Sage Journals: Discover world-class research

Abstract

Objective:

To test the hypothesis that a continuing educational strategy (ie, “the manual”) in primary health-care improves infant feeding practices among infants under 1 year of age.

Methods:

A before and after study was conducted at primary health-care units in Embu das Artes, Brazil. The intervention was the use of a manual created to support continuing educational activities on breastfeeding and complementary feeding to be performed by tutors of Estratégia Amamenta e Alimenta Brasil with health-care teams, in a period of 8 months. Five hundred sixty-one mothers before and 598 mothers after intervention were interviewed about breastfeeding and complementary feeding practices. Multivariate analysis was performed using Poisson multilevel regression to test the hypothesis.

Results:

Lack of minimum food diversity (before 62.9%; after 50.3%) and lack of food adequacy (before 77.5%; after 63.3%) decreased significantly. Regression analysis confirmed that infants after the intervention had lower prevalence of inadequacy of complementary feeding. While the intervention did not show significant association with exclusive breastfeeding, it showed association with the improvement of complementary feeding practices.

Conclusions:

The manual is a continuing educational strategy that improved complementary feeding practices in primary health care.

Keywords

breastfeeding infant nutrition infant feeding practices intervention studies

Introduction

Breastfeeding and adequate complementary feeding (CF) in the first years of life provide short- and long-term benefits for health and child development, in addition to economic and environmental advantages for children, women, and society.^1,2 According to the World Health Organization (WHO) recommendations, infants should be exclusively breastfed for the first 6 months and after that infants should receive nutritionally adequate and safe complementary foods while breastfeeding continues for up to 2 years of age or beyond.³ Such healthy feeding habits in early childhood can provide a lifetime of protection against chronic diseases such as overweight and obesity.^4

-7

Despite numerous global initiatives, breastfeeding rates, especially exclusive breastfeeding (EBF), are below recommended levels and CF is often offered at inadequate time (ie, early or late) and with low-quality food.^8

-11 Historically, in Brazil, breastfeeding promotion was focused on changing hospital practices by implementing The Baby Friendly Hospital Initiative, but more recently, this focus has expanded to promote EBF and adequate infant feeding in primary health-care settings.

In 2013, the Brazilian Ministry of Health developed an innovative program, the Estratégia Amamenta e Alimenta Brasil (EAAB), freely translated as the “Brazilian Breastfeeding and Complementary Feeding Strategy.” The EAAB’s goal is to improve the quality of the assistance provided by primary health-care teams to ultimately improving infant feeding practices in the Brazilian population. The EAAB is grounded on a problem-based learning methodology and proposes a continuing education process to both change working process in primary health-care settings and facilitate the transformation of knowledge in evidence-based practices. The continuing education in breastfeeding and CF is facilitated by a health professional who works as a coach (ie, EAAB tutor) supporting health-care teams within each primary health-care unit (PHU).¹²

Estratégia Amamenta e Alimenta Brasil implementation in the PHU starts with a 6-hour workshop conducted by the tutor with the entire PHU staff. The PHU can be certified in the EAAB program if they meet 6 certification criteria determined by the Ministry of Health: (1) developing systematic actions to promote breastfeeding and healthy CF; (2) monitoring breastfeeding and CF indicators; (3) maintaining a flowchart to organize the child care related to breastfeeding and CF; (4) complying with the Brazilian Code of Marketing of breast milk substitutes, foods for infants and toddlers, pacifiers, and bottles; (5) involving at least 85% of current staff in the workshops developed by tutors; (6) implementing at least one breastfeeding and one CF promotion action as agreed in the action plan developed by the health-care team during the workshops.¹² The tutor should support primary health-care teams to meet these 6 certification criteria to improve the quality of assistance to promote breastfeeding and CF.¹² However, data reported by the Brazilian Ministry of Health show that the number of tutors trained is 40 times higher than the number of PHU certified in the EAAB program. Thus, there is a clear challenge to advance toward EAAB certification in the Brazilian Unified Health System, where there are more than 40 000 PHUs.

