Link to the University of Pittsburgh Homepage
Link to the University Library System Homepage Link to the Contact Us Form

Informing Low-Dose Aspirin in Gestation and Reproduction Through Novel Methods in Causal Inference

Conzuelo Rodriguez, Gabriel (2021) Informing Low-Dose Aspirin in Gestation and Reproduction Through Novel Methods in Causal Inference. Doctoral Dissertation, University of Pittsburgh. (Unpublished)

Download (3MB) | Preview


Pregnancy loss is the most common complication of human reproduction, occurring in up to 20% of all recognized pregnancies. Aspirin, a widely available anti-inflammatory drug is hypothesized to improve pregnancy outcomes in women with a previous pregnancy loss if administered early in gestation. Under this premise, the Effects of Aspirin in Gestation and Reproduction (EAGeR) trial was devised to evaluate the benefits of assigning preconception low-dose aspirin on live birth. While the study findings suggest a moderate increase in live birth rate of 5.1% (95% CI -0.84 to 11.2), this is currently of limited use due to (1) potential effect modification of the aspirin effect among heterogenous subgroups in the EAGeR population; (2) low generalizability ensuing after demographic differences between the trial sample and the U.S. population; and (3) measurement error associated with time-varying treatments. Presently, there is a critical need to develop epidemiologic methods to overcome these limitations.
This dissertation will focus on evaluating and developing epidemiologic methods to address these limitations. In section 2, we will conduct a simulation study to evaluate the performance of nonparametric doubly robust estimators (i.e., Augmented Inverse Probability Weighting and Targeted Minimum Loss-Based Estimation) against correctly specified Generalized Linear Models to quantify effect modification. Then, we will apply these methods in 1,228 women enrolled in the EAGeR trial to quantify the extent to which the effect of low-dose aspirin on live birth is modified by pre-pregnancy body mass index. I¬¬n Section 3, we address generalizability concerns in EAGeR that result from its highly selective recruitment process. Specifically, we will adapt the parametric g-formula to generalize the intention-to-treat (ITT) and per-protocol (PP) effects of aspirin to a more representative U.S. sample of childbearing age women with a previous pregnancy loss (National Survey of Family Growth). Finally, in Section 4, we will develop an approach based on the parametric g-formula to correct for measurement error of time-varying exposures in complex longitudinal settings. The results from this work will improve our understanding on preconception aspirin role in pregnancy loss. Furthermore, our methods will help to overcome major limitations present in modern epidemiological studies.


Social Networking:
Share |


Item Type: University of Pittsburgh ETD
Status: Unpublished
CreatorsEmailPitt UsernameORCID
Conzuelo Rodriguez, Gabrielgac49@pitt.edugac49
ETD Committee:
TitleMemberEmail AddressPitt UsernameORCID
Committee ChairBodnar, Lisa Mbodnar@edc.pitt.edubodnar
Committee CoChairNaimi, Ashley Iashley.naimi@pitt.eduashley.naimi
Committee MemberBrooks, Maria Mmbrooks@pitt.edumbrooks
Committee MemberWahed, Abdus Swahed@pitt.eduwahed
Committee MemberKennedy, Edward
Date: 27 August 2021
Date Type: Publication
Defense Date: 12 July 2021
Approval Date: 27 August 2021
Submission Date: 9 August 2021
Access Restriction: 2 year -- Restrict access to University of Pittsburgh for a period of 2 years.
Number of Pages: 107
Institution: University of Pittsburgh
Schools and Programs: School of Public Health > Epidemiology
Degree: PhD - Doctor of Philosophy
Thesis Type: Doctoral Dissertation
Refereed: Yes
Uncontrolled Keywords: Epidemiologic methods
Date Deposited: 27 Aug 2021 16:45
Last Modified: 27 Aug 2023 05:15


Monthly Views for the past 3 years

Plum Analytics

Actions (login required)

View Item View Item