Evaluating the effect of Bolsa Familia, Brazil's conditional cash transfer programme, on maternal and child health: A study protocol.

BACKGROUND: Conditional Cash Transfer Programs have been developed in Latin America in response to poverty and marked social inequalities on the continent. In Brazil, the Bolsa Familia Program (BFP) was implemented to alleviate poverty and improve living conditions, health, and education for socioeconomically vulnerable populations. However, the effect of this intervention on maternal and child health is not well understood.
METHODS: We will evaluate the effect of BFP on maternal and child outcomes: 1. Birth weight; 2. Preterm birth; 3. Maternal mortality; and 4. Child growth. Dynamic retrospective cohort data from the 100 Million Brazilian Cohort (2001 to 2015) will be linked to three different databases: Live Birth Information System (2004 to 2015); Mortality Information System (2011 to 2015); and Food and Nutritional Surveillance System (2008 to 2017). The definition of exposure to the BFP varies according to the outcome studied. Those who never received the benefit until the outcome or until the end of the follow-up will be defined as not exposed. The effects of BFP on maternal and child outcomes will be estimated by a combination of propensity score-based methods and weighted logistic regressions. The analyses will be further stratified to reflect changes in the benefit entitlement before and after 2012. DISCUSSION: Harnessing a large linked administrative cohort allows us to assess the effect of the BFP on maternal and child health, while considering a wide range of explanatory and confounding variables.

Background

Poverty and social inequality have been identified as major social causes of poor health, requiring public policies and strategies to eradicate poverty and improve the most vulnerable populations’ living and health conditions [1-3]. Despite the advances observed on maternal and child health in the last decades, the slow decline in maternal mortality and the persistence of adverse outcomes, such as preterm birth (PTB), low birth weight (LBW), and child malnutrition, especially among the low and -middle-income countries (LMIC), hinder the achievement of the Sustainable Development Goals (SDGs) [4-8]. In Brazil, the maternal mortality ratio was 59.7 per 1,000 live births in 2015 (a 57% decline in 25 years), and the national prevalence of PTB and LBW were respectively 11.1%, and 8.4% [9]. Malnutrition estimates in children under five years old enrolled in the Bolsa Familia Program (BFP), a Brazilian Conditional Cash Transfer (CCT), showed a high prevalence of stunting (12.7%) and overweight/obesity (18.4%) in 2014 [10].

CCTs have been adopted as a strategy to promote maternal and child health [11-13]. Programs focused on combating immediate and future poverty may improve access to health, education, social assistance, employment, and income [14]. The BFP is one of the oldest and largest CCTs in the world, with over 13.2 million beneficiary families, corresponding to 96% coverage of the country’s poor households (estimates for February 2020) [15]. While Brazil was one of the pioneers in implementing the CCT in Latin America, there is still little evidence on the effect of BFP on maternal and child outcomes, especially from studies using robust methods and large-scale individual-level data with an extended period of follow-up [11, 12, 14]. Understanding the health and health equity impacts of social policies is important to inform policymaking, including decisions about ongoing investment in these schemes [12, 14, 16, 17].

The most important contribution of the proposed research will be developing robust evidence of the effect of the BFP on maternal and child outcomes, using a cohort which allow us assessing more robust statistical analyzes in the general population and separately for specific subpopulations.

Methods

Primary objective, study design, and overall population

We aim to evaluate the effect of BFP on maternal and child outcomes in the 100 Million Brazilian Cohort [18]. The main objective of the Cohort is to enable the study of the social determinants and the effects of social policies and programs on the different aspects of health in Brazil [18]. It is a dynamic retrospective cohort, the population of which is derived from more than 114 million individual records from the Single Registry for Federal Government’s Social Programs (CadÚnico). The cohort contains administrative records from CadÚnico and the BFP Payroll. CadÚnico identifies and characterizes low-income households and registration is required in order to receive any Federal Government’s social programs, such as the BFP [19]. The Cohort allows us to extract socioeconomic information from the individual, the household, and data related to receiving the benefit. The detailed variables and databases to be used are shown in Table 1.

Table 1

Structure and main components of the 100 Million Brazilian Cohort, sources of data, and relevant variables.

Components	Data source	Period	Number of Records	Relevant variables
Cohort Baseline	Single Registry (CadÚnico)	2001 to 2015	114,008,317	Socioeconomic and demographic conditions (information on family dynamics, childcare arrangements, parental employment, income, housing family formation, dissolution, social programs information, household characteristics).
Intervention (Exposure)	Bolsa Familia Program (BFP)	2004 to 2015	27,376,582	Start and end of data receipt of benefit, total value by family, and number of months received.
Outcomes	Live Birth Information System (SINASC)	2001 to 2015	44,485,274	Characteristics of the newborn (sex, Apgar score in the 1 and 5 minutes, birth weight, presence of an abnormality, congenital anomalies identified at birth), characteristics of the mother (age, marital status, education, race, place of residence), characteristics of pregnancy and delivery (number of previous pregnancies of live births, stillbirth or abortion, gestational age, place of birth, type of delivery, number of fetuses, number of prenatal visits, month that started prenatal). Some variables such as the month in which the woman started prenatal care and gestational age (continuous) are only available for the period from 2011 to 2015.
Outcomes	Mortality Information System (SIM)	2000 to 2015	17,829,111	Type of death, date of death, date of birth, sex, race, education, duration of the pregnancy, single or multiple pregnancies, type of delivery, age of mother, gestational age, birth weight, and death cause.
Outcomes	Food and Nutrition Surveillance System (SISVAN)	2008 to 2017	307,245,508	Date of birth, age, sex, race/ethnicity, traditional communities, anthropometric data (weight and height), measurement date, presence of chronic diseases (diabetes and cardiovascular diseases), and deficiencies and complications (diarrhea and anemia).

The primary objective will be achieved by linking the Cohort (2001 to 2015) and data from (i) the Live Birth Information System (SINASC) (2004 to 2015); (ii) the Mortality Information System (SIM) (2011 to 2015); and (iii) the Food and Nutrition Surveillance System (SISVAN) (2008 to 2017). We will use CIDACS Record Linkage (CIDACS-RL) to link the databases [20]. The linkage procedures are common for the 100 Million Cohort studies and consist of two stages. The first will be a deterministic linkage, and the second will be based on the similarity index. More detailed information can be consulted in previous publications [21, 22]. The CIDACS-RL is a tool for linking individual records based on identifiers: name, gender, age or date of birth, mother’s name, and the municipality of residence [22]. All linking procedures will be performed at CIDACS (Center for Data Integration and Knowledge for Health, Fiocruz) [23] in a strict data protection environment and complying with ethical and legal standards [24].

The Bolsa Familia Program (BFP)

We describe the policy in accordance with the TIDieR-PHP reporting guideline [25]. The checklist consists of nine items and helps researchers to describe the characteristics of population health and policy interventions. The BFP was implemented from a national decree in 2004, with eligibility criteria (poverty and extreme poverty cutoff points) and incorporation of benefits that varied over time [26-31]. The cut-off points and the eligibility criteria are shown in Table 2. The selection of households eligible for the BFP occurs through enrollment in the CadÚnico [26, 31]. Households served by the BFP receive a monthly cash benefit through a withdrawal card issued by the Caixa Econômica Federal [32].

Table 2

Changes in the eligibility criteria and inclusion of new groups of beneficiaries.

Year	Extreme poverty* (R$)	Poverty* (R$)	Inclusion of new groups (varying benefits)
2004	50.00	100.00	No change
2006	60.00	120.00	No change
2009	70.00	140.00	Concession of benefits to households with adolescents aged 16–17 years enrolled in education institutions
2012	No change	No change	Concession of benefits to households with children aged zero to six. Concession of varying benefits to pregnant women and nursing mothers
2014	77.00	154.00	No change
2016	85.00	170.00	No change
2018	89.00	178.00	No change

* Household units with a per capita household income less than or equal to the mentioned value.

The BFP is equipped with fraud prevention control mechanisms, with public access to beneficiaries’ individual data over the internet and semiannual comparison of CadÚnico’s enrolled data with other databases [21]. The suspension of households from BFP can occur due to failure to update the registration information, no longer fitting the profile (eligibility criteria), and noncompliance with conditionalities [32]. The program’s conditionalities are geared to participation in education, health, and social assistance. In the field of health, conditionalities include actions, such as immunization, prenatal care, and child growth monitoring [26, 31].

Logic models

We created a logic model to describe the hypothesized mechanisms through which the BFP might affect maternal and child outcomes (Fig 1). The socioeconomic characteristics can influence both the receipt of the benefit and maternal and child health outcomes [33-39]. Characteristics of particular relevance include targeting monetary resources preferentially to women and the fulfillment of conditionalities. Despite not being a guarantee, the BFP may increases women’s decision-making power [40], has the potential to transform women into heads of households with responsibility for directing the money received. The transfer of income to women can have a more immediate effect on maternal and child health outcomes, with female empowerment, the allocation of money for the purchase of food, and the use of health services [41-46].

