Prenatal care in Southern Brazil: coverage, trend and disparities

ABSTRACT OBJECTIVE To estimate coverage, examine trend and assess the disparity reduction regarding household income during prenatal care between mothers living in Rio Grande, state of Rio Grande do Sul, in 2007, 2010, 2013 and 2016. METHODS This study included all recent mothers living in this municipality, between 1/1 and 12/31 of those years, who had a child weighing more than 500 grams or 20 weeks of gestational age in one of the only two local maternity hospitals. Trained interviewers applied, still in the hospital and up to 48 hours after delivery, a unique and standardized questionnaire, seeking to investigate maternal demographic and reproductive characteristics, the socioeconomic conditions of the family and the assistance received during pregnancy and childbirth. To assess the adequacy of prenatal care, the criteria proposed by Takeda were used, which considers only the number of prenatal appointments and gestational age at initiation, and by Silveira et al., who in addition to these two variables, considers the achievement of some laboratory tests. Chi-square tests were used to compare proportions and assess the linear trend. RESULTS The total of 10,669 recent mothers were included in this survey (96.8% of the total). Prenatal coverage substantially increased between 2007 and 2016. According to Takeda, it rose from 69% to 80%, while for Silveira et al., it increased from 21% to 55%. This improvement occurred for all income groups (p < 0.01). The disparity between the extreme categories of income reduced, according to Takeda, and increased according to Silveira et al. CONCLUSIONS The provision of prenatal care, considering only the number of appointments and the early start, occurred in greater proportion among the poorest. However, only the richest recent mothers were contemplated with more elaborate care, such as laboratory tests, which increased the disparities in the provision of prenatal care.


INTRODUCTION
Prenatal care reduces the occurrence of intrauterine growth restriction, low birth weight, prematurity, neonatal mortality and malnutrition in children 1,2 . The better the quality of the services offered, the greater the impact on maternal and child health and the lower the morbidity and mortality rate in this population 1,3 .
This quality has been measured by the combination of care offered to pregnant women, including, in general, number of appointments, gestational age at initiation of the prenatal care and, more recently, laboratory tests. As a rule, three categories were originated: adequate, intermediate and inadequate [4][5][6] .
Several studies have reported the existence of inequality in prenatal care [7][8][9][10][11] . Most of them have done this after the identification of associated factors and found that pregnant women with the worst socioeconomic status are more likely to receive inadequate care [5][6][7][8][9][10] . Other studies have made more sophisticated analyses and shown that the inequality is decreasing because the proportion of pregnant women beginning prenatal care is increasing, and the number of appointments made by them approaches, more and more, those with a better socioeconomic level. The reduction in the difference between the rich and the poor suggests that the inequality in prenatal care has been decreasing 11,12 .
Perinatal studies conducted every three years in the municipality of Rio Grande, state of Rio Grande do Sul, since 2007, have been collecting information about the care received by pregnant women since the beginning of prenatal until the immediate postpartum. These data allow the completion of this assessment in a period of 10 years (2007-2016).
This study has three objectives: 1) estimate the adequate prenatal coverage according to the criteria proposed by Takeda andSilveira et al. between 2007 and in this municipality, 2) meet the trend of this adequacy and 3) verify how this coverage behaved during this period according to the household income.

METHODS
The municipality of Rio Grande is located in the coastal stretch of the south of Rio Grande do Sul state, 300 km from Porto Alegre. Between 2007 and 2016, the data collection period, its population increased from 197 thousand to about 213 thousand inhabitants. Throughout this period, the gross domestic product (GDP) increased from R$19.6 to R$41.3 thousand 13 , while the employment relationship doubled, going from 52,875 to 107.338 14 . The infant mortality coefficient remained around 11.5/1,000 15 , while little changed in the public health system, except for the Family Health Strategy (FHS), which went from 10 to 36 teams over the period. The same services remained in the others: 32 basic health units, three outpatient clinics and two hospitals, the University Hospital of the Universidade Federal do Rio Grande being destined entirely to the Unified Health System. This research is part of the Perinatal Studies of Rio Grande, conducted in the municipality every three years since 2007 to assess the care received by mothers from the beginning of pregnancy until immediate postpartum. Data presented here have a time series that included all births that occurred between 1/1 and 12 /31 2007, 2010, 2013 and 2016 in only two local maternity hospitals. To be included in the study, mothers should reside in the rural or urban area of the municipality of Rio Grande, and their children should have at least 500 grams or 20 weeks of gestational age at birth.
The sample size was estimated a posteriori based on the prevalence of 81.3 for Takeda 4 in 2016 and 43.8 for Silveira et al. 5 in 2010. Thus, from the "n" available (10,331), for a 95% confidence level, it would be possible to work with a margin of error of 2.0 and 1.0 percentage point, respectively 16 .
Data collection was performed using a single and standardized questionnaire, divided into blocks, which sought information about demographic, occupational, behavioral, and reproductive characteristics, as well as complications in this last pregnancy. We also collected information on housing conditions, possession of durable products, family crowding and socioeconomic level of the family.
The outcome, completion of adequate prenatal, was constituted from the criteria proposed by Takeda 4 and Silveira et al. 5 Takeda considered adequate when pregnant women began attending the appointments before the fifth month of pregnancy and attended six or more prenatal care appointments, while Silveira et al. classified as appropriate when pregnant women who attended six or more appointments began prenatal care before the fifth month of pregnancy and did two or more qualitative tests of urine, complete blood count and exams for the diagnosis of syphilis by the rapid test or the Venereal Disease Research Laboratory (VDRL).
Recent mothers were approached only once, in the maternity hospital, up to 48 hours after delivery. In the surveys of 2007, 2010 and 2013, printed and coded questionnaires were used by the interviewers and delivered at the headquarters of the project for further review, double typing, comparison and correction in EpiData 3.1 17 and EpiInfo 6.04 16 software. In 2016, the data were entered using tablets by the interviewers themselves via online Platform Web REDCap (Research Electronic Data Capture) 18 with the server of the Universidade Federal do Rio Grande (FURG) at the address www.redcap.furg.br. The data entered were verified on the next day, when any inconsistencies were corrected.
The consistency and categorization of variables were analyzed in Stata 11 19 . Then the variables were assembled to assess the prenatal adequacy according to the criteria proposed by Takeda 4 and Silveira et al. 5 The proportions and the linear trend between 2007 and 2016 were assessed using chi-square test and trend tests, respectively. The confidence level employed was 95%. For quality control, 7% of the interviews were redone, i.e. every 15 questionnaires, one was chosen at random to be partially repeated. This mini questionnaire contained questions from all blocks of questions of the questionnaire. The answers were then compared. The agreement was assessed

