Prevalence and determinants of unmet need for contraception in North Gonja District, Ghana

Background: Unmet need for contraception contributes to the burden of unwanted pregnancies, which are correlated with a host of adverse maternal and child outcomes. The aim of this study was to determine the prevalence and identify the determinants of unmet need for contraception in North Gonja District, Ghana. Methods: A cross-sectional survey involving 386 randomly selected women of childbearing age was conducted in North Gonja district, Ghana, with the use of a questionnaire in household interviews. Women were classified as having unmet need for contraception if they were fecund, sexually active and wished to postpone the next birth or halt childbearing completely but were not using any form of contraception. Chi-square/Fisher’s exact test and logistic regression analysis were used to identify the determinants of unmet need. Results: The mean age of the study population was 26.1 (±8.4) years and awareness on contraception was almost universal in the district (95.9%). The overall prevalence of unmet need for contraception was 38.9%, with 27.5% having unmet need for limiting and 12.2% unmet need for spacing. In multivariate analysis, compared to women aged 25-29 years, those aged 20-24 years [Adjusted Odds Ratio (AOR) 0.26; 95% Confidence Interval (CI) 0.11-0.58] and 30 years and above (AOR 0.25; 95% CI 0.09-0.73) were less likely to have unmet need for contraception. However, uneducated women (AOR 5.06; 95% CI 1.07-24.01) compared with those educated to tertiary level; those unaware of family planning (AOR 3.93; 95% CI 1.12-13.80) compared to those aware; and those who had not previously practised contraception (AOR 1.81; 95% CI 1.09-3.00) compared to those who did were more likely to have unmet need. Conclusions: The present study found high prevalence of both awareness on and unmet need for contraception among the study population. Unmet need for contraception is associated with age, educational status, awareness on family planning and previous contraception practice. Educational campaigns to promote contraception should prioritize women of middle age and low educational status. Further studies are needed to understand the low correlation between awareness on and unmet need for contraception. The study was a cross-sectional survey conducted in North Gonja District in March to June, 2016. The North Gonja District is located in the western part of Savanna Region and lies within longitude 1051 and 20,581 West and latitude 80,321 and 10,021 North. The district shares boundaries to the West with West Gonja and Wa East districts, to the East with Tolon District, to the North with Mamprugu-Moagduri and Kumbungu districts and to the South with Central Gonja district [16]. The district has a total land mass of about 4845.5sq km, representing about 6.9% of the total land size of the Region. About 75% of the communities in the district are difficult to access due to the poor nature of the road network and heavy flood during rainy season. This normally affects the delivery of health services in the district. The inhabitants of the district are predominantly Gonjas and Tamplumas but people of other ethnic groups such as Dagomba, Mamprusi, Hanga and Ewe can also be found there. The District is divided into four health sector operational sub-districts namely Bawena, Daboya, Lingbinsi and Mankarigu and has a total of sixty-one (61) communities. The district has an estimated population of 43,547 [16]. The study population consisted of women aged 15–45 years in the district. A minimum sample size of 384 was derived using the formula N = [z2 p (1-p)] ÷ ME2 [17], where z = confidence interval at 95% thus 1.96; p = assumed prevalence of unmet need for contraception in North Gonja District, Ghana, 50%; and ME = margin of error, 0.05. The sample size was rounded up to 386. Convenience sampling was used to select one sub-district (Daboya) out of the 4 sub-districts in the District because of the relative ease of access and 20 communities were also conveniently selected from this sub-district. Probability proportional to size was used to determine the number of women to sample from each of the selected communities. The respondent women were selected using simple random sampling. Four hundred women were approached to get the 386 women to interview giving a response rate of 96.5%. The 14 women not interviewed were not resident in the study area (n = 7), did not consent to participate (n = 5) and could not be interviewed because of language barrier (n = 2). Data were collected in face-to-face interviews in respondents’ homes by 4 trained community health nurses who were engaged in the provision of reproductive health services at their work places. The questionnaire used had sections on socio-demographic characteristics, reproductive and fertility history, and knowledge and practice of FP (See Additional file 1). The socio-demographic characteristics included age, highest educational status, marital status, occupation, and religious afilliation. Age in years was self-reported and this was categorised into 5-year groups (i.e., < 20 years, 20–24 years); marital status was categorized into married, and single or divorced (i.e., not currently married); and highest educational level ranged from no education to tertiary level education. Ethnicity had two main categories Gonja and Dagomba and all others were grouped together as “others”; and religion had Muslim and “others”. Information collected on reproduction and fertility history of respondents included age at first birth, number of children, previous contraceptive use, whether they would accept help in limiting/spacing childbirth and reasons for non-use of contraceptives. Issues explored on FP knowledge included what FP was and respondents were also asked to name at least one modern contraceptive. The dependent variable of the study was unmet need for modern contraception and the independent variables were socio-demographic factors, and contraceptive knowledge and practices of the women. The proportion of fecund, sexually active women who wish to stop childbearing  completely but are not using any modern contraceptive method. The proportion of fecund, sexually active women who want another child after 2 years but are not using any modern contraceptive method. Any woman having unmet need for limiting or spacing childbirth was considered as having unmet need for contraception [6]. Women were considered to be fecund if they were pregnant or postpartum amenorrheic. Non-pregnant and non-amenorrheic women were classified as fecund if they did not meet the criteria for infecundity i.e., they were married for at least 5 years but did not have a birth in the past 5 years or are not currently pregnant and have never used any contraceptive method; or self-reported infecundity, hysterectomy, or menopause or never menstruated or have been postpartum amenorrheic for at least 5 years; or they are not pregnant or amenorrheic and their last menstrual period occurred more than 6 months prior to the survey. All married women were considered to be sexually active and unmarried women were considered to be sexually active if they had a sexual encounter in the month preceding the survey [6]. The interviewers underwent a 2-day training on the contents of the questionnaire and techniques of questionnaire administration. Prior to data collection, the questionnaire was pretested in a community in the North Gonja district that was not part of the study communities and questions that were unambigous or did not elicit the appropriate responses were rephrased. Data management and analysis were performed using Stata (version 13, Stata Corporation, College Station, Texas, USA). The determinants of unmet for contraception were identified in a two-stage process. Bivariate associations between unmet need for contraception and the characteristics of the women were explored using Chi-square or Fisher’s exact test and those characteristics that were statistically significant were retained and entered into a logistic regression model. In the logistic regression analysis, odds ratios and their 95% confidence intervals were computed to reflect the strength of association between unmet need and its independent determinants. In all analyses, a p-value less than 0.05 was considered statistically significant. The protocol for this study (Protocol Number 04–2016) was approved by the Joint Ethics Board of School of Medicine and Health Sciences and School of Allied Health Sciences, University for Development Studies, Tamale. Written informed consent was obtained from the study women and from the parents of subjects less than 16 years old before interviews were conducted. It was made clear to all subjects that participation was voluntary.