Fig 1

Logical model of the impact of the Bolsa Familia Program (BFP) in reducing adverse maternal and child outcomes.

On the other hand, BFP also requires the fulfillment of conditionalities, using services during pregnancy, puerperium, and early childhood [26, 31]. Using health services is an important determinant of maternal and child outcomes [4, 47–54] because it can have an immediate effect on these outcomes, with immunization, nutritional counseling and preventive behaviors during prenatal care, monitoring of comorbidities, and connected to the place of birth [47, 55–57]. The reduction of adverse maternal and child outcomes depends on joint efforts that ensure access to quality health services and lower social inequalities [12, 16, 17, 34, 35, 58].

Secondary objectives, study population, definition of exposure, and outcomes

The definitions of outcomes, study populations, and exposure to BFP will be presented separately (detailed information in Chart 1, as supplementary material), according to the objectives:

assess the effect of BFP on birth weight, small and large for gestational age (SGA/LGA) and on preterm births

evaluate the effect of BFP on maternal mortality

assess the effect of BFP on child malnutrition.

i) Birth weight, SGA, LGA, and preterm birth

Study population. The study will include baseline data from the “100 Million Brazilian Cohort” linked to SINASC (Table 1). The study population will consist of the first and the second live birth of women registered in the cohort baseline, from 2004 to 2015, with ages ranging from 10 to 49 years. The study population for the SGA, LGA and PTB refers to 2012 to 2015 period, due to the inclusion gestational age as a continuous variable in 2011.

Multiple births and newborns with congenital anomalies will be excluded to avoid bias, as these conditions are known to be strongly associated with low birth weight and PTB [52, 59–61]. Fetal viability criteria can be applied [62-65]. Regarding birth weight, the inclusion of the first live birth is a strategy to capture the effect of receiving the BFP during the first and the second pregnancy. Ordering the live births will allow us to select/extract the population of interest and obtain previous birth information such as inter-birth interval, low birth weight, and preterm birth.

Since nulliparous women are at increased risk of chronic and acute medical and obstetrical complications leading to preterm birth, we will restrict the population related to PTB to singleton live births whose mothers in reproductive age have at least one child before joining the Cohort. Only the firstborn after enrolment will be included.

Exposure to BFP. The exposure is defined as having started receiving BFP before the birth of their child in the 2004 to 2015 period (or 2012 to 2015 for SGA, LGA and PTB) and did not stop receiving from pregnancy to delivery. Live births of women who did not receive the benefit at any time until delivery will be considered as not exposed.

Outcomes. Birth weight will be considered as (1) birth weight, in grams (continuous variable), and (2) birth weight categorized into very low, low, normal, and high (see Table 3) [66].

Table 3

Description of the outcomes that will be considered in studies by assessing the impact of the Bolsa Familia Program (BFP).

Objective	Original variables used to construct the outcome	Outcome
To evaluate the effect of BFP on birth weight, small and large for gestational age and preterm birth	Birth weight in grams	Birth weight in grams (continuous variable)
		Adequate birth weight (≥2500g) vs. low birth weight (<2500g)
		Adequate weight (2500-3999g) vs. extremely low weight (<1000g), very low weight (1000-1499g), low birth weight (1500-2499g) and macrosomia (≥4000g)
	Weight in grams and Gestational age in full weeks (available from 2011)	Adequate for gestational age (between 10th and 90th percentiles) vs. Small for gestational age (<10th percentile) and Large for gestational age (>90th percentile)
		Extreme weights for gestational age: 10th to 90th percentile vs. <3rd percentile; 3rd to 9th percentile, 91st to 97th percentile and >97th percentile
	Gestational age in categories	non-preterm birth (≥37 gestational weeks) vs. preterm birth (<37 gestational weeks)
		Non-PTB vs. moderate-to-late PTB (32 to 36 gestational weeks), very PTB (28–31 gestational weeks) e extreme PTB (< 28 gestational weeks)
To assess the effect of BFP on maternal mortality	Underlying cause of death	Non-death vs. death of a woman during pregnancy or up to 42 days after the end of pregnancy, due to any cause related to or aggravated by the pregnancy, but not due to accidental or incidental causes.
	Intermediate cause of death
To assess the effect of BFP on child malnutrition	Length/height in centimeters, age in months, and sex	Height-for-age z-score (HAZ)
		HAZ ≥ –2 (benchmark) vs. HAZ < –2 (stunting)
		HAZ ≥ –2 (benchmark) vs. HAZ <-3 (severe stunting) and HAZ ≥ –3 to HAZ < –2 (moderate stunting)
	Weight in grams, age in months, and gender	Weight-for-age z-score (WAZ)
		WAZ ≥ –2 to ≤ +2 (benchmark) vs. WAZ < –2 (underweight)
		WAZ ≥ –2 to ≤ +2 (benchmark) vs. WAZ < –3 (severe underweight) and WAZ ≥ –3 to < –2 (moderate underweight)
	Weight in grams, length/height in centimeters, and gender	Weight-for-height z-score (WHZ)
		WHZ ≥ –2 and ≤ +2 (benchmark) vs. WHZ < –2 (wasting)
		WHZ ≥ –2 and ≤ +2 (benchmark) vs. WHZ < –3 (severe wasting) and WHZ ≥ –3 and < –2 (moderate wasting)
		WHZ ≥ –2 and ≤ +2 (benchmark) vs. WHZ > +2 (overweight/obesity)
		WHZ ≥ –2 and ≤ +2 (benchmark) vs. WHZ > +3 (obesity) and WHZ ≤ +3 to > +2 (overweight)

Small for Gestational Age will be defined as birth weight according to gestational age and gender below the 10th percentile; Adequate for Gestational Age, between the 10th and 90th percentiles; and Large for Gestational Age, above the 90th percentile [66, 67]. Categories will also be explored, including weight extremes for gestational age (Table 3).

Preterm birth will be defined as 1. PTB (22 to <37 gestational weeks) vs. non-PTB (37 to 42 gestational weeks); and 2. stratified (Table 3), according to the degree of severity [66].

ii) Maternal mortality

Study population. The study will include data from 100 Million Brazilian Cohort linked to SINASC and SIM. The study population will consist of women of reproductive age (10 to 49 years) according the surveillance criteria in Brazil, registered in the Cohort baseline, in their last pregnancy in the 2004 to 2015 period.

Exposure to BFP. The exposure is defined as having started receiving the BFP before or during pregnancy and did not stop receiving the benefit before the outcome or until childbirth. Women who have not received the benefit at any time until childbirth or the puerperium will be considered as not exposed.

Outcome. Maternal death will be defined as the death of women during pregnancy or up to 42 days after the end of pregnancy, due to any cause related to or aggravated by the pregnancy, but not due to accidental or incidental causes. We will evaluate the follow outcome according the International Classification of Diseases–ICD-10: ICD-10 “XV” codes will be considered (Pregnancy, childbirth and the puerperium (O00-O99) to compose cases of maternal death, except for deaths after 42 days, “O96” and “O97”; and other ICD-10 chapters (A34, F53, M83.0, B20 to B24, D39.2, and E23.0) [68].

iii) Child malnutrition

Study population. The study will include data from the 100 Million Brazilian Cohort linked to SISVAN and SINASC. The study population will consist of children aged 0 to 5 years registered in the cohort baseline from 2004 to 2015. Anthropometric information from the last visit in the 2008 to 2017 period will be used.

Definition of exposure. Exposure in the studied population will consist of children who started and did not stop receiving the BFP before the last visit (2008 to 2017) to answer the objective of interest. Those not exposed will be the ones who have not received the benefit at any time until the date of the child’s last visit.

Outcome. Nutritional status in children will be computed according to the WHO growth references and cutoff points for standardized height-for-age z-score (HAZ), weight-for-age z-score (WAZ), and weight-for-height z-score (WHZ) [69]. Anthropometric indices will be considered as continuous and categorized measures (Table 3).

Statistical analysis

The effect of BFP on birth weight, preterm birth, maternal mortality, and child growth will be estimated based on propensity score-based methods (PS). The PS can be characterized as the conditional probability of receiving the treatment (to be a BFP beneficiary or not), given its observable characteristics [70]. These methods are different from the others in that they avoid multidimensionality and can be implemented using a control variable, which is the propensity score itself [46].

First, the missing data patterns will be evaluated for the variables considered in the calculation of the PS. Depending on these analyses, the PS calculation can be performed only with complete data or including the missing data as a category in each variable. The PS will be estimated using a logit model with baseline covariables related to receipt of BFP according Chart 1 (supplementary material).

