Factors influencing unmet need for family planning among Ghanaian married/union women: a multinomial mixed effects logistic regression modelling approach

Chris Guure; Ernest Tei Maya; Samuel Dery; Baaba da-Costa Vrom; Refah M Alotaibi; Hoda Ragab Rezk; Alfred Yawson

doi:10.1186/s13690-019-0340-6

Factors influencing unmet need for family planning among Ghanaian married/union women: a multinomial mixed effects logistic regression modelling approach

Arch Public Health. 2019 Mar 12:77:11. doi: 10.1186/s13690-019-0340-6. eCollection 2019.

Authors

Chris Guure¹, Ernest Tei Maya², Samuel Dery¹, Baaba da-Costa Vrom¹, Refah M Alotaibi³, Hoda Ragab Rezk^{3

4}, Alfred Yawson¹

Affiliations

¹ 1Department of Biostatistics, School of Public Health, University of Ghana, Legon, Accra Ghana.
² 2Department of Population, Family and Reproductive Health, School of Public Health, University of Ghana, Legon, Accra Ghana.
³ 3Department of Mathematical Sciences, Faculty of Science, Princess Nourah bint Abdulrahman University, Riyadh, Saudi Arabia.
⁴ 4Princess Nourah bint Abdulrahman University, Riyadh, Saudi Arabia.

Abstract

Background: Unmet need for family planning is high (30%) in Ghana. Reducing unmet need for family planning will reduce the high levels of unintended pregnancies, unsafe abortions, maternal and neonatal morbidity and mortality. The purpose of this study was to examine factors that are associated with unmet need for family planning to help scale up the uptake of family planning services in Ghana.

Methods: This cross sectional descriptive and inferential study involved secondary data analysis of women in the reproductive age (15-49 years) from the Ghana Demographic and Health Survey 2014 data. The outcome variable was unmet need for family planning which was categorized into three as no unmet need, unmet need for limiting and unmet need for spacing. Chi-squared test statistic and bivariate multilevel multinomial mixed effects logistic regression model were used to determine significant variables which were included for the multivariable multilevel multinomial mixed effects logistic regression model. All significant variables (p < 0.05) based on the bivariate analysis were included in the multinomial mixed effects logistic regression model via model building approach.

Results: Women who fear contraceptive side effects were about 2.94 (95% CI, 2.28, 3.80) and 2.58 (95% CI, 2.05, 3.24) times more likely to have an unmet need for limiting and spacing respectively compared to those who do not fear side effects. Respondents' age was a very significant predictor of unmet need for family planning. There was very high predictive probability among 45-49 year group (0.86) compared to the 15-19 year group (0.02) for limiting. The marginal predictive probability for spacing changed significantly from 0.74 to 0.04 as age changed from 15 to 19 to 45-49 years. Infrequent sexual intercourse, opposition from partners, socio-economic (wealth index, respondents educational level, respondents and partner's occupation) and cultural (religion and ethnicity) were all significant determinants of both unmet need for limiting and spacing.

Conclusions: This study reveals that fear of side effect, infrequent sex, age, ethnicity, partner's education and region were the most highly significant predictors of both limiting and spacing. These factors must be considered in trying to meet the unmet need for family planning.

Keywords: Estimating unobserved effects; Family planning; Multilevel modelling; Unmet need.