did not explicitly report on the relevant effective
components of the mechanisms that were assumed to bring about
FGM/C related behavior change
April 8, 2015
29
SUMMARY
 All programs based on a theory that provision of information
improves cognitions about FGM/C
– All measured change in knowledge or beliefs related to FGM/C;
positive results from six programs
 Success contingent upon contextual factors:
– Integrating the issue of FGM/C in a larger set of community-relevant
issues facilitated acceptance
– Alliance with religious leaders
 Process factors:
– Participants not aware of or signed up to take account of the research
dimension of the study; information was not recalled/retained
– Role conflict or uncertainties for staff
– Insufficient measures in place to reduce confounding
– Adverse prevailing program and evaluation climate
April 8, 2015
30
ACKNOWLEDGEMENTS
 Financial support: 3ie (International Initiative for
Impact Evaluation)
 Colleague: Eva Denison
April 8, 2015
31
THANK YOU
Contact details:
Rigmor "Rimo" C Berg rib@nokc.no
April 8, 2015
32