Artigo Carlota de F. Lelis Et All - 04 - 09 - 2023 - ENG - VERSION
Artigo Carlota de F. Lelis Et All - 04 - 09 - 2023 - ENG - VERSION
Artigo Carlota de F. Lelis Et All - 04 - 09 - 2023 - ENG - VERSION
ORCID: 0001-6482-2562
carlotalelis@yahoo.com.br
ORCID: 0002-0222-8768
prietsch@gmail.com
ORCID: 0003-0864-0486
juraci.a.cesar@gmail.com
1
ABSTRACT
The study aims to estimate the proportion of puerperae with an unplanned pregnancy,
evaluate trends and identify factors associated with its occurrence in Rio Grande (RS),
residing in the municipality in 2007, 2010, 2013, 2016 and 2019. The chi-square test
compared proportions and the Poisson regression with robust variance adjustment in the
multivariate analysis. The prevalence ratio (RP) was the effect measure employed. The
study includes 12,415 puerperae (98% of the total). The unplanned pregnancy rate was
pregnancy were observed among younger, black women, living without a partner, with
multiparous and smokers. The rate of unplanned pregnancy is high and stable, with a
higher propensity among women those with the highest risk of unfavorable events
during pregnancy and childbirth. Reaching these women in high schools, companies,
services and health professionals, in addition to the mass media, can be strategies to
2
INTRODUCTION
the most recent estimate available, were unplanned. 1 This type of pregnancy is
associated with late initiation of appointments, inadequate prenatal care, and adverse
outcomes such as low birth weight, prematurity, need for induced labor, and longer
without a partner, of low socioeconomic status, and very often victims of intimate
partner violence.6,7 For these reasons, not planning a pregnancy is a global public health
issue.8
million births/year.9 Among Brazilian women, besides the factors already mentioned,
also identified as significantly associated with this outcome are Brown, Yellow, and
and tobacco use.9-12 The most recent of these studies dates from 2020 and included only
one hospital in eight Brazilian capitals.12 The only one to consider trends was conducted
in Pelotas, RS, finding an unplanned pregnancy rate of 62.7% in 1993, 65.9% in 2004,
and 52.2% in 2015.11 Since then, no other study evaluating trends in Brazil has been
The Rio Grande Perinatal Studies, conducted every three years since 2007,
collected a range of information from the six months before pregnancy to the immediate
postpartum period. Among this information is one addressing pregnancy planning. Five
This article aims to measure prevalence, assess trends, and identify factors
3
METHODS
The present study was conducted in Rio Grande (RS), Brazil, from 2007 to 2019.
This municipality is located on the south coastline, about 300 km from Porto Alegre and
250 km from the border with Uruguay. Its population increased from 195 thousand to
212 thousand during this period. Its economy is based on trade and agribusiness, mainly
rice production and extensive livestock, fertilizer industries, and fishing and port
activities.
The public health network has changed little over these 13 years, with 36 PHC
units, four specialty centers, and two general hospitals, one with mixed care and the
other exclusively dedicated to the Unified Health System (SUS). The Human
Development Index (HDI) reached 0.744, while infant mortality increased from 9.3 to
11/1000 births.14
The data presented in this article are nested in the Rio Grande Perinatal Studies,
surveys conducted in 2007, 2010, 2013, 2016, and 2019 to monitor the care offered
during pregnancy and childbirth in the municipality. The inclusion criteria were having
had a child between January 1 and December 31 in those years and having reached at
The design is cross-sectional, and the respondents were approached only once,
within 48 hours after delivery while still in the hospital. Moreover, it is a census study
because it includes all puerperae living in urban and rural areas. 15 Data were collected
sought information from pregnancy planning to the immediate postpartum period. Most
of the questions and variables originated from blocks D, E, and F for this study. These
4
born alive or dead, age at first pregnancy, and first delivery), maternal lifestyle and
behavior (tobacco use and alcohol/coffee/mate consumption, and physical activity), and
residents in the household, degree of kinship, age, schooling, occupation, and individual
income in the month immediately preceding the interview). Further details on the
The questionnaires were applied by four interviewers trained for 40 hours, who
underwent a pilot study in the month immediately before the beginning of data
collection. Daily, they visited the maternity ward and all the wards of each hospital in
search of births whose mothers resided in Rio Grande. When meeting a puerperae, the
interviewer explained that the study invited her to participate. She signed two copies of
the Informed Consent Form (ICF) if accepted. One copy was delivered to the mother,
while the other was filed at the Faculty of Medicine of the Federal University of Rio
Grande (FURG) project headquarters. The interview was started only after this step.