In order to strengthen the implementation of EAAB, a manual with supporting material was developed. In partnership with the Ministry of Health the Manual to Support EAAB Tutors (the manual) was outlined to clarify the EAAB tutor’s role and to suggest some comprehensive training activities to be led by him/her with the health-care team. The themes and contents of the comprehensive training activities were designed to guide PHU to EAAB certification. The manual follows a problem-based learning methodology to propose an educational process of teaching (“coaching”) and encouraging changes in health workers’ practice and improvements in the work process within the PHU to provide a high-quality standard care to promote breastfeeding and CF. The use of the manual is the intervention evaluated in this study.

Our hypothesis is that the use of the manual as a tool of continuing education in primary health care can strengthen the EAAB implementation and thereafter improve infant feeding practices. Thus, this article aims to evaluate the effectiveness of using the manual by analyzing infant feeding practices.

Methods

Design and Study Setting

In this before and after study were conducted 2 cross-sectional surveys on infant feeding practices at the 13 PHU located in the urban area of the city of Embu das Artes. Embu das Artes is a low-income city in Brazil, part of the metropolitan region of Sao Paulo, with its total population of 240 230 inhabitants.

A formal permission from the authority of the Health Department of Embu das Artes was obtained as well as written informed consent from mothers of infants who participated in the study. The research protocol was approved by the Research Ethics Committee of the Public Health School of the University of Sao Paulo, Brazil (protocol 43317315.0.0000.5421, approved May 07, 2015).

Intervention

At the beginning of the study, the municipality of Embu das Artes had already 13 EAAB tutors trained but they had not started the EAAB implementation. The intervention consisted in the use of the manual by EAAB tutors who performed some comprehensive training activities with health-care teams, in a period of 8 months. The manual was presented to the tutors by the research coordination and they were asked to follow the manual as written, without any interference of the researchers in the intervention. The manual proposes 5 comprehensive training activities (CAs) with step-by-step suggestions for the tutor to carry out such activities with health-care teams. Each CA was estimated to last no longer than 2 and a half hours. The CAs include clinical case-studies (a narrative situation of clinical difficulties that a mother/family may face during breastfeeding or CF) and discussion guides. The 5 themes (outlined to support the PHU in meeting the EAAB certification criteria) cover: (1) action plan to promote breastfeeding and CF; (2) how to integrate the promotion of breastfeeding and CF into existents primary health-care activities; (3) clinical management to encouraging breastfeeding in primary health-care units; (4) how to promote healthy CF in primary health-care units; and (5) feeding and nutrition surveillance.

Figure 1 presents the Program Impact Pathways of the intervention (the use of the manual) clarifying the mechanisms that lead to the expected outcomes,¹³ that is, changes in the infant feeding practices considering the context of EAAB implementation. Thus, in this study the focus of the analysis will be on the intervention impact, that is on the EBF and CF indicators.

Figure 1.

Program impact pathways (PIP) diagram of the implementation of the use of the manual to support Estratégia Amamenta e Alimenta Brasil tutors.

Participants

Participants consisted of mothers of infants under 1 year of age attended at 1 of the 13 PHUs. A required sample size of 460 infants under 6 months of age was estimated to be able to detect an increase of 6% in the prevalence of EBF after intervention. For children aged 6 to 12 months, the sample size was based on the prevalence of consumption of stuffed cookies, chips, or salty or otherwise seasoned crackers. A sample size of 140 children was estimated to be required in order to detect a reduction of 10% in the prevalence of consumption of unhealthy foods.¹⁴ In total, the estimated sample size of 600 children under 1 year of age was needed for each cross-sectional survey, considering the sample size required to compare independent samples.¹⁵ Considering the number of children attending each PHU, the total estimated sample was proportionally distributed among the PHU, totaling a sample size from 35 to 60 infants per PHU. For the obtaining sample, mothers were interviewed if their child had an appointment in the PHU in the period of data collection. The sample size was calculated assuming a margin of error of 5% and a statistical power of 90%.

