Buprenorphine Retention on Acute Opioid-Related Healthcare Encounters During Pregnancy and Postpartum Restricted; Files Only

Abstract

Pregnant and postpartum individuals with opioid use disorder (OUD) experience heightened risks of overdose, emergency department visits, and withdrawal-related hospitalizations. While buprenorphine is one of the first-line treatment recommendations for OUD, many individuals struggle with treatment retention. Additionally, the impact of buprenorphine retention on health outcomes is dynamic—changing over time and in response to evolving clinical and social circumstances. This dissertation applies time-varying causal methods to evaluate the effect of buprenorphine retention on acute opioid-related healthcare encounters during pregnancy and postpartum, addressing critical methodological and clinical questions.

 

Aim 1: We used a time-varying incremental propensity score estimator to examine how weekly shifts in buprenorphine retention affect the risk of opioid-involved overdose, emergency visits, and managed withdrawal admissions. Our findings suggest that even small, sustained increases in buprenorphine retention over time can reduce the occurrence of these critical health events.

 

Aim 2: The presence of psychiatric and substance use comorbidities may modify the effectiveness of buprenorphine treatment. Using g-estimation of structural nested mean models, we evaluated how the effect of buprenorphine retention on withdrawal-related hospitalizations varies over time in the presence of these comorbidities. We found that buprenorphine retention provided acute protection for those without psychiatric or substance use conditions, but this protective effect diminished among individuals with these comorbidities—underscoring the need for integrated, multimodal care strategies.

 

Aim 3: Many perinatal studies analyzing buprenorphine treatment exclude individuals who enter care late, potentially biasing results. Using a plasmode simulation study, we assessed how different patterns of left truncation (missing early data) impact time-varying treatment effect estimates. We compared the performance of different methods, finding that inverse probability weighting and non-iterative g-computation yielded stable estimates over time, while iterative g-computation produced more bias. Our findings are consistent with previous work on the behavior of methods for time-dependent confounding adjustment and extend our understanding into the additional impact of left truncation. 

 

Conclusion: Buprenorphine retention plays a time-sensitive role in preventing opioid-related healthcare encounters during pregnancy and postpartum. However, the effectiveness of retention varies based on an individual’s evolving comorbid conditions. These findings emphasize the need for dynamic, individualized treatment strategies that account for evolving health risks over time. By leveraging time-varying causal methods, this research provides insights into optimizing buprenorphine treatment and improving maternal health outcomes in the context of the opioid epidemic.

Table of Contents

CHAPTER 1 INTRODUCTION      1

CHAPTER 2  INCREMENTAL PROPENSITY SCORE EFFECT OF BUPRENORPHINE RETENTION ON OPIOID-INVOLVED OVERDOSE AND ACUTE HOSPITAL ENCOUNTERS DURING PREGNANCY AND POSTPARTUM       7

ABSTRACT  8

BACKGROUND       10

METHODS    11

RESULTS      19

DISCUSSION 22

CHAPTER 3  TIME-VARYING EFFECT MODIFICATION OF BUPRENORPHINE RETENTION ON MANAGED WITHDRAWAL DURING PREGNANCY AND POSTPARTUM          38

ABSTRACT  39

BACKGROUND       41

METHODS    43

RESULTS      48

DISCUSSION 59

CHAPTER 4  IMPLICATIONS OF LEFT TRUNCATION IN COMPLEX LONGITUDINAL DATA: A SIMULATION STUDY        64

ABSTRACT  65

BACKGROUND       67

METHODS    68

RESULTS      80

DISCUSSION 85

CHAPTER 5  CONCLUSIONS       94

SUMMARY OF KEY FINDINGS    95

STRENGTHS & LIMITATIONS      95

RECOMMENDATIONS FOR FUTURE RESEARCH AND PRACTICE  99

CONCLUDING REMARKS 101

REFERENCES          103

 

About this Dissertation

Rights statement

• Permission granted by the author to include this thesis or dissertation in this repository. All rights reserved by the author. Please contact the author for information regarding the reproduction and use of this thesis or dissertation.

School	Laney Graduate School
Department	Epidemiology
Degree	Ph.D.
Submission	Dissertation
Language	English
Research Field	Health Sciences, Epidemiology
Keyword	opioid use disorder perinatal epidemiology causal inference
Committee Chair / Thesis Advisor	Patrick Sullivan, DVM, PhD, Emory University Ashley Naimi, PhD, Emory University
Committee Members	Michael Kramer, PhD, Mercer University Hannah Cooper, ScD, Emory University

Last modified

Primary PDF

Thumbnail	Title	Date Uploaded	Actions
	File download under embargo until 22 May 2027	2025-09-09 11:44:48 -0400	File download under embargo until 22 May 2027

Supplemental Files

Thumbnail	Title	Date Uploaded	Actions