Mooney, Shauna (2025) Advancing the Use of Service Statistics for Estimating Modern Contraceptive Use through Bayesian Modelling Approaches. PhD thesis, National University of Ireland Maynooth.
Preview
ShaunaMooney_phd_thesis.pdf
Available under License Creative Commons Attribution Non-commercial Share Alike.
Download (1MB) | Preview
Abstract
Monitoring family planning progress requires accurate and timely estimates of key
indicators such as the modern contraceptive prevalence rate (mCPR), defined as
the proportion of women of reproductive age using modern contraceptive methods.
However, large-scale survey data, often the primary source of these mCPR estimates,
are infrequently collected, leading to data gaps. Family planning service statistics,
routinely collected during service delivery, provide a supplementary data source.
From these statistics, an indicator known as Estimated Modern Use (EMU) can be
derived, but it is a biased estimator of mCPR and has uncertainties that need to
be considered. This thesis focuses on advancing the methodology, application, and
usability of EMU by quantifying and accounting for biases and uncertainties and
ultimately better supporting low- and middle-income countries in tracking family
planning progress.
First, we refine the derivation of EMU by improving upon necessary adjustments for
missing private sector contributions to family planning service statistics databases.
Previous methods carried out adjustments assuming constant contraceptive supply
share distributions over time and without quantifying uncertainty associated with
supply share estimates. We update and improve upon the EMU calculation to
reflect time-varying contraceptive supply and capture uncertainty in the private
sector adjustment, resulting in observation-specific uncertainty previously unseen
in EMU derivation. These improvements are demonstrated through country-level
case studies.
Next, we develop a new approach to incorporating EMUs into the Family Planning
Estimation Tool (FPET), which generates estimates and short-term projections
of mCPR. We use a Bayesian hierarchical modelling approach to estimate data
type-specific EMU uncertainty and across country variance parameters before
incorporating the resulting estimates into FPET. We introduce a Bayesian hierarchical
model when using EMUs in FPET to capture uncertainty, accounting
for country- and type-specific uncertainties through the hierarchically estimated
variance hyperparameters. Model validation results and anonymised country-level
case studies highlight the impact to mCPR estimates when including EMU data in
FPET using this approach. Validation findings demonstrate improved predictive
performance with EMU inclusion compared to relying on survey data alone, while
case studies provide further insights into its effects across different country contexts.
Finally, we present a paper to describe the details and implementation of ss2emu,
an open-source R package, developed to perform the most advanced SS-to-EMU
calculation process in R. This tool complements existing workflows performed by
country-level data experts, providing reproducible datasets and visualisations for
use in FPET. By offering a scalable and user-friendly solution, the tool enhances
accessibility and empowers users, such as family planning monitoring and evaluation
officers, to make more informed decisions in family planning monitoring.
Together, these contributions improve the accuracy, integration, and usability of
EMU as a family planning indicator, enabling countries to better monitor progress
toward family planning goals and address data gaps with confidence.
Item Type: | Thesis (PhD) |
---|---|
Keywords: | Use of Service Statistics; Modern Contraceptive Use; Bayesian Modelling Approaches; |
Academic Unit: | Faculty of Science and Engineering > Research Institutes > Hamilton Institute |
Item ID: | 20667 |
Depositing User: | IR eTheses |
Date Deposited: | 09 Oct 2025 15:25 |
URI: | https://mural.maynoothuniversity.ie/id/eprint/20667 |
Use Licence: | This item is available under a Creative Commons Attribution Non Commercial Share Alike Licence (CC BY-NC-SA). Details of this licence are available here |
Repository Staff Only (login required)
Downloads
Downloads per month over past year