Data Collection

Data collection was carried out in the before and after intervention phases, from July to September 2015 and from August to September 2016, respectively, at the 13 PHU. Mothers of children under 1 year of age were invited to participate in the study while awaiting appointment at the PHU. In both phases, the interviews were conducted by trained staff and supervised by the leading researcher. The questionnaire was composed of 2 parts. The first had closed-ended questions related to socioeconomic, demographic, and biomedical conditions of the mother, child, and other family members and some questions about the care received in the PHU. In the second part, data were collected on food consumption adopting the Food Intake Markers Questionnaire for infants between 6 and 24 months, as proposed by the Brazilian Food and Nutrition Surveillance System. These questionnaires consist of yes or no answers about food consumption, such as breast milk, water, fluids, other types of milk, food groups, food frequency, food consistency, and unhealthy foods. All questions were applied to child consumption in the previous 24 hours of the interview. This questionnaire enables assessment of food consumption markers once it identifies healthy and unhealthy eating practices.¹⁶

In the before and after intervention phases, the manager from each PHU was asked about the presence of some criteria required by the Ministry of Health to certify a PHU for EAAB.

Study Variables

Outcomes

The outcomes were defined as follows:

Exclusive breastfeeding when a child from 0 to 5 months of age is fed with breast milk exclusively. The specific questions (yes/no) that mothers were asked to determine exclusivity were: (a) breast milk consumption, (b) porridge, (c) water/tea, (d) cow milk, (e) infant formula, (f) fruit juice, (g) fruit, (h) salt meals (solid, semi-solid, or soft foods), and (i) any other foods or beverages. A child was considered in EBF if the mother answer “yes” for breast milk consumption and “no” for all other questions.

Minimum food diversity (MFD): when a child from 6 to 11 months of age received at least one serving from each of these 6 food groups (yes/no): (1) complex carbohydrates (like rice, roots and tubers); (2) breast milk or dairy products; (3) meats or eggs; (4) beans; (5) vitamin-A-rich fruits and vegetables; and (6) other fruits or vegetables.

Food adequacy (FA): yes/no variable calculated based on 3 information, MFD, minimum frequency, and proper consistency. Minimum food diversity was estimated as describe above. Minimum frequency and proper consistency were estimated at the same time and was considered when a 6-month-old child received solid, semi-solid, or soft foods once a day whether in a mashed consistency or in chunks; and when a 7 to 11-month-old child received solid, semi-solid, or soft foods twice a day whether in a mashed consistency or in chunks.

Independent variable

The main independent variable was the intervention, that is, before intervention or after intervention.

Covariables

To estimate the individual influence of the outcome-associated variables, we used a multiple hierarchical model. The covariables were grouped into blocks and ordered according to the influence that they would have on each outcome (ie, EBF and CF indicators). Figure 2 presents the theoretical models that guided the 2 levels of the analysis, contextual, and individual variables. The maternal education level was used as a proxy for socioeconomic status. The contextual variable number of certification criteria fulfilled was created based on manager’s responses before and after the intervention. Each PHU was categorized in accordance with the number of certification criteria fulfilled: 6 criteria; 5 to 4 criteria; 3 to 2 criteria; 0 to 1 criterion.

Figure 2.

Hierarchical theoretical model of outcomes: Exclusive breastfeeding (A) and minimum food diversity and food adequacy (B).

Data Analysis

All data were double entered for subsequent validation in the Epi-Info, version 3.5 software (Center for Disease Control and Prevention, Atlanta, Georgia). Data management and statistical analysis were performed using the Stata statistical software package version 14.1 (Stata Corp, College Station, Texas).

In the analyzes it was not excluded PHU where no CAs were carried out as we considered that the availability of the manual could influence in some way, even if the tutor did not support the health team as proposed in the manual.

A comparison between the before- and after-intervention groups for the characteristics of child, family/maternal and care received in the PHU was conducted using χ² test. Differences in the prevalence of the outcomes before and after intervention were tested by the application of χ² and were explored among the specific age ranges.

To analyze the associations between the intervention and the 3 outcomes, a Poisson multilevel regression was conducted with robust variance following the hierarchical model in Figure 2. Poisson regression was adopted because of the cross-sectional data with nonrare outcomes.¹⁷ In the multilevel analysis, the contextual level (level 2) included the PHU variable to organize individual data according to clusters; the individual level (level 1) included family/maternal and child data and characteristics of care received in the PHU. Considering the multilevel structure, initially, the prevalence ratio (PR) and confidence intervals (95% CI) were estimated for each variable and outcome (step 1). Variables with P < .50 were entered into multiple internal analyses of each block (step 2). Then, variables with P < .20 in the internal analysis of each block were included in the multilevel modeling analysis (step 3).