The models will be weighted by the Inverse Probability Treatment Weighting (IPTW) and by the Kernel weights. Balancing will be performed before and after weighting to ensure that the procedure used controlled for the available confounders. Finally, weighted Logistic Regressions and the Average Treatment Effect on Treated (ATT) will be calculated using non-linear and linear models, depending on the analyzed outcome [71, 72].

Robustness analysis for propensity score-based methods

As it is a dynamic cohort, analyses will be considered according to the treatment exposure time. Supplementary analyses will also be carried out with subpopulations with similar lengths of time since entering the cohort to balance the time until the outcome between recipients and controls. Also, analyses will be carried out for municipalities with a higher quality of information from vital statistics and according to the quantiles of coverage of the Family Health Strategy, region of residence and the decentralized (municipal) management index (IGD) of the BFP; and for subpopulations of maternal reproductive age (15 to 49 years or 10 to 49 years) [73, 74] and prenatal care follow-up.

Ethical considerations

The Research Ethics Committee of the Institute of Collective Health, Federal University of Bahia (ICS-UFBA), approved the studies involved in this protocol under Opinion N° CAAE: 41695415.0.0000.5030 on May 30, 2017.

The linkage of the databases will be carried out in a secure environment, following a strict internal information security procedure to ensure data privacy and confidentiality [21]. A non-identified database will be used for the analyses, which can only be accessed by previously authorized researchers, and all steps after obtaining the data will be carried out following the CIDACS information security culture.

Discussion

This study will use propensity score-based methods to assess the BFP effect on maternal and child health outcomes in a large sample of poor and impoverished Brazilian households. The BFP might be expected to result in positive effects in all conditions related to difficulties in accessing health, education, social assistance, employment, and income, thus, improving maternal and child health conditions. The study will follow internationally recognized guidelines for conducting and disseminating the results of impact assessment studies, providing transparency in conducting data analysis, and greater comparability of results [25, 75, 76].

Some limitations must be considered. Information systems can include missing data and lack of relevant information on potential outcome and confounding variables, such as more specialized access and quality of prenatal or postnatal care indicators, which could allow a better understanding of the nuances of the intervention (for example, distance to the clinic or ability and training of health professionals). We will not explore the results of the BFP concerning the amount of the transfers granted. BFP is a binary variable in our study, and nuances related to the amount received and poverty levels will not be explored in this first proposal.

On the other hand, the large-scale data set will allow us to investigate comprehensively and in subpopulations the effects of BFP on maternal and child outcomes. The use of these databases will allow us exploring rarer outcomes with a high level of statistical power. The databases used in this study have national coverage, low under-registration, and some have already documented reliability [59]. Thus, this study will provide a comprehensive and representative analysis of the poor and extremely poor Brazilian population and reinforce the adequacy of these bases for epidemiological investigations [59]. The availability of a cohort with socioeconomic information linked to maternal and child health data provides us with the possibility to assess the effect of the BFP on these outcomes, considering a wide range of explanatory and confounding variables, and enabling the use of methods based on propensity scores.

Dissemination of knowledge

This evaluation of BFP will provide tools and evidence to program management focused on poverty reduction and reduction of adverse outcomes related to maternal and child health. We will disseminate the data in scientific journals, reports, and policy briefings targeting policymakers and civil society.

(DOCX)

Click here for additional data file.

4 Nov 2021

Evaluating the impact of Bolsa Familia, Brazil’s conditional cash transfer programme, on maternal and child health: a study protocol

PLOS ONE

Dear Dr. Falcão,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

The present study protocol sits amid important research questions and production of quality evidence regarding a major social program and maternal and child outcomes. The rationale is also well stated. However, as consistently pointed out by the reviewers, there are several methodological aspects that need to be clarified, including steps for data management and availability, statistical procedures (also considering the huge sample size), and approaches for sensitivity analysis --all crucial for ensuring reproducibility. In addition, please note that more detail on the planned developments may be useful to highlight the need to register this study protocol, as opposed to bringing such information in the methodology section of derived original studies (for instance, by the same research group: https://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.1003509).

Please submit your revised manuscript by Dec 17 2021 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Bárbara Hatzlhoffer Lourenço, Ph.D.

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. Please include your tables as part of your main manuscript and remove the individual files. Please note that supplementary tables (should remain/ be uploaded) as separate "supporting information" files"

3. We note that the grant information you provided in the ‘Funding Information’ and ‘Financial Disclosure’ sections do not match.

When you resubmit, please ensure that you provide the correct grant numbers for the awards you received for your study in the ‘Funding Information’ section.

4. Thank you for stating the following financial disclosure: "The funders had and will not have a role in study design, data collection and analysis, decision to publish, or preparation of the manuscript."

At this time, please address the following queries:

a) Please clarify the sources of funding (financial or material support) for your study. List the grants or organizations that supported your study, including funding received from your institution.

b) State what role the funders took in the study. If the funders had no role in your study, please state: “The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.”

c) If any authors received a salary from any of your funders, please state which authors and which funders.

d) If you did not receive any funding for this study, please state: “The authors received no specific funding for this work.”

Please include your amended statements within your cover letter; we will change the online submission form on your behalf.

5. In your Data Availability statement, you have not specified where the minimal data set underlying the results described in your manuscript can be found. PLOS defines a study's minimal data set as the underlying data used to reach the conclusions drawn in the manuscript and any additional data required to replicate the reported study findings in their entirety. All PLOS journals require that the minimal data set be made fully available. For more information about our data policy, please see http://journals.plos.org/plosone/s/data-availability.

Upon re-submitting your revised manuscript, please upload your study’s minimal underlying data set as either Supporting Information files or to a stable, public repository and include the relevant URLs, DOIs, or accession numbers within your revised cover letter. For a list of acceptable repositories, please see http://journals.plos.org/plosone/s/data-availability#loc-recommended-repositories. Any potentially identifying patient information must be fully anonymized.

Important: If there are ethical or legal restrictions to sharing your data publicly, please explain these restrictions in detail. Please see our guidelines for more information on what we consider unacceptable restrictions to publicly sharing data: http://journals.plos.org/plosone/s/data-availability#loc-unacceptable-data-access-restrictions. Note that it is not acceptable for the authors to be the sole named individuals responsible for ensuring data access.

We will update your Data Availability statement to reflect the information you provide in your cover letter.

6. We note that you have indicated that data from this study are available upon request. PLOS only allows data to be available upon request if there are legal or ethical restrictions on sharing data publicly. For more information on unacceptable data access restrictions, please see http://journals.plos.org/plosone/s/data-availability#loc-unacceptable-data-access-restrictions.

In your revised cover letter, please address the following prompts:

a) If there are ethical or legal restrictions on sharing a de-identified data set, please explain them in detail (e.g., data contain potentially sensitive information, data are owned by a third-party organization, etc.) and who has imposed them (e.g., an ethics committee). Please also provide contact information for a data access committee, ethics committee, or other institutional body to which data requests may be sent.

b) If there are no restrictions, please upload the minimal anonymized data set necessary to replicate your study findings as either Supporting Information files or to a stable, public repository and provide us with the relevant URLs, DOIs, or accession numbers. For a list of acceptable repositories, please see http://journals.plos.org/plosone/s/data-availability#loc-recommended-repositories.

We will update your Data Availability statement on your behalf to reflect the information you provide.

7. Your abstract cannot contain citations. Please only include citations in the body text of the manuscript, and ensure that they remain in ascending numerical order on first mention.

8. Your ethics statement should only appear in the Methods section of your manuscript. If your ethics statement is written in any section besides the Methods, please delete it from any other section.

9. Please upload a new copy of Figure 1 as the detail is not clear. Please follow the link for more information: " ext-link-type="uri" xlink:type="simple">https://blogs.plos.org/plos/2019/06/looking-good-tips-for-creating-your-plos-figures-graphics/" " ext-link-type="uri" xlink:type="simple">https://blogs.plos.org/plos/2019/06/looking-good-tips-for-creating-your-plos-figures-graphics/"

10. Please include captions for your Supporting Information files at the end of your manuscript, and update any in-text citations to match accordingly. Please see our Supporting Information guidelines for more information: http://journals.plos.org/plosone/s/supporting-information.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Does the manuscript provide a valid rationale for the proposed study, with clearly identified and justified research questions?

The research question outlined is expected to address a valid academic problem or topic and contribute to the base of knowledge in the field.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

**********

2. Is the protocol technically sound and planned in a manner that will lead to a meaningful outcome and allow testing the stated hypotheses?

The manuscript should describe the methods in sufficient detail to prevent undisclosed flexibility in the experimental procedure or analysis pipeline, including sufficient outcome-neutral conditions (e.g. necessary controls, absence of floor or ceiling effects) to test the proposed hypotheses and a statistical power analysis where applicable. As there may be aspects of the methodology and analysis which can only be refined once the work is undertaken, authors should outline potential assumptions and explicitly describe what aspects of the proposed analyses, if any, are exploratory.