The 2007, 2010, and 2013 surveys were based on physical questionnaires, in
which the interviewers coded and revised the closed-ended questions at the end of each
working day. The following day, the questionnaires were delivered to the project
headquarters, where they were revised and entered twice by different professionals and
in the reverse order of the first. Entries were compared at each block of 100
questionnaires and, if necessary, corrected. All these steps were performed using the
free software Epi Data16 and Epi Info.17 In the 2016 and 2019 surveys, data were entered
simultaneously with the interview using tablets and the REDCap (Research Electronic
Data Capture) application.18 At the end of each day, these questionnaires were
5
The outcome of this study was defined based on the negative response to the
following question: “Did you plan to have this child, or did you accidentally become
pregnant?”. All puerperae who answered “not having planned” or “having become
The initial analysis consisted of listing the frequency of the variables of interest.
Then, a bivariate analysis was performed to verify the distribution of the outcome
concerning different exposures, which was evaluated using Pearson’s chi-square test.
robust variance adjustment.19,20 This last stage obeyed a previously defined hierarchical
model with the independent variables allocated in three levels: distal (demographic and
(behavioral and lifestyle). The variables located at a hierarchically higher level than the
variable in question were considered potential confounders vis-à-vis the outcome, in this
case, not planning the pregnancy. The p-value associated with the outcome should be
≤0.2021 to be maintained in the model in the adjusted analysis. The effect measure was
the prevalence ratio for a 95% confidence interval (95%CI). Wald tests for
heterogeneity and linear trends were used for ordinal exposures. 19 All these analyses
Approximately 10% of the interviews were partially repeated within 15 days after
the initial interview to evaluate the agreement of the answers provided by the mothers
shortly after birth. The Kappa index of agreement ranged from 0.61 to 0.99, with most
protocols under the following numbers: 2007 (Opinion N° 05369/2006); 2010 (Opinion
6
N° 06258/2009); 2013 (Opinion N° 02623/2012); 2016 (Opinion N° 0030-2015) and
RESULTS
A total of 12,663 puerperae were identified in the five surveys conducted in Rio
Table 1 shows the proportional distribution of mothers who did not plan the
proportion of this condition among younger mothers (≤ 19 years old), less educated (≤ 8
study years), and belonging to the first and second quartile of household income.
residents in the household, higher parity, and smokers. The penultimate column on the
right of this same table shows the variation between the extremes (2007 and 2019). The
In contrast, the smallest decline occurred among those with 12 or more schooling
years, reaching 8.1%. The last column shows the linear trend test. Only one of the
categories of variables included in this table (age 20-29 years) was not statistically
Table 2 shows that only the “unemployed partner” and “previous abortions”
variables lost statistical significance after controlling for confounders among all the
variables in the table. All other variables had an independent effect on the outcome.