Step 3 consists of the stepwise hierarchical multilevel modeling following Figure 2. Both variables PHU and infant age were added to the empty model (data not shown). Model 1 included the number of certification criteria fulfilled (contextual level); this variable remained in the model as an adjustment for subsequent models. Subsequently, maternal education (proxy for socioeconomic status) was added to model 2 after adjustment for the model 1 variable, and then remained as another control variable for the subsequent model. Likewise, variables related to characteristics of care received in the PHU and the variable intervention were included in model 3 after adjustments for the model 1 and 2 variables and then remained as control variables for the subsequent model. A similar procedure was adopted for variables related to maternal and family characteristics (model 4) and to child characteristics (model 5). Significant variables were maintained in the model even if they became insignificant after the inclusion of lower order variables. A significance level of 5% was adopted after adjustment for variables within the same model and for upper order variables. The quality of model adjustment was assessed using the 2-log likelihood test. In the multilevel analysis, the fixed effects/random intercept model described by Snijders and Bosker¹⁸ was used.

Considering that this study has 3 different outcomes, the steps for data analysis were separately reproduced for each outcome.

Results

The results of the implementation process (ie, use of the manual) were as follows: 9 of 13 tutors conducted Cas with the health-care teams of 11 PHU. The reasons why 4 tutors did not remain in the study were retirement, dismissal, and maternity leave. Tutors from 7 PHU conducted 5 CAs; 1 tutor conducted 4 CAs; 3 tutors conducted 2 CAs; and in 2 PHU none activity was developed. A total of 153 health workers were exposure to the CAs. The number of attendees in each CAs varied widely among PHU. Only 21 health-care workers participated in 5 CAs.

Before the intervention, of the 571 mothers interviewed initially, 561 children were enrolled into the study (an 1.7% loss); after the intervention, 609 mothers were interviewed initially, and 598 children were enrolled into the study (an 1.8% loss), totalizing 1159 children under 1 year of age analyzed in this study. Figure 3 outlines the stages of the study and shows the reasons for losses. Refusals were negligible.

Figure 3.

Flowchart of the intervention study in primary health-care units. Embu das Artes, Brazil, 2015 to 2016.

Table 1 shows the comparison of mother, child, family, PHU, and care characteristics between the before- and after-intervention groups. There was no difference between the 2 groups in terms of gender, birth weight, or use of pacifier or bottle. Maternal age and education were similar for all participants, and a significant difference was observed in the maternal work. Regarding the care received, there were some significant differences between the groups. The number of certification criteria fulfilled by PHU was significantly higher in the after-intervention group (P = .000).

Table 1.

Number and Proportion of Infants Under 1 Year of Age According to Child, Maternal, Family, Care Received and Primary Health-Care Unit Characteristics, by Study Group. Embu das Artes, Sao Paulo, Brazil, 2015 to 2016.