Reviewer #1: Yes

Reviewer #2: Partly

Reviewer #3: Partly

**********

3. Is the methodology feasible and described in sufficient detail to allow the work to be replicable?

Descriptions of methods and materials in the protocol should be reported in sufficient detail for another researcher to reproduce all experiments and analyses. The protocol should describe the appropriate controls, sample size calculations, and replication needed to ensure that the data are robust and reproducible.

Reviewer #1: No

Reviewer #2: No

Reviewer #3: Yes

**********

4. Have the authors described where all data underlying the findings will be made available when the study is complete?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception, at the time of publication. The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: No

Reviewer #3: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above and, if applicable, provide comments about issues authors must address before this protocol can be accepted for publication. You may also include additional comments for the author, including concerns about research or publication ethics.

You may also provide optional suggestions and comments to authors that they might find helpful in planning their study.

(Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: Review:

Thank you for this informative paper on your proposed study protocol for exploring the impact of Brazil’s Bolsa Familia Program on maternal and child health outcomes. I appreciate the time you’ve put into describing your protocols and methods before undertaking a study, which is important for transparency in scientific research.

Suggested revisions:

1. In the first paragraph of the ‘Background’ section (lines 53-57), I’m not sure if this sentence refers to Brazil, lower-middle-income/higher-middle-income countries, globally. Please clarify.

2. In the first paragraph of the ‘Methods’ section, could you briefly elaborate on the original purpose of the 100 Million Brazilian Cohort survey.

3. In the second paragraph of the ‘Methods’ section, I find the sentence on lines 97-98 unclear. Can you briefly explain the two stages? Is this unique to your study? Or is this standard procedure for linking surveys to government data?

4. The second paragraph of the ‘Methods’ section is generally a bit disjointed with the CIDACS acronym being defined at the end, CIDACS-RL being mentioned, then linkages, then an explanation of CIDACS-RL. It could be improved for the ease of comprehension.

5. Line 105, it would be helpful to general readers to briefly explain the purpose of TIDieR-PHP reporting guidelines and why you used them.

6. Line 167, is definition b) not covered by definition a)? If so, then definition b) is redundant. If not, please rephrase b) for better clarity.

7. Line 180, do you want a ‘2)’ before “…stratified…”? Is it not a second definition?

8. You’ve mentioned Regression Discontinuity Design (RDD) in your keywords. I see no reference to RDD analyses in the main text. How do you plan to use RDD with your data? What questions do you hope to answer?

9. Rationale for stratification/sub-analyses of samples post-2011 is not clearly laid out in the text. It is briefly explained under “iii) child malnutrition - study population”. But post-2011 sub-analyses are suggested before that without explanation (LGA, SGA and prematurity). I figure this is due to changes in the BFP, but this is not clearly laid out in the text.

10. Table 1:

• In the text you refer to the Bolsa Familia Program, but in the table, it is the Family Grant Program. I suggest you use Bolsa Familia here for consistency.

• Under relevant variables from SINASC, it says “month that started prenatal after 2011”. Do you mean prenatal classes? Prenatal vitamins? Please check this.

• Are all the main variables you intend to use in your analyses listed in this table? It would be very important to know other health-related variable of the mothers (i.e. pre-existing diabetes or gestational diabetes; smoking; etc.) as these will be relevant confounders for prematurity, LGA and SGA analyses.

11. Table 3: There is an extra ‘e’ in front of “extremely preterm” in the outcomes column.

12. Figure 1: Resolution needs to be checked as it is barely readable at the moment.

For further consideration:

Your proposed rationale is reasonable, but have you considered how cash transfers that are conditional and preferentially paid to mothers may not increase purchasing power/ empowerment for all women. For example, the responsibility of getting children to school and to regular medical appointments for working mothers with partners may further entrench domestic/care work as women’s roles – potentially at the expense of paid employment, social networks, self-care, etc. Indeed, there appears to be a heterogeneous effect of CCT on women’s empowerment that may need further consideration in additional analyses:

De Brauw, A., Gilligan, D. O., Hoddinott, J., Roy, S. (2014). The impact of Bolsa Família on women’s decision-making power. World Development, 59, 487-504.

Reviewer #2: It is not clear to me that what the manuscript describes warrants a study protocol. The construction of the database itself has been published elsewhere by the same group. All aspects described in the section “Secondary objectives, study population, definition of exposure, and outcomes” are minor and would be well suited to methodology sections of different papers. The statistical methods are solid for natural experiment studies in public health. I firmly believe that the authors should expand the details of their methodological decisions and processes in a future submission and publish the product of this development as a supplemental file to their methodology.

Nevertheless, if the authors decide to proceed with the submission, some key points should be addressed.

Keeping in mind that reproducibility is one of the main pillars of study protocols at PLOS ONE, the authors should provide a comprehensive background of how the databases that compose the 100 Million Brazilian Cohort can be accessed for research purposes. If only governmental officials can access the data, the authors should consider another type of publication for this manuscript.

Given the 100 Million Brazilian complexity, the author should also expand on how they plan to address bias in all three outcomes.

A more thorough explanation should be provided concerning data cleaning decisions and the linkage process.

Minor points to be addressed:

Some aspects of the study population, exposure to PBF, and outcome should be standardized between sections. For example, picking either the “2004-2015” or “2004 to 2015” to declare year ranges.

In the logic model, “Linkage to the place of birth” is not a product but a process and does not belong to this logic model.

Reviewer #3: “Evaluating the Impact of Bolsa Familia, Brazil’s conditional cash transfer programme on, on maternal and child health: a study protocol,” submitted to PLOS ONE (PONE-08640)

The protocol outlines a substantive analysis using various large administrative health and social program databases from Brazil to the so-called 100 million Brazilian cohort. It represents an ambitious research agenda (likely leading to multiple papers) that has potential to improve understanding of the influences of PBF. The authors make clear that despite its enormous size and importance, careful empirical assessment of the effect of PBF, particularly on child and maternal outcomes, is sparse. They identify an appropriate set of outcomes based on available data and the literature. There are likely to have sufficient power for even very small impacts and rare events such as maternal mortality (making it important to judge not only statistical impact but size of impact). The large sample sizes and observational nature of the data make the research design, i.e., arguments for assessing causal impact and not just associations, crucial.

Main Comments:

1. One important reason there is not more evidence on PBF is the lack of a strong research design for assessing program impacts, as was done for example via randomization of Progresa in Mexico. Another challenge when examining impact at a national level, I believe, are the differences in program administration across municipalities played. The team proposes resolving identification of the causal effects of the intervention via propensity score matching techniques. This approach is preferable to simpler comparisons but still relies on key assumptions of non-confoundedness across treatment and comparison groups after matching. After controlling for the observed factors available, the assumption is that there are no unobserved differences in those taking up the treatment and those not taking it up. Central problem is that those who enter, despite observed characteristics, might be different – ie more likely to benefit or have unobserved wealth or something we do not observe. So, sign of bias is difficult to ascertain. If untreated are better off on other characteristics, for example, might be able to argue results are conservative or underestimates of beneficial program benefits. Unconfoundedness cannot be proven but the matching literature provides various approaches for assessing it in the articles cited and my expectation is many of these will be done in your analyses.

2. In practice, carrying out PS or other matching techniques on these large data sets will involve dozens if not hundreds of decisions regarding the specifics (on which variables, functional form, common support etc.) and possibly variations on those decisions to assess sensitivity. It could be useful to say a bit more about how this will be approached, including the specifics of the data for readers less familiar with it – for example specific variables/measures that are included or links to those descriptions or an appendix.

3. In their work (and related approaches), Imbens and coauthors develop other types of matching such as Nearest Neighbor (implemented in Stata using the command nnmatch). It may not be feasible with such large data sets to follow those approaches but a key aspect of them is allowing “exact” matching on certain types of characteristics. One that may be particularly important in this context is location – I noted mention of some subgroup analyses but think my suggestion here is a little different. Taking for example geographic location, to help ensure important elements like potential differences in municipality health systems are not leading to bias, a strategy of only matching treated cases with untreated in the same municipality can be used in the overall analysis. This permits an arguably better comparison than allowing geographic location, for example, to enter only via the combined propensity score.

4. Because PBF had an income cut-off, I did wonder whether there was any scope for an alternative approach to identification, related to regression discontinuity designs (RDD). I believe this could be done in conjunction with ps matching, but it would require availability of income measures (but these appear to be available). Or if not explicit, limiting comparison samples to those with incomes nearer to the cutoff, for example.