Thus, puerperae aged 11-19, Black, who lived with a partner, with up to eight schooling
years, and belonging to the worst quartile of income showed a higher PR of not having
7
family planning than those aged 30 years or more, White, living with a partner, with
12+ schooling years, and belonging to the best income quartile, respectively. Mothers
who did not engage in paid work during pregnancy, had two or more children, and
smoked before or during pregnancy showed a higher PR than the others for not planning
the pregnancy. Finally, the greater the number of residents in the household, the greater
8
DISCUSSION
Practically two out of three Rio Grande women did not plan the pregnancy, and
this rate was high at the end of the period, especially among those at greater risk for
unfavorable events during pregnancy and childbirth. The adjusted analysis showed that
younger puerperae, Black, living without a partner, with more significant household
crowding, lower schooling and household income, higher parity, and smokers had the
similar to that observed in the baseline of the 1993 Pelotas cohort (62.7%), slightly
lower than in 2004 (65.9%), but higher than the 2015 rate (52.2%). 10 In São Luís,
Maranhão, this prevalence reached 68.1% in 2010 24, while it was 53.8% in Ribeirão
Preto, São Paulo.25 In Brazil, a hospital-based study conducted between 2011-12 found a
eight university hospitals identified 67.5%.12 This high level in different locations
indicates that this problem requires actions at different levels of public service
management. Besides the health sector, education should also be included since most of
The younger the age, the greater the proportion of parturients who did not plan the
pregnancy in Rio Grande. This pattern was also identified in other studies. It can be
immaturity, difficulty accessing safe contraceptive methods, concern about the side
Black mothers also showed a higher PR for not planning the pregnancy,
mentioned, Black mothers struggle more in accessing health services that are often
9
insufficient and have inadequate quality.28-31 This situation indicates structural
is not recent, and is not easily perceived by clients. However, it will require efforts from
civil society and, above all, the Federal Government through policies and
programs.28,30,31
The lower the level of education and the worse the income, the greater the PR for
not planning the pregnancy. No single factor has as many benefits to maternal and child
information to care for their health and their children’s, acquire argumentative skills,
become better aware of their rights, and start to participate more effectively in social
life, including holding positions in the labor market. 4,27 By doing so, they increase
household income, improve purchasing power, and their living conditions and their
family’s. These advances undoubtedly contribute to better planning of their lives, which
includes choosing the most appropriate moment to become a mother 27,31, which helps
explain the higher PR observed for not planning a pregnancy among those with the
Not having a partner was a risk factor for unplanned pregnancy. Unstable
relationships, even with the same partner, hinder the continued use of contraceptive
methods, which increases exposure to pregnancy. 5,24 In Ribeirão Preto, São Paulo, the
odds ratio for not planning a pregnancy among women without a partner was 7.56
(95%CI: 5.98-9.56) times higher than the others.25 In Rio Grande perinatal surveys, the
outcomes.11,13 This analysis reinforces the importance of a partner, often the biological
10
The possibility of not planning the pregnancy also increases with a higher number
of residents in the household. The dose-response effect was evident, even more so than
the values observed for other variables, such as household income and maternal
education. In environments with many people, as they are responsible for most
household chores, women often neglect their health care. 27 This situation may lead to the
irregular use of contraceptive methods, especially when used daily, such as oral
contraceptives, which are widely adopted. 23 The incorrect use of contraceptives can
Smoking was an independent risk factor for not planning pregnancy. The Birth in
Brazil Study also identified an odds ratio of 1.23 (95%CI: 1.06-1.43) for not planning a
pregnancy among puerperal smokers.9 Few studies have shown this association.
Considering that it is widely known that smoking is harmful to the health of both the
mother and the fetus and even so some women continue to smoke, the lack of family
planning may indicate a lack of concern for their health. It would be interesting to
evaluate this finding in other studies. In any case, in Rio Grande, tobacco use appeared
by health teams.
When interpreting these results, we should consider that data were collected over
13 years, during which essential changes occurred in the local economic setting, from
the results presented, at least until 2013. However, these changes do not change the
direction or eliminate the results found. It is essential to highlight that this work has the
shortest interval between data collection ever carried out in Brazil. It included all
11
puerperae in a medium-sized municipality with a response rate of at least 96% in the
those at greater risk of adverse events during pregnancy and childbirth. We observed an
evident overload among the most vulnerable, further increasing this group’s
campaigns and the availability of safe, reversible contraceptive methods, with adequate
guidance from public health services to prevent unplanned pregnancies, which would
12
References
2. Dibaba Y, Fantahun M, Hindin MJ. The effects of pregnancy intention on the use of
antenatal care services: systematic review and meta-analysis. Reprod Health. 2013;
10:50.
pregnant and low birth weight and preterm birth: a systematic review. Matern Child
Health J. 2011;15:205-216.
neonatal outcomes: findings of new high-risk newborns in Peru. Int Health. 2023;
21:ihad018.
