Menu
Menu
Menu
Menu
The fee exemption policy for EmONC in Mali aims to lower the financial barrier to care. The objective of the study was to evaluate the direct and indirect expenses associated with caesarean interventions performed in EmONC and the factors associated with these expenses. Data sampling followed the case control approach used in the large project (deceased and near-miss women). Our sample consisted of a total of 190 women who underwent caesarean interventions. Data were collected from the health workers and with a social approach by administering questionnaires to the persons who accompanied the woman. Household socioeconomic status was assessed using a wealth index constructed with a principal component analysis. The factors significantly associated with expenses were determined using multivariate linear regression analyses. Women in the Kayes region spent on average 77,017 FCFA (163 USD) for a caesarean episode in EmONC, of which 70 % was for treatment. Despite the caesarean fee exemption, 91 % of the women still paid for their treatment. The largest treatment-related direct expenses were for prescriptions, transfusion, antibiotics, and antihypertensive medication. Near-misses, women who presented a hemorrhage or an infection, and/or women living in rural areas spent significantly more than the others. Although abolishing fees of EmONC in Mali plays an important role in reducing maternal death by increasing access to caesarean sections, this paper shows that the fee policy did not benefit to all women. There are still barriers to EmONC access for women of the lowest socio-economic group. These included direct expenses for drugs prescription, treatment and indirect expenses for transport and food.
The study population in the larger project consisted of 484 women, who were both maternal deaths and near-misses and experienced a caesarean section in the latter case–control study. A near-miss was defined as a pregnant women who survived to severe medical complications such as prenatal or postpartum hemorrhage, severe pre-eclampsia, eclampsia, miscarriage, or uterine rupture of the uterus or obstructed. Cases were restricted to the following four obstetric complications—hemorrhage, eclampsia, postpartum infection and uterine rupture—as their first signs were easily recognized by the community and health professionals when following the national clinical guideline. When a woman had experienced two or more complications, the most serious or the lethal one was selected. Since the criterion of inclusion in the present study included only women who delivered by caesarean in EmONC, a total of 190 women were finally selected (95 deceased and 95 near-misses). For consistency purposes, one same sampling design was conducted for near-misses and deceased women. The interviewer traced back the woman’s steps and conducted the investigation starting in the last place the woman was treated (district health center or regional hospital), moving on to the community health center (CSCom) and ending at the woman’s household. Medical records were recorded from the health professionals who treated the woman. A social interview was conducted with the head of the household and/or any household member who was present during the caesarean episode. The questionnaire administered during the interview recorded socio-demographic data, information on the episode, such as transport time, and any data of expenses related to caesarean intervention. Because household incomes were not available, the socio-economic status was estimated using a wealth index based on assets. This estimated wealth index was obtained using principal components analysis (PCA), as commonly performed in some other studies [26–28]. This index has been calculated based on the possession (Yes/No) of eight variables assets as follow: the household’s items (cellular phone, stereo, motorcycle), the quality of the home’s building materials (roof and floor), ownership of the house, cattle for commerce and cattle for consumption. The first two dimensions of the PCA explained 55.26 % of the total variation after varimax rotation (37.14 % for dimension 1 and 18.12 % for dimension 2). Only the dimensions with an eigenvalue greater than 1 were retained for further analyses. For each selected dimension, responses to the items were weighted and averaged to create an overall score and thus an integrative wealth index was calculated. The wealth index was used as a categorical variable for analysis purposes (three categories). Although the women were paired in the larger project (Causes et effets du premier retard sur la létalité des urgencies obstétricales dans la region de Kayes (Mali)), the present analyses based on the caesarean sections used the variable near-misses versus deceased women as a covariate in statistical analyses. One woman was excluded from the analyses as an outlier due to extremely high treatment expenses (356,000 FCFA-754 USD) (1 USD = 472 FCFA, 2008–2011 average). Four categories of expenses were considered in the statistical analyses, as follow: total amount of expenses, expenses for treatment, transportation and other expenses. For a total of six women for who only the total amount of expenses was available, a mean imputation procedure was computed to replace missing values in the other categories of expenses such as treatment, transportation and others. Descriptive analyses were used to display the profile of factors Figs. 1, ,22 and Tables 1, ,3,3, ,4.4. For the Figs. 3 and and4,4, marginal predicted values, that are the expected value of a typical observation from some level of a categorical factor when all the other factors have been set to neutral values, were used. Mann–Whitney U test was used to compare the four categories of expenses between women according to their residence or the diagnosis. Following a forward stepwise procedure, a series of multivariate linear regression analyses were completed to identify which factors were significantly associated with each category of expense. Outcomes of the caesarean episode (near-misses versus deceased women), diagnosis, permanent residence of the parturient, and wealth index were considered as potential factors and In all regression analyses, the absence of collinearity among variables was assessed with the tolerance and variance inflation factor (VIF) measures. Homoscedasticity of residuals was also verified. Finally, any influential data was tested with Cook’s distance. Mean proportion of expenses by residence (n = 190, FCFA*) Mean proportion of expenses by diagnosis (n = 190, FCFA*) Characteristics of the two-dimension of the principal component analysis (factorial loading of selected items, eigenvalues, % of variance explained and Cronbach’s alpha values on each dimension are provided), varimax rotation Marginal predicted values of expenses by residence with 95 % confidence interval (n = 190, FCFA*) Marginal predicted values of treatment expenses with 95 % confidence interval (n = 190, FCFA*) Factors associated with the expenses of women who had caesareans in Kayes region aStandardized beta, b Student’s t test, * p < 0.05, ** p < 0.01 Treatment expenses for women by type of treatment in FCFA*, for women with positive expenses (n = 174) * 1 FCFA = 0,00212 USD The limit of significance for statistical tests was set at ≤0.05. Analyses were performed using SPSS version 20 (SPSS Inc., Chicago, IL) and JMP software package version 5.1 (SAS institute INC, Cary, NC, USA). This study was approved by the research ethics committee of the University of Montreal Hospital Research Centre (CRCHUM, Canada) and by the ethics committee of the Faculty of Medicine, Pharmacy and Dentistry of the University of Bamako (Mali).
N/A