5. Administrative data match quality: I am unfamiliar with the various administrative data the study will use. It has been my experience in other settings, however, that combining administrative data across systems can be error ridden. To that end, greater support for the case that merging administrative records across the data sets is feasible and result in high quality (and high %) matches would strengthen confidence in the research design and the ultimate findings. Differences in quality of administrative match across the different data sets may influence findings in the three domains differently. It was unclear to me what the “similarity index” (page 3) approach was, but I presume on subsets of information (eg birth date, gender, location but not quite exact name spelling). A clear distinction in the final papers between the administrative matching across data sets and the ps matching procedures needs to be made. Characteristics of those matched and those not matched could shed light on potential biases.

6. Are there any statistical considerations relevant to having particularly large sample sizes?

7. It was unclear to me whether length of exposure (beyond pregnancy periods) for outcomes would be considered, but I may have missed this.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

Reviewer #3: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

8 Feb 2022

To the Editor and reviewers,

First, we would like to express our appreciation for your consideration of our manuscript, especially in light of the crisis the world is currently experiencing. We are very grateful for the pertinent criticism offered, and believe the incorporation of the reviewer’s suggestions will greatly contribute to the quality of our publication. Please find our point-by-point responses below to the criticism raised by each reviewer:

Editor:

The present study protocol sits amid important research questions and production of quality evidence regarding a major social program and maternal and child outcomes. The rationale is also well stated.

However, as consistently pointed out by the reviewers, there are several methodological aspects that need to be clarified, including steps for data management and availability, statistical procedures (also considering the huge sample size), and approaches for sensitivity analysis --all crucial for ensuring reproducibility.

In addition, please note that more detail on the planned developments may be useful to highlight the need to register this study protocol, as opposed to bringing such information in the methodology section of derived original studies (for instance, by the same research group: https://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.1003509).

- Please find our point-by-point responses below to the criticism raised by each reviewer.

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. Please include your tables as part of your main manuscript and remove the individual files. Please note that supplementary tables (should remain/ be uploaded) as separate "supporting information" files"

- We apologize for this oversight, and we have included the tables and figures in manuscript.

3. We note that the grant information you provided in the ‘Funding Information’ and ‘Financial Disclosure’ sections do not match.

When you resubmit, please ensure that you provide the correct grant numbers for the awards you received for your study in the ‘Funding Information’ section.

- We apologize for this oversight, and we have changed this information.

4. Thank you for stating the following financial disclosure: "The funders had and will not have a role in study design, data collection and analysis, decision to publish, or preparation of the manuscript."

- We appreciate your comment. We have included the financial disclosure (ls. 362-64).

At this time, please address the following queries:

a) Please clarify the sources of funding (financial or material support) for your study. List the grants or organizations that supported your study, including funding received from your institution.

b) State what role the funders took in the study. If the funders had no role in your study, please state: “The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.”

c) If any authors received a salary from any of your funders, please state which authors and which funders.

d) If you did not receive any funding for this study, please state: “The authors received no specific funding for this work.”

- Thank you for your comment. We have revised and separated in "Funding information" the individual and institutional fundings.

Please include your amended statements within your cover letter; we will change the online submission form on your behalf.

- We have added the statements to the cover letter as requested.

5. In your Data Availability statement, you have not specified where the minimal data set underlying the results described in your manuscript can be found. PLOS defines a study's minimal data set as the underlying data used to reach the conclusions drawn in the manuscript and any additional data required to replicate the reported study findings in their entirety. All PLOS journals require that the minimal data set be made fully available. For more information about our data policy, please see http://journals.plos.org/plosone/s/data-availability.

- This study protocol will use linked data from the 100 Million Brazilian Cohort, coordinated and housed by the Center for Data and Knowledge Integration for Health (CIDACS). In accordance with the policy of CIDACS and the Ministry of Health and Ministry of Citizenship (database providers), restrictions on the availability of this data apply. Currently, only national and international researchers who collaborate with CIDACS and authorized staff from government agencies can access de-identified or anonymized linked data. Any person who wishes to receive authorization must: (i) be affiliated to CIDACS or be accepted as collaborators; (ii) present a detailed research project together with approval by an appropriate Brazilian institutional research ethical committee; (iii) provide a clear data plan restricted to the objectives of the proposed study and a summary of the analyses plan intended to guide the linkage and or data extraction of the relevant set of records and variables; (iv) sign terms of responsibility regarding the access and use of data; and (v) perform the analyses of datasets provided using the CIDACS data environment, a safe and secure infrastructure that provides remote access to de-identified or anonymized datasets and analysis tools. For more information, please visit the CIDACS website [https://cidacs.bahia.fiocruz.br/] or contact us via email [cidacs@fiocruz.br].

Upon re-submitting your revised manuscript, please upload your study’s minimal underlying data set as either Supporting Information files or to a stable, public repository and include the relevant URLs, DOIs, or accession numbers within your revised cover letter. For a list of acceptable repositories, please see http://journals.plos.org/plosone/s/data-availability#loc-recommended-repositories. Any potentially identifying patient information must be fully anonymized.

Important: If there are ethical or legal restrictions to sharing your data publicly, please explain these restrictions in detail. Please see our guidelines for more information on what we consider unacceptable restrictions to publicly sharing data: http://journals.plos.org/plosone/s/data-availability#loc-unacceptable-data-access-restrictions. Note that it is not acceptable for the authors to be the sole named individuals responsible for ensuring data access.

- We updated and included the Data Availability in the cover letter.

6. We note that you have indicated that data from this study are available upon request. PLOS only allows data to be available upon request if there are legal or ethical restrictions on sharing data publicly. For more information on unacceptable data access restrictions, please see http://journals.plos.org/plosone/s/data-availability#loc-unacceptable-data-access-restrictions.

In your revised cover letter, please address the following prompts:

a) If there are ethical or legal restrictions on sharing a de-identified data set, please explain them in detail (e.g., data contain potentially sensitive information, data are owned by a third-party organization, etc.) and who has imposed them (e.g., an ethics committee). Please also provide contact information for a data access committee, ethics committee, or other institutional body to which data requests may be sent.

- This study protocol will use linked data from the 100 Million Brazilian Cohort, coordinated and housed by the Center for Data and Knowledge Integration for Health (CIDACS). The cohort is composed of admistrative databases, which the owners, the Brazil’s Ministry of Health and Ministry of Citizenship, have given CIDACS the custody and authorization to conduct research, with the guarantee that all data processing takes place in a safe and private environment. Therefore, importante restrictions apply to the availability of this data.

Currently, only national and international researchers who collaborate with CIDACS and authorized staff from government agencies can access de-identified or anonymized linked data. Any person who wishes to receive authorization must: (i) be affiliated to CIDACS or be accepted as collaborators; (ii) present a detailed research project together with approval by an appropriate Brazilian institutional research ethical committee; (iii) provide a clear data plan restricted to the objectives of the proposed study and a summary of the analyses plan intended to guide the linkage and or data extraction of the relevant set of records and variables; (iv) sign terms of responsibility regarding the access and use of data; and (v) perform the analyses of datasets provided using the CIDACS data environment, a safe and secure infrastructure that provides remote access to de-identified or anonymized datasets and analysis tools. For more information, please visit the CIDACS website [https://cidacs.bahia.fiocruz.br/] or contact us via email [cidacs@fiocruz.br].

This study protocol has already received the approval and authorization from CIDACS. In addition, the Research Ethics Committee of the Institute of Collective Health, Federal University of Bahia (ICS-UFBA), approved the study under Opinion N° CAAE: 41695415.0.0000.5030. All steps subsequent to obtaining the data will be carried out following the CIDACS information security protocols.

b) If there are no restrictions, please upload the minimal anonymized data set necessary to replicate your study findings as either Supporting Information files or to a stable, public repository and provide us with the relevant URLs, DOIs, or accession numbers. For a list of acceptable repositories, please see http://journals.plos.org/plosone/s/data-availability#loc-recommended-repositories.

- In line with the previous answer, it is important to note that there are restrictions that apply to the availability of these data, which are licensed for use in the present study only and therefore are not publicly available. We updated the Data Availability statement as requested avove. Recently, an article was published that gives more details about the databases used to build the 100 Million Brazilian Cohort (doi: 10.1093/ije/dyab213)

7. Your abstract cannot contain citations. Please only include citations in the body text of the manuscript, and ensure that they remain in ascending numerical order on first mention.

- We apologize for this oversight, and we have removed this citation from the abstract.

8. Your ethics statement should only appear in the Methods section of your manuscript. If your ethics statement is written in any section besides the Methods, please delete it from any other section.

- Thank you for your comment. Our ethics statement only appears in the Methods section and in Declarations, according Plos One requirements.

9. Please upload a new copy of Figure 1 as the detail is not clear. Please follow the link for more information: https://blogs.plos.org/plos/2019/06/looking-good-tips-for-creating-your-plos-figures-graphics/" https://blogs.plos.org/plos/2019/06/looking-good-tips-for-creating-your-plos-figures-graphics/"

- We added a new copy of Figure 1 into the manuscript.