Unplanned pregnancy and the association with maternal health and pregnancy
6. Goossens J, Van Den Branden Y, Van der Sluys L, Delbaere I, Van Hecke A,
al. Pregnancy intendedness and the association with physical, sexual and emotional
2015;15:120.
13
9. Theme-Filha MM, Baldisserotto ML, Fraga AC, Ayers S, da Gama SG, Leal MD.
2016;13(Suppl 3):118.
10. Moreira LR, Ewerling F, Dos Santos IS, Wehrmeister FC, Matijasevich A, Barros
AJD, Menezes AMB, Gonçalves H, Murray J, Domingues MR, Silveira MF. Trends
Pelotas, Brazil. Int J Public Health. 2020. Int J Public Health. 2020;65:1635-1645.
11. Prietsch SO, González-Chica DA, Cesar JA, Mendoza-Sassi RA. Gravidez não
Publica. 2011;27:1906-16.
12. Nilson TV, Amato AA, Resende CN, Primo WQSP, Nomura RMY, Costa ML, et al.
13. Cesar JA, Mendoza-Sassi RA, Marmitt LP. Evolução da assistência à gestação e ao
<https://cidades.ibge.gov.br/brasil/rs/rio-grande/panorama>.
15. Silva IS. Cancer epidemiology: principles and methods. Lyon: World Health
16. Lauritsen JM. (Ed.) EpiData Data Entry, Data Management and basic Statistical
em: http://www.epidata.dk.
14
17. Dean A, Arner T, Sunki G, Friedman R, Lantinga M, Sangam S, et al. Epi InfoTM,
a database and statistics program for public health professionals. Atlanta, GA:
CDC; 2011.
18. Harris PA, Taylor R, Thielke R, Payne J, Gonzalez N, Conde JG. Research
19. Kirkwood BR, Sterne JAC. Essentials of medical statistics. 2ª Ed. London:
20. Barros AJ, Hirakata VN. Alternatives for logistic regression in cross-sectional
21. Victora CG, Huttly SR, Fuchs SC, Olinto MT. The role of conceptual frameworks
7.
22. StataCorp. Stata Statistical Software: release 11.2. College Station: StataCorp LP;
2011.
24. Costa ACM, Oliveira BLCA, Alves MTSS. Prevalence and factors associated with
25. Vieira CS, Braga GC, Lugarinho PTC, Stifani BM, Bettiol H, Barbieri MA,
Cardoso VC, Cavalli RC. Sociodemographic factors and prenatal care behaviors
15
26. Fundo das Nações Unidas para a População (UNFPA). Relatório sobre a Situação
em: https://brazil.unfpa.org/pt-br/publications/situacao-da-populacao-mundial-
2022.
27. Ameyaw EK, Budu E, Sambah F, Baatiema L, Appiah F, Seidu AA, Ahinkorah BO.
14:0220970.
