Page 242 - Special Topic Session (STS) - Volume 2

STS489 Chibuzor C. N. et al.

Modelling and mapping prevalence of Female
Genital Mutilation/C (FGM/C) among 0-14 years
old girls in Kenya, Nigeria and Senegal
Chibuzor Christopher Nnanatu , Glory Atilola , Paul Komba , Lubanzadio
1
1
1
Mavatikua , Zhuzhi Moore , Dennis Matanda , Ngianga-Bakwin Kandala
1
3
2
1
1 Northumbria University, Newcastle, UK
2 Population Council, Kenya
3 Independent Consultant

Abstract
World Health Organisation defines Female Genital Mutilation/cutting (FGM/C)
as all forms of injury caused to the external female genitalia for non-medical
reasons. FGM/C is a public health and human right issue, which is strongly
anchored on customs and traditions, without any established benefit. The
practice has both short- and long-term consequences ranging from
haemorrhage to complications during child birth. It is estimated that over 200
million women and girls alive today globally, have undergone FGM/C at some
point in their lives. FGM/C is rampant in Africa where it is feared that some 3
million girls are at risk of being cut each year. Recent studies showed that
FGM/C prevalence among women aged 15-49 in Kenya was estimated at
27.1% in 2008-9. On the other hand, in 2017, FGM/C prevalence among girls
aged 0-14 years was estimated at 14.0% and 25.3% in Senegal and Nigeria,
respectively. There are several change-provoking interventions geared
towards eliminating the practice. Consequently, change has begun but rather
sluggish, and this calls for the generation of credible statistical evidence that
sufficiently describes where, when and how change is taking place. Robust
Bayesian hierarchical space-time models which simultaneously accounted for
unobserved effects of space and time, as well as space-time interactions, whilst
controlling for other linear and nonlinear covariates were employed. These
models were developed and fitted on the available datasets in a coherent
mixed models regression framework. Posterior inference was carried out using
Markov Chain Monte Carlo (MCMC) techniques, while model fit and
complexity assessments utilised Deviance Information Criterion (DIC)
approach. The approach adopted in this study allowed us to jointly account
for individual-, household-, community-level factors, map and identify
patterns and spatial and temporal variations in the practice, thus unmasking
FGM/C hotspots and patterns over time, as well as their characteristics across
the three countries. Factors found to associate with higher risk of the practice
included mother’s FGM/C status, support for FGM/C continuation, household
wealth index, level of education of mother, region and type of place of
residence, marital status and religion. Our findings are important in various
ways: First, it is now clear, at least to an extent, where and when changes are

231 | I S I W S C 2 0 1 9