RESULTS
In 2007, 2010, 2013 and 2016, the number of births was 10,669, and the mothers lived in the municipality of Rio Grande. Of this total, 10,331 (96.8%) mothers were interviewed. Table 1 shows the main socioeconomic characteristics of recent mothers included in these surveys. Between 2007 and 2016, the rate of teenage mothers decreased 16% and labors among those of brown skin color increased 23%. The proportion of those who lived with a partner remained stable at around 84%. The schooling rate increased, especially among mothers who have graduated high school (≥ 12 years). The household income has improved, especially in categories of lower (up to 0.99 minimum wage) and higher (four minimum wages or more) income. The proportion of mothers who worked during pregnancy increased 23% over the period. The absence of prenatal appointments decreased and there was a gain, on average, of one prenatal appointment in the period.

DISCUSSION
This study showed an increase in the coverage of adequate prenatal care in all income categories of both criteria used. It also showed the disparity reduction according to Takeda and increase regarding the criterion proposed by Silveira et al.
As to the assessment of the adequacy of prenatal care using the criterion by Takeda  Regarding the trend, for both criteria, we found that the provision of adequate prenatal care has increased consistently in the municipality, which is in agreement with the results observed in the cohorts of Pelotas 25 , São Luís and Ribeirão Preto. 9 However, even for Takeda, undemanding criterion to achieve adequacy, there is room for improvement, while for Silveira et al., more robust criterion that requires two laboratory tests, the challenge is much bigger because almost half of them (45%) have not reached adequacy yet.
Despite the increase in the provision of adequate prenatal in all income categories in Rio Grande, we verified reduction and increase in disparity according to Takeda and Silveira et al., respectively. In the case of Takeda, the reduction was basically due to the expansion of the Family Health Strategy in the municipality, which went from 10 teams in 2007 to 36 teams in 2016, increasing the coverage of FHS from 38% to 57% over the period 27 . These teams were primarily installed in neglected areas or with insufficient capacity to meet the demand for prenatal care in these locals.
The fact that only the poor have a low number of appointments or do not receive prenatal care decreased the difference between extreme categories, reducing significantly the inequality in prenatal care, finding already evidenced in other studies 11,12 .
According to Silveira et  and 2016 in Rio Grande, the coverage for these three tests among poor women increased from 91% to 94%, while it increased from 93% to 99% among the rich. These data may explain the increased disparity between these groups, since those with higher income achieved practically the universalization, while the poor had less evident improvement.
Finally, the increase in disparity regarding prenatal care found in Rio Grande is not in accordance with other studies on the subject 12,30 . A possible explanation is that other studies assessed only the number of appointments, but not the complementary exams as our study.
If this analysis included only the criterion proposed by Takeda, restricted to the number of appointments and gestational age at initiation of prenatal care, the result would be similar.
When interpreting these results, one should consider the limitations inherent to the design used and the fact that they include questions since the planning of pregnancy until the immediate postpartum, that the questionnaire was applied a few hours after delivery, and that it took between 60 and 80 minutes. In addition, we included a technical question related to the kind of disparity assessed: our aim was to assess the difference between extremes, so we did not considerate the intermediate groups 31 . The assessment of intermediate groups requires more complex analysis of difficult interpretation, which would differ from our goal to denounce the increasing disparity between income extremes.
The provision of adequate prenatal care has been increasing consistently, but we must prioritize the provision of care for pregnant women with the worst household income. We suggest that other variables are included besides the number of appointments when assessing the adequacy of prenatal care, in order to provide a more in-depth analysis about the prenatal care, extremely relevant to the health care of mother and child.