Data and study population
This cross-sectional study was based on data collected from 2013 to 2018 as part of the Fistula Heath Care Program from women whose fistulas were repaired at Panzi Hospital in South Kivu Province and six other provinces in the DRC. The fistula repairs were performed by the Panzi Hospital’s mobile team and by doctors working at local hospitals that hosted the Panzi mobile team. The Panzi teams, consisting of at least two surgeons, a surgical assistant, a nurse, and an anesthesiologist at each site, are deployed annually or bi-annually, depending on the number of fistulas and the financial and logistical resources available, or at the request of the host sites through NGOs, local churches, civil society, or women’s associations. Once the site is identified, an agreement is signed that allows the Panzi Mobile Team to provide expertise in patient consultation, surgical equipment, medications, and quarterly patient follow-up. Each host site has at least one medical director to manage operations and teams, a nursing director to oversee pre- and post-operative care, and a religious or civil society leader to provide information to patients and their families and to oversee the nutrition of women whose fistulas have been repaired [19].
Data were collected using a ten-page form that included sociodemographic information, gynecologic and obstetric history (pregnancies, abortions, child survival status, current health status of the patient, surgical history and information on recent fertility, fertility aspirations, knowledge, attitudes and practices regarding fistula, diagnosis, treatment and outcome of surgery). Subsequently, the data were entered into a database designed on Epi-Info for this purpose.
Outcome and variables
The outcome variable in this study was the woman’s desire to have another child after fistula repair. This variable was measured using the following question: ‘Do you want to have children after your medical treatment?’ Responses to this question were coded 1 (yes) if the woman wanted to have children after fistula repair and 0 (no) otherwise.
We tested the effect of sociodemographic variables including age (15–19, 20–24, 25–29, 30–34, 35–39, 40–44, 45–49 years), marital status (married, separated/divorced, widow), occupation (farmer, housekeeper, seller, other/unknown), religion (Protestant, Catholic, other), highest level of formal education (no formal education, primary school, secondary school), year of fistula repair (from 2013 to 2018), province (Equateur, Kasai-Oriental, Katanga, Province Orientale, and Kivu, which includes South Kivu, North Kivu, and Maniema, combined into Kivu due to small sample size), place of residence (urban or rural) and parity. Following other studies [20, 21] and to balance the sample size between categories, parity was coded as 0–2 children (low parity), 3–4 children (medium parity) and ≥ 5 children (high parity). We also tested the effect of variables related to women’s health: fistula duration (0–9, 10–19, ≥ 20 years), years since last delivery (0–4, 5–9 and ≥ 10), number of abortions (0, 1–2, ≥ 3), and number of surgeries before fistula repair (0, 1–2, ≥ 3). To maintain a large sample size, a separate category ‘Unknown/Not applicable’ was used for missing information on any variable.
Statistical analysis and analysis strategy
Frequency and contingency tables were used to describe data. A chi-square test was used to test the relationship between each independent variable and the desire to have a child. Adjusted odds ratios along with their 95% confidence intervals (CI) from a binary logit model were used to analyze factors associated with the desire to have a child after fistula repair. We tested for collinearity between the explanatory variables using generalized variance inflation factors (GVIF). One variable (province) that was more collinear with other variables, GVIF > 3 (see Supplementary Table 1), was dropped from the final model. Statistical significance was set at P < 0.05. Analyses were performed using lme4, car and gtsummary packages in R.
We first conducted analyses on all fistula patients aged 15–49, as fertility analysis conventionally focuses on women of fertile age, 15–49 years [22]. Results for all women aged 15–49 may be useful, for example, for predicting fertility based on fertility desires or intentions. All analyses were stratified by parity for three reasons. First, parity is an important determinant of fertility desires [23]. Second, parity may be a confounding factor in that it is associated with other factors such as the number of surgeries or the number of abortions [24]. Third, certain authors suggest that childbearing decisions are made sequentially [24, 25]. This parity-specific design allowed us to investigate how child demand varies for each parity level.
We also conducted analyses on women aged 20–34 years. The 20–35 age group would be preferable, but the age variable was available in five age groups. There are four reasons for using this strategy. First, age is another major factor in fertility and fertility desires and may be correlated with other factors. Second, 20–35 is the most fertile age group. In other words, women tend to have more children in this age range. Third, childbearing is less risky in this age interval [26, 27]. Fourth, this age group may be more relevant for policy development; that is, policymakers can still help them achieve their desired family size without compromising their health.
Add Comment