10. Please include captions for your Supporting Information files at the end of your manuscript, and update any in-text citations to match accordingly. Please see our Supporting Information guidelines for more information: http://journals.plos.org/plosone/s/supporting-information.

- We included captions for our Supporting Information files at the end of our manuscript.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Does the manuscript provide a valid rationale for the proposed study, with clearly identified and justified research questions?

The research question outlined is expected to address a valid academic problem or topic and contribute to the base of knowledge in the field.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

- We thank the reviewers for his/her careful reading of our text.

________________________________________

2. Is the protocol technically sound and planned in a manner that will lead to a meaningful outcome and allow testing the stated hypotheses?

The manuscript should describe the methods in sufficient detail to prevent undisclosed flexibility in the experimental procedure or analysis pipeline, including sufficient outcome-neutral conditions (e.g. necessary controls, absence of floor or ceiling effects) to test the proposed hypotheses and a statistical power analysis where applicable. As there may be aspects of the methodology and analysis which can only be refined once the work is undertaken, authors should outline potential assumptions and explicitly describe what aspects of the proposed analyses, if any, are exploratory.

Reviewer #1: Yes

Reviewer #2: Partly

Reviewer #3: Partly

- We appreciate your important comment. We describe all the objectives of the studies to be carried out. In addition, we present a logic model to describe the hypothetical mechanisms by which BFP may affect maternal and infant outcomes (Figure 1).

________________________________________

3. Is the methodology feasible and described in sufficient detail to allow the work to be replicable?

Descriptions of methods and materials in the protocol should be reported in sufficient detail for another researcher to reproduce all experiments and analyses. The protocol should describe the appropriate controls, sample size calculations, and replication needed to ensure that the data are robust and reproducible.

Reviewer #1: No

Reviewer #2: No

Reviewer #3: Yes

- Working with population databases gives us many possibilities to assess specific subpopulations and explore hypotheses, but it has some limitations. We will work with a database of people eligible for social programs, which can be up to 90% beneficiaries and 10% non-beneficiaries. Thus, we do not have a sample, we cannot previously define a number for the control group, and we need appropriate methods. Thus, we believe that matching may not be the best way, but the methods based on Propensity Score (weighting), could be applied.

________________________________________

4. Have the authors described where all data underlying the findings will be made available when the study is complete?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception, at the time of publication. The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: No

Reviewer #3: Yes

- As mentioned earlier, it is important to note that there are restrictions that apply to availability of such data, which are licensed for use only in the present study and therefore not able to make available publicly. Recently, two articles were published that provide more details on the databases used to build the 100 Million Brazilian Cohort (doi: 10.1093/ije/dyab213) and CIDACS Birth Cohort (DOI: 10.1093/ije/dyaa255)

________________________________________

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

- We thank the reviewers.

________________________________________

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above and, if applicable, provide comments about issues authors must address before this protocol can be accepted for publication. You may also include additional comments for the author, including concerns about research or publication ethics.

You may also provide optional suggestions and comments to authors that they might find helpful in planning their study.

(Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: Review:

Thank you for this informative paper on your proposed study protocol for exploring the impact of Brazil’s Bolsa Familia Program on maternal and child health outcomes. I appreciate the time you’ve put into describing your protocols and methods before undertaking a study, which is important for transparency in scientific research.

Suggested revisions:

1. In the first paragraph of the ‘Background’ section (lines 53-57), I’m not sure if this sentence refers to Brazil, lower-middle-income/higher-middle-income countries, globally. Please clarify.

- We appreciate the reviewer’s comments. The sentence refers to low and-middle-income countries. The sentence in question incorporated this notion: “…low birth weight, and child malnutrition, especially among the low and -middle-income countries (LMIC), hinder the achievement of the Sustainable Development Goals (SDGs) [4-8].”

2. In the first paragraph of the ‘Methods’ section, could you briefly elaborate on the original purpose of the 100 Million Brazilian Cohort survey.

- We agree with this suggestion. We have added appropriate citations in the text: “The main objective of the 100 million Brazilian cohort is to enable the study of the social determinants and the effects of social policies and programs on the different aspects of health in Brazil (https://doi.org/10.1093/ije/dyab213). ”

3. In the second paragraph of the ‘Methods’ section, I find the sentence on lines 97-98 unclear. Can you briefly explain the two stages? Is this unique to your study? Or is this standard procedure for linking surveys to government data?

- The linkage procedures are common for 100 Million Brazilian Cohort studies. We have included other recent publication and the mention: “We will use CIDACS Record Linkage (CIDACS-RL) to link the databases (https://doi.org/10.1186/s12911-020-01285-w). The linkage procedures consist of two stages. The first will be a deterministic linkage, and the second will be based on the similarity index. More detailed information can be consulted in previous publications (https://doi.org/10.1186/s12911-020-01192-0) (https://doi.org/10.3389/fphar.2019.00984).”

4. The second paragraph of the ‘Methods’ section is generally a bit disjointed with the CIDACS acronym being defined at the end, CIDACS-RL being mentioned, then linkages, then an explanation of CIDACS-RL. It could be improved for the ease of comprehension.

- We apologize for this oversight and have added an explanation in line 100.

5. Line 105, it would be helpful to general readers to briefly explain the purpose of TIDieR-PHP reporting guidelines and why you used them.

- We have added an explanation in ls. 113-4. The checklist consists of 9 items and helps researchers to describe the characteristics of population health and policy interventions.

6. Line 167, is definition b) not covered by definition a)? If so, then definition b) is redundant. If not, please rephrase b) for better clarity.

- Thank you. We have removed the definition b).

7. Line 180, do you want a ‘2)’ before “…stratified…”? Is it not a second definition?

- We have added the “2” before “…stratified…”.

8. You’ve mentioned Regression Discontinuity Design (RDD) in your keywords. I see no reference to RDD analyses in the main text. How do you plan to use RDD with your data? What questions do you hope to answer?

- We have removed RDD among the keywords.

9. Rationale for stratification/sub-analyses of samples post-2011 is not clearly laid out in the text. It is briefly explained under “iii) child malnutrition - study population”. But post-2011 sub-analyses are suggested before that without explanation (LGA, SGA and prematurity). I figure this is due to changes in the BFP, but this is not clearly laid out in the text.

- We agree with your comment. After discussion with the team, we came to the conclusion that it will not be possible, considering the impossibility of working with the variable income. We included a mention about SGA and LGA (ls 161-2). For the "SGA" and "LGA" outcomes, it is only possible to assess after 2011, as we do not have any information on gestational age before that.

10. Table 1:

• In the text you refer to the Bolsa Familia Program, but in the table, it is the Family Grant Program. I suggest you use Bolsa Familia here for consistency.

- We apologize for this mistake. We changed for Bolsa Familia Program.

• Under relevant variables from SINASC, it says “month that started prenatal after 2011”. Do you mean prenatal classes? Prenatal vitamins? Please check this.

- This refers to the month of pregnancy in which the woman started the prenatal consultations. We only have the variables number of consultations (2004 to 2015) and month in which prenatal care started (2011 to 2015). To make the information clearer, we have included the information in the table: “Some variables such as the month in which the woman started prenatal care and gestational age (as a continuous variable) are only available for the period from 2011 to 2015. ”

• Are all the main variables you intend to use in your analyses listed in this table? It would be very important to know other health-related variable of the mothers (i.e. pre-existing diabetes or gestational diabetes; smoking; etc.) as these will be relevant confounders for prematurity, LGA and SGA analyses.

- We have included all possible variables to be evaluated. Unfortunately, the database does not have important variables such as, smoking, physical activity or alcohol use.

11. Table 3: There is an extra ‘e’ in front of “extremely preterm” in the outcomes column.

- We apologize for this mistake. We changed this.

12. Figure 1: Resolution needs to be checked as it is barely readable at the moment.

- We have included a figure with better resolution.

For further consideration:

Your proposed rationale is reasonable, but have you considered how cash transfers that are conditional and preferentially paid to mothers may not increase purchasing power/ empowerment for all women. For example, the responsibility of getting children to school and to regular medical appointments for working mothers with partners may further entrench domestic/care work as women’s roles – potentially at the expense of paid employment, social networks, self-care, etc. Indeed, there appears to be a heterogeneous effect of CCT on women’s empowerment that may need further consideration in additional analyses:

De Brauw, A., Gilligan, D. O., Hoddinott, J., Roy, S. (2014). The impact of Bolsa Família on women’s decision-making power. World Development, 59, 487-504.

- We appreciate and agree with yourcomment. Indeed, the effect of Bolsa Família on female empowerment still needs further study due to the heterogeneous results, as mentioned.