28. Pacheco VC, Silva JC, Mariussi AP, Lima MR, Silva TR. As influências da raça/cor
29. Victora CG, Matijasevich A, Silveira MF, Santos IS, Barros AJD, Barros FC. Socio-
economic and ethnic group inequities in antenatal care quality in the public and
31. Victora CG, Barreto ML, Leal MC, Monteiro CA, Schmidt MI, Paim J, Bastos FI,
Working Group. Health conditions and health-policy innovations in Brazil: the way
16
Box 1. Hierarchical model of analysis for unplanned pregnancy
Level Characteristics
(variables)
Demographic: Socioeconomic:
17
Table 1. Distribution of puerperae for some of their characteristics by unplanned
Survey year
Characteristics 2007 (%) 2010 (%) 2013 (%) 2016 (%) 2019 (%) 2007-19 (%) p-tend
Maternal age
11-19 72.6 (374) 74.2 (327) 75.9 (346) 73.0 (327) 82.6 (247) +13.8 <0.001
20-29 61.5 (816) 63.4 (785) 62.4 (826) 61.1 (808) 68.6 (787) +11.5 0.163
30-47 59.9 (407) 57.9 (391) 56.0 (470) 52.5 (461) 59.0 (486) -1.5 <0.001
Skin color
White 60.2 (1059) 62.4 (1022) 59.3 (1025) 58.5 (1042) 64.6 (1121) +7.3 <0.001
Brown 68.8 (318) 65.8 (320) 67.1 (393) 61.5 (368) 74.2 (256) +7.8 <0.001
Black 73.1 (220) 70.0 (161) 73.4 (224) 68.9 (186) 75.3 (143) +3.0 <0.001
Living with partner 58.7 (1224) 59.3 (1162) 58.7 (1318) 55.6 (1231) 63.0 (1218) +7.3 <0.001
Schooling (years)
0-8 67.7 (833) 71.7 (764) 71.4 (745) 66.6 (647) 78.1 (554) +15.4 <0.001
9-11 60.5 (638) 59.5 (624) 59.0 (691) 63.3 (665) 67.5 (723) +11.6 <0.001
12+ 53.2 (126) 47.5 (115) 51.0 (206) 45.4 (284) 49.6 (243) -6.8 <0.001
Household monthly income
(quartiles)
1º (worst) 69.7 (428) 72.8 (399) 71.8 (535) 69.8 (467) 78.9 (448) +13.2 <0.001
2º 63.9 (419) 64.9 (351) 62.4 (339) 65.5 (431) 70.8 (408) +10.8 <0.001
3º 64.7 (372) 61.7 (322) 61.8 (400) 57.2 (419) 66.8 (356) +3.2 <0.001
4º (best) 54.0 (329) 53.6 (281) 51.7 (325) 40.9 (174) 49.6 (265) -8.1 <0.001
Mothers who worked during
pregnancy 60.6 (572) 59.0 (595) 58.5 (670) 55.2 (667) 61.2 (591) +1.0 <0.001
Unemployed partner 61.7 (1275) 61.7 (1240) 61.2 (1436) 56.7 (1187) 63.5 (1162) +2.9 <0.001
Household residents
2 47.8 (354) 47.8 (385) 50.2 (455) 47.9 (487) 58.9 (352) +23.2 <0.001
3 58.3 (398) 62.7 (421) 60.3 (476) 59.1 (481) 61.0 (487) +4.6 <0.001
4+ 76.9 (845) 79.4 (697) 77.0 (711) 76.9 (628) 78.0 (681) +1.4 <0.001
Parity
1 58.3 (581) 55.6 (569) 56.1 (694) 55.3 (634) 62.1 (533) +6.5 <0.001
2 58.0 (505) 63.5 (544) 64.5 (702) 56.2 (532) 62.4 (526) +7.6 <0.001
3+ 78.0 (511) 82.1 (390) 84.0 (246) 77.5 (430) 81.2 (461) +4.1 <0.001
Previous abortions 58.1 (281) 57.7 (184) 60.1 (236) 56.9 (230) 59.5 (213) +2.4 <0.001
Tobacco use before or
during pregnancy 73.8 (428) 73.9 (362) 74.4 (360) 70.2 (236) 81.7 (228) +10.7 <0.001
Unplanned pregnancy
prevalence 63.3 (1597) 63.8 (1503) 62.7 (1642) 60.3 (1596) 67.0 (1520) +5.8 <0.001
N 2,523 2,355 2,619 2,648 2,270 12,415
Total Mean prevalence (and 95% CI) from 2007 to 2019: 63.3% (95% CI: 62.5%-64.1%)
18
Table 2. Crude and adjusted analyses for factors associated with unplanned
19
Adjusted variables in each level:
- Level I: maternal age, skin color, living with a partner, residents per household, maternal schooling,
household income, worked during pregnancy, and employed partner.
- Level II: maternal age, skin color, living with partner, residents per household, maternal schooling,
household income, worked during pregnancy, parity, and previous abortions.
- Level III: maternal age, skin color, living with a partner, residents per household, maternal schooling,
household income, mother worked during pregnancy, parity, and maternal tobacco use before
pregnancy.
* Wald’s test for heterogeneity
** Wald’s test for linear trend
20