Dissertation/Thesis Abstract

Determinants of Contraceptive Choice: Factors Affecting Contraceptive Nonuse among Urban Women Utilizing Title X Services
by Bommaraju, Aalap, M.S., University of Cincinnati, 2013, 123; 1548538
Abstract (Summary)

Background: Disparities in unintended pregnancy are partially due to ineffective contraceptive method choice among vulnerable populations. Improved understanding of the ecological, individual, and health system related determinants of contraceptive choice can provide guidance for how to reduce ineffective contraceptive method choice among women at high risk for unintended pregnancy.

Objectives: Secondary data analysis is performed on visit data from women utilizing Title X Family Planning services at the Cincinnati Health Department’s Reproductive Health & Wellness Program (RHWP) to determine the significance of predisposing factors (age, African- American race, education), enabling factors (income, health insurance status, socio-behavioral risks), need factors (having had a recent birth, number of living children) on choosing an ineffective contraceptive method. Mediation analysis is performed to determine if health system factors mediated the effect of these explanatory variables.

Methods: Using data from 1,119 RHWP clients who were not seeking pregnancy, multinomial logistic regression is used to compare pill, patch, and ring users, depot medroxyprogesterone acetate (DMPA) users, and long-acting reversible contraception (LARC) users with a reference group of ineffective method users. Multinomial logistic regression is first performed with all independent variables except health system mediation. Then, it is performed with inclusion of health system mediation. Multiple linear regression analysis is performed to determine significant relationships between independent factors and health system mediation. Mediation analysis is performed for any independent variable that is significantly correlated with both contraceptive method choice and health system mediation for the purpose of determining if any witnessed mediation effect is statistically significant.

Results: The model including health system mediation is found to account for more variance in the data than the model excluding health system mediation (Nagelkerke R-squared = 0.195 and 0.158, respectively). Through both models, and in all three comparisons, higher age is found to reduce the odds of choosing a more effective method with only minor mediation effects. Having more children results in higher odds of both DMPA and LARC uptake in both models and is unmediated by the health system. Being of African-American race results in lower odds of LARC use in both models – a finding that is also not mediated by the health system. Health system mediation effects are found in the LARC and DMPA comparisons and not in the pills, patch, or ring comparison. In the DMPA comparison, inclusion of health system mediation eliminates income and insurance status as significant explanatory variables. In the LARC comparison, inclusion of health system mediation eliminates the explanatory significance of having a recent birth.

Conclusions: Analysis of the role of health system mediation suggests that health system level factors play a large role in explaining contraceptive choice among contraceptive methods that require increased health care utilization. The impact of health system mediation on the significance of other predisposing, enabling, and need variables implies that health care access, utilization, and quality are important factors that should be included in future models for understanding contraceptive choice.

Indexing (document details)
Advisor: Mase, William
Commitee: Mooney, Jennifer
School: University of Cincinnati
Department: Public Health
School Location: United States -- Ohio
Source: MAI 52/04M(E), Masters Abstracts International
Source Type: DISSERTATION
Subjects: Medicine, Womens studies, Public health
Keywords: Contraceptive use, Health care access, Health system, Mediation, Title X services, Vulnerable populations
Publication Number: 1548538
ISBN: 9781303577734
Copyright © 2019 ProQuest LLC. All rights reserved. Terms and Conditions Privacy Policy Cookie Policy
ProQuest