Variable	Before Intervention (n = 561)		After Intervention (n = 598)		P
Variable	n ^a	%	n ^a	%	P
Child characteristics
Age (months)
0-3	168	29.95	235	39.30	002
3-6	195	34.76	204	34.11
6-9	120	21.39	96	16.05
9-12	78	13.90	63	10.54
Gender
Male	290	51.69	307	51.34	.904
Female	271	48.31	291	48.66
Low birth weight
No	495	90.99	540	90.45	.753
Yes	49	9.01	57	9.55
Breastfeeding within first hour of life
Yes	350	64.81	257	43.19	.000
No	190	35.19	338	56.81
Bottle feeding
No	218	41.52	255	43.0057.00	.618
Yes	307	58.48	338	43.0057.00	.618
Pacifier use
No	265	50.48	308	51.85	.646
Yes	260	49.52	286	48.15
Maternal and family characteristics
Maternal age (years)
<20	107	19.14	98	16.44	.479
20-34	376	67.26	412	69.13
≥35	76	13.60	86	14.43
Primiparous
Yes	274	49.10	243	40.77	.004
No	284	50.90	353	59.23
Maternal education (years)
0-7	77	14.86	66	11.06	.162
8-11	392	75.68	470	78.73
≥12	49	9.46	61	10.22
Maternal work
Not working outside/maternity leave	417	75.27	505	84.59	.000
Working outside	137	24.73	92	15.41
Lives with the child’s father
Yes	444	80.00	458	76.72	.177
No	111	20.00	139	23.28
Lives with child’s grandmother
Yes	94	16.94	100	16.75	.933
No	461	83.06	497	83.25
Maternal daily dedication to child care
Full time	455	81.98	501	84.92	.181
Part time	100	18.02	89	15.08
Difficulties to breastfeed
Yes	140	26.77	118	19.97	.007
No	383	73.23	473	80.03
Care received and PHU characteristics
Primary health-care model
PHU with ESF	149	26.56	162	27.09	.931
Traditional PHU	127	22.64	130	21.74
Combination PHU	285	50.80	306	51.17
Delivery
Vaginal	321	58.68	379	63.80	.076
Cesarean	226	41.32	215	36.20
Time of first appointment after child’s birth
First week of child’s life	220	40.00	134	22.71	.000
After first week of child’s life	330	60.00	456	77.29
Home visit after child’s birth
Yes	114	20.61	93	15.55	.025
No	439	79.39	505	84.45
PHU schedules regular appointments
Yes	486	92.75	524	89.57	.064
No	38	7.25	61	10.43
PHU attends without appointment
Yes	334	62.31	287	48.89	.000
No	202	37.69	300	51.11
Advice or support provided on BF
Yes	450	81.23	470	78.73	.290
No	104	18.77	127	21.27
Advice provided on CF ^b
Yes	137	73.26	108	69.68	.464
No	50	26.74	47	30.32
Number of EAAB certification criteria fulfilled
6	30	5.35	206	34.45	.000
4-5	245	43.67	281	46.99
2-3	116	20.68	111	18.56
0-1	170	30.30	0	0

Abbreviations: BF, breastfeeding; CF, complementary feeding; EAAB, Estratégia Amamenta e Alimenta Brasil; PHU, primary health-care unit.

^aThe total was smaller for some variables due to missing information.

^bOnly infants from 6 to 11 months and 29 days, n = 357.

Regarding the outcomes, in descriptive analyses, the prevalence of EBF cessation presented nonsignificant difference between the groups (before 58.0%; after 53.7%; P = .229); while the lack of MFD (before 62.9%; after 50.3%; P = .023) and the lack of FA (before 77.5%; after 63.3%; P = .005) decreased significantly.

Table 2 presents the bivariate analysis and the covariates included in the multivariate model (P < .50) for each outcome. The multilevel hierarchical analysis indicates that mothers having difficulties in breastfeeding, maternal daily dedication to child care, and pacifier use were independent factors associated with EBF cessation, but the intervention did not show association with this outcome (Table 3). For both CF indicators, the multilevel hierarchical analysis shows that infants in the after-intervention group had a lower chance of nutritional inadequacy: lack of MFD PR = 0.70 (0.50-0.97; P < .05); and lack of FA PR = 0.74 (0.55-0.99; P < .05; Table 4).

Table 2.

Poisson Bivariate Analysis of Exclusive Breastfeeding (EBF) Cessation, Lack of Minimum Food Diversity (MFD) and Lack of Food Adequacy (FA) in Infants Under 1 Year of Age. Embu das Artes, Sao Paulo, Brazil, 2015 to 2016.

