Date of Award


Degree Type


Degree Name

Master of Public Health (MPH)


Public Health

First Advisor

Katherine Masyn

Second Advisor

Huey Chen


INTRODUCTION: The residents in the Low Country region of South Carolina consist of a predominantly low-income, African American population with a history of trauma and experiences of racism. Chronic conditions, unintended pregnancies, and adverse birth outcomes are prevalent. Many women experience rapid repeat pregnancies (RRP) due to lack of access to choices in contraceptive methods or lack of education on the dangers of RRP and prevention through contraception. Low Country Healthy Start (LCHS) aims to ensure that perinatal women and adolescents in the service area who enrolled received adequate prenatal and postpartum care, educational and counseling services, and contraceptive methods, including a Depo Provera injection at discharge (D1) after their index birth in LCHS. Previous research agree that black women, adolescents, low education, mental health, and past trauma are all associated with RRP, and lack adolescents are less likely to retain a form of contraception that requires maintenance and proper usage.

AIM: To (1) examine the effect of D1 and other variables on time to RRP; and to (2) examine the effect of receiving various forms of contraception and their use over time, including the Depo injection (D2) on time to RRP.

METHODS: Clients included in the analysis either delivered a baby while enrolled in LCHS or had complete data on all necessary variables (n=761). The Cox regression model was fitted to model the effect of receiving different contraceptive methods as well as relevant and statistically significant (α=0.05) risk factors on time to RRP.

RESULTS: For Aim 1, D1 resulted in a hazard rate about 46% lower than that of a non-D1 (unadjusted HR = 0.54, 95% CI: 0.36- 0.83; adjusted HR = 0.52, 95% CI: 0.34-0.8). However, after adjusting for other variables (age, unplanned index pregnancy, physical abuse during pregnancy, and postpartum depression score) and the time-varying effect of D1, D1 resulted in a HR of 29.63 (β = 3.39, 95% CI: 6.049- 145.141), that decreased at a natural log function of time (HR = 0.22, β = -1.53, 95% CI: 0.12-0.40).

For Aim 2, D2 resulted in a lower hazard rate than non-D2 (unadjusted HR = 0.17, 95% CI: 0.09-0.32; adjusted HR = 0.16, 95% CI: 0.08-0.31). Adjusting all variables in Aim 2, including D2, D1 resulted in a statistically insignificant lower HR of 0.88 (p = 0.544, 95% CI: 0.57-1.34). There was no significant interaction between D1and D2 or between D1 and any other contraceptive type. LARC showed a highly protective but not statistically significant effect against RRP (adjusted HR = 0.05, p = 0.093, 95% CI: 0.002-2.26), but that protective effect decreased multiplicatively by about .25 with each passing month (HR = 1.25, p = 0.029, 95% CI: 1.02-1.53).

DISCUSSION: These findings indicate that the Depo injection, although important to receive at discharge, must be continued consistently to have a significant protective effect in preventing a RRP. LARC methods in general are strong protective factors. Being issued a contraceptive method that required adherence predicted a shorter inter-pregnancy interval (IPI), but this reflects the client’s adherence to the contraceptive method, and not its biological effectiveness. Future research should examine the effect of receiving the Depo injection at discharge on the continuation of different contraceptive methods, as well as the effect of counseling and educational services on contraceptive use and time to RRP.