Reviewer #2: It is not clear to me that what the manuscript describes warrants a study protocol. The construction of the database itself has been published elsewhere by the same group. All aspects described in the section “Secondary objectives, study population, definition of exposure, and outcomes” are minor and would be well suited to methodology sections of different papers. The statistical methods are solid for natural experiment studies in public health. I firmly believe that the authors should expand the details of their methodological decisions and processes in a future submission and publish the product of this development as a supplemental file to their methodology.

Nevertheless, if the authors decide to proceed with the submission, some key points should be addressed.

- We appreciate your suggestion. We strongly believe that designing a research protocol and submittingg it for critical peer review and publication before conducting studies, especially intervention studies, are important to describe the methods in suficiente detail and prevent undisclosed flexibility in the experimental procedures and analysis. We believe that it is a good practice for research, providing more quality and clarity to future findings. For these reasons we have decided to proceed with submission.

Keeping in mind that reproducibility is one of the main pillars of study protocols at PLOS ONE, the authors should provide a comprehensive background of how the databases that compose the 100 Million Brazilian Cohort can be accessed for research purposes. If only governmental officials can access the data, the authors should consider another type of publication for this manuscript.

Given the 100 Million Brazilian complexity, the author should also expand on how they plan to address bias in all three outcomes.

A more thorough explanation should be provided concerning data cleaning decisions and the linkage process.

- This study protocol will use linked data from the 100 Million Brazilian Cohort, coordinated and housed by the Center for Data and Knowledge Integration for Health (CIDACS). The cohort is composed of admistrative databases, which the owners, the Brazil’s Ministry of Health and Ministry of Citizenship, have given CIDACS the custody and authorization to conduct research, with the guarantee that all data processing takes place in a safe and private environment. Therefore, importante restrictions apply to the availability of this data.

Currently, only national and international researchers who collaborate with CIDACS and authorized staff from government agencies can access de-identified or anonymized linked data. Any person who wishes to receive authorization must: (i) be affiliated to CIDACS or be accepted as collaborators; (ii) present a detailed research project together with approval by an appropriate Brazilian institutional research ethical committee; (iii) provide a clear data plan restricted to the objectives of the proposed study and a summary of the analyses plan intended to guide the linkage and or data extraction of the relevant set of records and variables; (iv) sign terms of responsibility regarding the access and use of data; and (v) perform the analyses of datasets provided using the CIDACS data environment, a safe and secure infrastructure that provides remote access to de-identified or anonymized datasets and analysis tools. For more information, please visit the CIDACS website [https://cidacs.bahia.fiocruz.br/] or contact us via email [cidacs@fiocruz.br].

We have been working with different Linked databases. Recently, two articles were published that provide more details on the databases used to build the 100 Million Brazilian Cohort (doi: 10.1093/ije/dyab213) and CIDACS Birth Cohort (DOI: 10.1093/ije/dyaa255).

As mentioned earlier, the linkage procedures are common for 100 Million Cohort studies. We have included a recent publications and the mention: “We will use CIDACS Record Linkage (CIDACS-RL) to link the databases (https://doi.org/10.1186/s12911-020-01285-w). The linkage procedures consist of two stages. The first will be a deterministic linkage, and the second will be based on the similarity index. Others detailed information can be consulted in previous publications (https://doi.org/10.3389/fphar.2019.00984 and https://doi.org/10.1186/s12911-020-01192-0).” It is important to note that this protocol covers many studies and each specific objective involves a different database. Analyzes and cleaning procedures are specific to each database covered in this protocol. We believe that the mention of methodological studies, already published, can provide more detailed information about common procedures (linkage method) and general characteristics of the cohort.

Minor points to be addressed:

Some aspects of the study population, exposure to PBF, and outcome should be standardized between sections. For example, picking either the “2004-2015” or “2004 to 2015” to declare year ranges.

- We have standardized the ranges as suggested.

In the logic model, “Linkage to the place of birth” is not a product but a process and does not belong to this logic model.

- We agree with your comment, and we withdraw the term from the logic model.

Reviewer #3: “Evaluating the Impact of Bolsa Familia, Brazil’s conditional cash transfer programme on, on maternal and child health: a study protocol,” submitted to PLOS ONE (PONE-08640)

The protocol outlines a substantive analysis using various large administrative health and social program databases from Brazil to the so-called 100 million Brazilian cohort. It represents an ambitious research agenda (likely leading to multiple papers) that has potential to improve understanding of the influences of PBF. The authors make clear that despite its enormous size and importance, careful empirical assessment of the effect of PBF, particularly on child and maternal outcomes, is sparse. They identify an appropriate set of outcomes based on available data and the literature. There are likely to have sufficient power for even very small impacts and rare events such as maternal mortality (making it important to judge not only statistical impact but size of impact). The large sample sizes and observational nature of the data make the research design, i.e., arguments for assessing causal impact and not just associations, crucial.

Main Comments:

1. One important reason there is not more evidence on PBF is the lack of a strong research design for assessing program impacts, as was done for example via randomization of Progresa in Mexico. Another challenge when examining impact at a national level, I believe, are the differences in program administration across municipalities played. The team proposes resolving identification of the causal effects of the intervention via propensity score matching techniques. This approach is preferable to simpler comparisons but still relies on key assumptions of non-confoundedness across treatment and comparison groups after matching. After controlling for the observed factors available, the assumption is that there are no unobserved differences in those taking up the treatment and those not taking it up. Central problem is that those who enter, despite observed characteristics, might be different – ie more likely to benefit or have unobserved wealth or something we do not observe. So, sign of bias is difficult to ascertain. If untreated are better off on other characteristics, for example, might be able to argue results are conservative or underestimates of beneficial program benefits. Unconfoundedness cannot be proven but the matching literature provides various approaches for assessing it in the articles cited and my expectation is many of these will be done in your analyses.

- We appreciate your careful reading and we are very grateful for the pertinent criticism offered. Firstly we agree that results probably will be conservative or underestimate of beneficial program benefits. Second, we propose the effec evaluation via methods based on propensity score (weighting). Working with population databases gives us many possibilities to assess specific subpopulations and explore hypotheses, but it has some limitations. We will work with a database of people eligible for social programs, which can be up to 90% beneficiaries and 10% non-beneficiaries. Thus, we do not have a sample, we cannot previously define a number for the control group, and we need appropriate methods. In this sense, we believe that methods based on Propensity Score (weighting), could be applied. If it is known that these are procedures related to quasi-experimental methods, we are careful with the use of the term “causal effect”. We will use a large dataset with many confounding covariates, but unfortunately we cannot guarante unconfoundedness.

We will perform balancing after weighting to ensure that the procedure used controlled for available confounders. We have included the mention in the text: “Balancing will be performed before and after weighting to ensure that the procedure used controlled for the available confounders.” We will also do analyzes according to subpopulations (Robustness analysis for propensity score-based methods section).

2. In practice, carrying out PS or other matching techniques on these large data sets will involve dozens if not hundreds of decisions regarding the specifics (on which variables, functional form, common support etc.) and possibly variations on those decisions to assess sensitivity. It could be useful to say a bit more about how this will be approached, including the specifics of the data for readers less familiar with it – for example specific variables/measures that are included or links to those descriptions or an appendix.

- We appreciate your suggestion and have included a chart (chart 1) as supplementary material.

3. In their work (and related approaches), Imbens and coauthors develop other types of matching such as Nearest Neighbor (implemented in Stata using the command nnmatch). It may not be feasible with such large data sets to follow those approaches but a key aspect of them is allowing “exact” matching on certain types of characteristics. One that may be particularly important in this context is location – I noted mention of some subgroup analyses but think my suggestion here is a little different. Taking for example geographic location, to help ensure important elements like potential differences in municipality health systems are not leading to bias, a strategy of only matching treated cases with untreated in the same municipality can be used in the overall analysis. This permits an arguably better comparison than allowing geographic location, for example, to enter only via the combined propensity score.

- We appreciate your pertinent criticism. As mentioned earlier, we propose the effec evaluation via methods based on propensity score (weighting). We will work with a database of people eligible for social programs, which can be up to 90% beneficiaries and 10% non-beneficiaries. Thus, we do not have a sample, we cannot previously define a number for the control group, and we need appropriate methods. In this sense, we believe that methods based on Propensity Score (weighting), could be applied. We are still not sure if it would be possible to include the location in the propensity score calculation of with a level of disaggregation greater than the Region. This depends on the cases according to state or municipality. We intend to perform analysis by location subgroup (region, state...) and we can do it because we are using weighting (not matching), unless you have more than one region without cases.

4. Because PBF had an income cut-off, I did wonder whether there was any scope for an alternative approach to identification, related to regression discontinuity designs (RDD). I believe this could be done in conjunction with ps matching, but it would require availability of income measures (but these appear to be available). Or if not explicit, limiting comparison samples to those with incomes nearer to the cutoff, for example.