Variables	EBF interruption^a			Lack of MFD^b			Lack of FA^b
	(n = 794)			(n = 322)			(n = 315)
	% (n)	PR_crude	P	% (n)	PR_crude	P	% (n)	PR_crude	P
Main independent variable
Intervention
Before-intervention	58.03 (206)	1		62.87 (105)	1	.14	77.50 (124)	1
After-intervention	53.76 (236)	0.93 [0.77 to 1.12]	.42	50.32 (78)	0.80 [0.60 to 1.08]		63.23 (98)	0.81 [0.62 to 1.06]	.13
Proxy for socioeconomic status
Maternal education (years)
≥12	51.16 (44)	1		50.00 (12)	1		66.67 (16)	1
8-11	54.41 (321)	1.06 [0.78 to 1.46]	.7	57.83 (144)	1.16 [0.64 to 2.08]	.62	70.66 (171)	1.06 [0.63 to 1.77]	.82
0-7	62.11 (59)	1.21 [0.82 to 1.79]	.33	60.47 (26)	1.21 [0.61 to 2.39]	.59	74.42 (32)	1.12 [0.61 to 2.03]	.72
Care received and PHU characteristics
Delivery
Vaginal	52.17 (252)	1		–	–	–	–	–	–
Cesarean	61.18 (186)	1.17 [0.97 to 1.42]	.09
Breastfeeding within first hour of life
Yes	53.83 (225)	1		–	–	–	–	–	–
No	57.49 (211)	1.07 [0.88 to 1.29]	.49
Time of first appointment after child’s birth
First week of child’s life	56.56 (125)	1		–	–	–	–	–	–
After first week of child’s life	55.28 (309)	0.98 [0.79 to 1.20]	.83
Home visit after child’s birth
Yes	55.88 (76)	1		–	–	–	–	–	–
No	55.42 (363)	0.99 [0.77 to 1.27]	.95
PHU schedules regular appointments
Yes	56.20 (385)	1		55.71 (161)	1		70.67 (200)	1
No	51.81 (43)	0.92 [0.67 to 1.26]	.61	86.67 13)	1.55 [0.88 to 2.74]	.12	93.33 14)	1.32 [0.77 to 2.27]	.31
PHU attends without appointment
Yes	51.96 (225)	1		55.90 (90)	1		69.87 (109)	1
No	58.46 (197)	1.12 [0.93 to 1.36]	.22	57.79 (89)	1.03 [0.77 to 1.38]	.82	71.05 (108)	1.01 [0.78 to 1.33]	.9
Advice or support provided on BF
Yes	55.00 (352)	1		–	–	–	–	–	–
No	57.43 (85)	1.04 [0.82 to 1.32]	.72
Advice provided on CF
Yes	–	–	–	54.50 (121)	1		67.44 (145)	1
No				61.29 (57)	1.12 [0.82 to 1.54]	.46	75.27 (70)	1.11 [0.84 to 1.48]	.45
Number of EAAB certification criteria fulfilled by the PHU
6
4-5	48.85 (85)	1		63.16 (36)	1	.43	73.68 (42)	1
2-3	57.93 (201)	1.18 [0.92 to 1.53]	.19	53.90 (83)	0.85 [0.58 to 1.26]	.85	68.21 (103)	0.92 [0.65 to 1.32]	.67
0-1	60.24 (100)	1.23 [0.92 to 1.65]	.15	60.38 (32)	0.95 [0.59 to 1.54]	.58	76.92 (40)	1.04 [0.68 to 1.61]	.85
	52.43 (56)	1.07 [0.76 to 1.50]	.69	55.17 (32)	0.87 [0.54 to 1.40]		67.27 (37)	0.91 [0.59 to 1.42]	.69
Maternal and family characteristics
Maternal age (years)
<20	53.71 (282)	1		56.60 (133)	1	.32	70.56 (163)	1
20-34	59.59 (87)	1.10 [0.87 to 1.41]	.39	68.75 (33)	1.21 [0.83 to 1.78]	.31	76.09 (35)	1.08 [0.75 to 1.55]	.69
≥35	60.33 (73)	1.12 [0.86 to 1.45]	.38	43.24 (16)	0.76 [0.45 to 1.28]		63.89 (23)	0.90 [0.58 to 1.40]	.66
Primiparous
Yes	51.22 (231)	1		58.68 (98)	1		69.33 (113)	1
No	61.40 (210)	1.20 [0.99 to 1.44]	.06	54.97 (83)	0.94 [0.70 to 1.25]	.66	72.30 (107)	1.04 [0.80 to 1.36]	.76
Maternal work
Not working outside/maternity leave	51.89 (344)	1		57.71 (131)	1		70.59 (156)	1
Working outside	75.59 (96)	1.46 [1.16 to 1.83]	0	54.35 (50)	0.94 [0.68 to 1.30]	.72	70.33 (64)	0.99 [0.74 to 1.33]	.98
Lives with the child’s father
Yes	53.19 (333)	1		58.47 (145)	1		72.84 (177)	1
No	64.63 (106)	1.21 [0.98 to 1.51]	.08	51.39 (37)	0.88 [0.61 to 1.26]	.48	62.86 (44)	0.86 [0.62 to 1.20]	.38
Lives with child’s grandmother
Yes	62.02 (80)	1		52.73 (29)	1		66.04 (35)	1
No	54.31 (359)	0.87 [0.69 to 1.11]	.28	57.74 (153)	1.09 [0.73 to 1.63]	.65	71.54 (186)	1.08 [0.75 to 1.55]	.66
Maternal daily dedication to child care
Full time	51.15 (355)	1		59.56 (134)	1		71.95 (159)	1
Part time	87.10 (81)	1.70 [1.34 to 2.17]	0	52.75 (48)	0.88 [0.64 to 1.23]	.47	69.32 (61)	0.96 [0.72 to 1.29]	.8
Difficulties to breastfeed
No	49.48 (286)	1		–	–	–	–	–	–
Yes	75.92 (145)	1.53 [1.26 to 1.87]	0
Child characteristics
Gender
Male	58.29 (239)	1		60.84 (101)	1		73.62 (120)	1
Female	52.86 (203)	0.91 [0.75 to 1.09]	.3	52.56 (82)	0.86 [0.64 to 1.16]	.32	67.11 (102.	0.91 [0.70 to 1.19]	.49
Low weight
No	54.29 (392)	1		–	–	–	–	–	–
Yes	68.66 (46)	1.26 [0.93 to 1.72]	.13
Bottle feeding
No	–	–	–	55.95 (47)	1		69.05 (58)	1
Yes				56.58 (129)	1.01 [0.72 to 1.41]	.95	71.04 (157)	1.03 [0.76 to 1.39]	.85
Pacifier use
No	41.92 (166)	1		53.33 (88)	1		69.14 (112)	1
Yes	69.55 (265)	1.66 [1.37 to 2.01]	0	59.86 (88)	1.12 [0.83 to 1.50]	.44	72.03 (103)	1.04 [0.80 to 1.36]	.76

