Search the site

The University of Manchester Library

In April 2016 Manchester eScholar was replaced by the University of Manchester’s new Research Information Management System, Pure. In the autumn the University’s research outputs will be available to search and browse via a new Research Portal. Until then the University’s full publication record can be accessed via a temporary portal and the old eScholar content is available to search and browse via this archive.

Predictive modelling of assisted conception data withembryo-level covariates.Statistical issues and application.

Stylianou, Christos

[Thesis]. Manchester, UK: The University of Manchester; 2011.

Access to files

Abstract

Statistical modelling of data from the embryo transfer process of In-VitroFertilization (IVF) treatments is motivated by the need to perform statisticalinference for potential factors and to develop predictive models for thesetreatments. The biggest issue arising when modelling these treatments is that anumber of embryos are transferred but unless all of the embryos get implantedor fail to implant then it is not possible to identify which of the embryosimplanted. Little work has been done to address this partial observability of theoutcome as it arises in this contextWe adopt an Embryo-Uterus (EU) framework where a patient response hasdistinct uterine and embryo components. This framework is used to constructstatistical models, expand them to allow for clustering effects and develop apackage that will enable the fitting and prediction of these models in STATA.The capabilities of this package are demonstrated in two real datasets, aimed ininvestigating the effect of a new embryo prognostic variable and the effect ofpatient clustering in these treatments.In a simulation study EU models are shown to be capable of identifying apatient covariate either as a predictor of uterine receptivity or embryo viability.However a simulation case study shows that a considerable amount ofinformation about the embryo covariate is lost due to the partial observability ofthe outcome. Further simulation work evaluating the performance of a numberof proposed alternatives to the EU model shows that these alternatives areeither biased or conservative.The partially observed cycles are finally considered as a missing data problemand two novel modelling approaches are developed which are able to handlethe structure of these treatments. These novel models, based on multipleimputation and probability weighting, are compared to the EU model usingsimulation in terms of predictive accuracy and are found to have similarpredictive accuracy to the EU model.

Additional content not available electronically

An accompanying CD-ROM including the EU library constructed in Chapter 2. The list of files included in the CD-ROM is in Appendix A3.

Bibliographic metadata

Type of resource:
Content type:
Administered thesis
Form of thesis:
Traditional
Type of submission:
Doctoral level ETD - final
Thesis title:
Predictive modelling of assisted conception data withembryo-level covariates.Statistical issues and application.
Degree type:
Doctor of Philosophy
Degree programme:
PhD Medicine (Community Based Medicine)
Publication date:
2011-09-14T13:24:08
Institution:
The University of Manchester
Location:
Manchester, UK
Total pages:
245
Abstract:
Statistical modelling of data from the embryo transfer process of In-VitroFertilization (IVF) treatments is motivated by the need to perform statisticalinference for potential factors and to develop predictive models for thesetreatments. The biggest issue arising when modelling these treatments is that anumber of embryos are transferred but unless all of the embryos get implantedor fail to implant then it is not possible to identify which of the embryosimplanted. Little work has been done to address this partial observability of theoutcome as it arises in this contextWe adopt an Embryo-Uterus (EU) framework where a patient response hasdistinct uterine and embryo components. This framework is used to constructstatistical models, expand them to allow for clustering effects and develop apackage that will enable the fitting and prediction of these models in STATA.The capabilities of this package are demonstrated in two real datasets, aimed ininvestigating the effect of a new embryo prognostic variable and the effect ofpatient clustering in these treatments.In a simulation study EU models are shown to be capable of identifying apatient covariate either as a predictor of uterine receptivity or embryo viability.However a simulation case study shows that a considerable amount ofinformation about the embryo covariate is lost due to the partial observability ofthe outcome. Further simulation work evaluating the performance of a numberof proposed alternatives to the EU model shows that these alternatives areeither biased or conservative.The partially observed cycles are finally considered as a missing data problemand two novel modelling approaches are developed which are able to handlethe structure of these treatments. These novel models, based on multipleimputation and probability weighting, are compared to the EU model usingsimulation in terms of predictive accuracy and are found to have similarpredictive accuracy to the EU model.
Additional digital content not deposited electronically:
An accompanying CD-ROM including the EU library constructed in Chapter 2. The list of files included in the CD-ROM is in Appendix A3.
Keyword(s):
Thesis main supervisor(s):
Thesis co-supervisor(s):
Thesis advisor(s):
Funder(s):
Degree grantor:
Language:
en

Institutional metadata

University researcher(s):
Academic department(s):

Record metadata

Manchester eScholar ID:
uk-ac-man-scw:130553
Created by:
Stylianou, Christos
Created:
14th September, 2011, 12:24:09
Last modified by:
Stylianou, Christos
Last modified:
18th November, 2011, 13:04:53

Can we help?

The library chat service will be available from 11am-3pm Monday to Friday (excluding Bank Holidays). You can also email your enquiry to us.