- We appreciate your comment. Due to changes to the income variable in the Cadúnico form, we cannot use this variable for the period from 2004 to 2015 (excessive missings and zeros). Thus, unfortunately, we will not be able to explore the RDD due to the impossibility of working with the income variable.

5. Administrative data match quality: I am unfamiliar with the various administrative data the study will use. It has been my experience in other settings, however, that combining administrative data across systems can be error ridden. To that end, greater support for the case that merging administrative records across the data sets is feasible and result in high quality (and high %) matches would strengthen confidence in the research design and the ultimate findings. Differences in quality of administrative match across the different data sets may influence findings in the three domains differently. It was unclear to me what the “similarity index” (page 3) approach was, but I presume on subsets of information (eg birth date, gender, location but not quite exact name spelling). A clear distinction in the final papers between the administrative matching across data sets and the ps matching procedures needs to be made. Characteristics of those matched and those not matched could shed light on potential biases.

- We have been working with different linked databases. Recently, two articles were published that provide more details on the databases used to build the 100 Million Brazilian Cohort (doi: 10.1093/ije/dyab213) and CIDACS Birth Cohort (DOI: 10.1093/ije/dyaa255). As mentioned earlier, the linkage procedures are common for 100 milion Cohort studies. We have included a recent publications and the mention: “We will use CIDACS Record Linkage (CIDACS-RL) to link the databases (https://doi.org/10.1186/s12911-020-01285-w). The linkage procedures are common for 100 milion Cohort studies and consist of two stages. The first will be a deterministic linkage, and the second will be based on the similarity index. Others detailed information can be consulted in previous publications (https://doi.org/10.3389/fphar.2019.00984 and https://doi.org/10.1186/s12911-020-01192-0). It is important to note that this protocol covers many studies and each specific objective involves a database. Analyzes are specific to each database covered in this protocol. We believe that the mention of methodological studies, already published, can provide more detailed information about common procedures (linkage method) and general characteristics of the cohort.

6. Are there any statistical considerations relevant to having particularly large sample sizes?

- Large sample sizes provide comprehensive data to conduct analyses on subgroups of interest while maintaining sufficient power to gain insights into the direction and size of the effects. It is important to highlight that larger samples provide great opportunities for empirical research, but also may lead to equivocal interpretations due to the detection of statistical significance. Below we indicate some publications.

Gelman A. P values and statistical practice. Epidemiology (Cambridge, Mass). 2013;24(1):69-72.

Siontis GCM, Ioannidis JPA. Risk factors and interventions with statistically significant tiny effects. International journal of epidemiology. 2011;40(5):1292-1307.

7. It was unclear to me whether length of exposure (beyond pregnancy periods) for outcomes would be considered, but I may have missed this.

- We apologize for the confusion of definitions. We created a chart to make these definitions clearer (Chart 1 in supplement).

________________________________________

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

Reviewer #3: No

Submitted filename: Rev_resp_190122.docx

Click here for additional data file.

14 Mar 2022

20 Apr 2022

To the Editor and reviewers,

First, we would like to express our appreciation for your consideration of our manuscript. We are very grateful for the pertinent criticism offered, and believe the incorporation of the reviewer’s suggestions will greatly contribute to the quality of our publication. Please find our point-by-point responses below to the criticism raised by each reviewer:

Journal Requirements:

Please review your reference list to ensure that it is complete and correct. If you have cited papers that have been retracted, please include the rationale for doing so in the manuscript text, or remove these references and replace them with relevant current references. Any changes to the reference list should be mentioned in the rebuttal letter that accompanies your revised manuscript. If you need to cite a retracted article, indicate the article’s retracted status in the References list and also include a citation and full reference for the retraction notice.

- We appreciate the suggestion. We have carefully reviewed our list of references. We did not have articles with retractions. We included newer government decrees, but kept those in force in the period defined by the study.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Does the manuscript provide a valid rationale for the proposed study, with clearly identified and justified research questions?

The research question outlined is expected to address a valid academic problem or topic and contribute to the base of knowledge in the field.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

________________________________________

2. Is the protocol technically sound and planned in a manner that will lead to a meaningful outcome and allow testing the stated hypotheses?

The manuscript should describe the methods in sufficient detail to prevent undisclosed flexibility in the experimental procedure or analysis pipeline, including sufficient outcome-neutral conditions (e.g. necessary controls, absence of floor or ceiling effects) to test the proposed hypotheses and a statistical power analysis where applicable. As there may be aspects of the methodology and analysis which can only be refined once the work is undertaken, authors should outline potential assumptions and explicitly describe what aspects of the proposed analyses, if any, are exploratory.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

________________________________________

3. Is the methodology feasible and described in sufficient detail to allow the work to be replicable?

Descriptions of methods and materials in the protocol should be reported in sufficient detail for another researcher to reproduce all experiments and analyses. The protocol should describe the appropriate controls, sample size calculations, and replication needed to ensure that the data are robust and reproducible.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

________________________________________

4. Have the authors described where all data underlying the findings will be made available when the study is complete?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception, at the time of publication. The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

________________________________________

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

________________________________________

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above and, if applicable, provide comments about issues authors must address before this protocol can be accepted for publication. You may also include additional comments for the author, including concerns about research or publication ethics.

You may also provide optional suggestions and comments to authors that they might find helpful in planning their study.

(Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: Overall, the manuscript is much more clear. The authors have made important improvements. There are, however, a few remaining issues to be addressed:

1- Please check your acronyms. Sometimes you use PBF, and other times BFP.

- We appreciate your careful review. We revised the text and changed the acronym that was present in reference 15, line 449.

2- Line 158-159: Why drop babies of less than 500g or born before 22 gestational weeks? I know they may be less likely to survive but could extreme prematurity and/or very low birthweight not be related to the socioeconomic standing of the mother? Is this not conditioning on the outcome? I understand the rationale for multiple births or congenital abnormalities as they are unlikely to be related to SES but related to birthweight and prematurity, but excluding the babies based on their likelihood of survival does not make sense to me.

- We appreciate your comment. We consider that survival may be related to SES and better access to a specialized delivery service, we used this feasibility criterion (500g and 22sg). As pointed out, women with better SES could be more likely to have babies with 300g or 20sg (criteria used in developed countries), but we must emphasize that we are working with populations of socioeconomically vulnerable women. In a way, we agree with your comment and removed the mention of the criterion that will be applied, keeping only the phrase “Fetal viability criteria can be applied [62-65]. "

3- On line 241, do you mean the parent's education or the child's?

- We appreciate the comment. We apologize, but we could not find this passage in the text. When we refer to education in the text, we are usually referring to maternal schooling. Except in some passages of the text (when citing the BFP conditionalities), we refer to school attendance by adolescents (17).

4- Line 270: You say "quasi-experimental approaches". You should explain how your approaches are quasi-experimental (i.e. exogenous changes with the BF program and subsequent outcomes, or regression discontinuity designs around the cut points for BF eligibility). I'm not sure I would use this characterization here based on how your have described your proposed analyses.

- We appreciate your comment and agree that the term used was not so appropriate. We rewrite the sentence to “This study will use propensity score-based methods to assess the BFP effect on maternal and child health outcomes in a large sample of poor and impoverished Brazilian households.”

Reviewer #2: The authors addressed all the points raised. I believe that this work have the potential to improve methodological rigor in future researches exploring maternal and child health outcomes

- We appreciate your careful review.

Reviewer #3: Thank you for the clarifications and revisions. In finalizing, I suggest you consider reconsidering words like "effect" and "impact" if you wish to be more careful around causal language (see you response to my first comment). In chart 1, column 1 it often says "access" but I believe you mean "assess"

- We appreciate your careful review. We consider maintaining the term “effect”. We consider this to be justified by the limitations of each study generated from this protocol. We know that we cannot sustain the “causal effect” and that the effect of the Bolsa Família found will probably be associative, due to the characteristics of the program and the proposed analysis (adequate to the particularities of the program and the database).

- We have modified the terms in Chart 1 to “To evaluate”.

________________________________________

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

Reviewer #3: Yes: John A. Maluccio

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

- We appreciate the suggestion. The figure was generated from the PACE digital diagnostic tool and attached to this submission.

Submitted filename: Asw_2rev_19042022.docx

Click here for additional data file.

4 May 2022

Evaluating the effect of Bolsa Familia, Brazil’s conditional cash transfer programme, on maternal and child health: a study protocol

PONE-D-21-08640R2

Dear Dr. Falcão,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Bárbara Hatzlhoffer Lourenço, Ph.D.

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

12 May 2022

PONE-D-21-08640R2

Evaluating the effect of Bolsa Familia, Brazil’s conditional cash transfer programme, on maternal and child health: a study protocol

Dear Dr. Falcão:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Bárbara Hatzlhoffer Lourenço

Academic Editor

PLOS ONE