Abbreviations: BF, breastfeeding; CF, complementary feeding; EAAb, Estratégia Amamenta e Alimenta Brasil; EBF, exclusive breastfeeding; PHU, primary health-care unit; PR, prevalence ratio.

^aInfants under 6 months.

^bInfants from 6 to 11 months and 29 days.

Table 3.

Multilevel Hierarchical Regression Models to Estimate the Prevalence Ratio of Exclusive Breastfeeding (EBF) Cessation in Infants Under 6 Months According to Child, Maternal, Family, Care Received and Primary Healthcare Unit (PHU) Characteristics, Adjusted by Infant Age. Embu das Artes, SP, Brazil, 2015 to 2016.^a,b

EBF Interruption	Model 1^c	Model 2^d	Model 3^e	Model 4^f	Model 5^g
Fixed effects—constant	0.27 (0.21-0.36)^f	0.25 (0.17-0.37)^f	0.46 (0.27-0.65)^f	0.32 (0.21-0.51)^f	0.26 (0.16-0.41)^f
Maternal education (years)
≥12	–	1	–	–	–
8-11	–	1.06 (0.78-1.46)	–	–	–
0-7	–	1.26 (0.85-1.87)	–	–	–
Intervention
Before-intervention	–	–	1	–	–
After-intervention	–	–	0.97 (0.77-1.22)	–	–
PHU attends without appointment
Yes	–	–	1	–	–
No	–	–	1.15 (0.95-1.40)	–	–
Difficulties to breastfeed
No	–	–	–	1	–
Yes	–	–	–	1.58 (1.28-1.95) ^f	–
Lives with the child’s father
Yes	–	–	–	1	–
No	–	–	–	1.20 (0.95-1.51)	–
Maternal daily dedication to child care
Full time	–	–	–	1	–
Part time	–	–	–	1.71 (1.33-2.22) ^f	–
Pacifier use
No	–	–	–	–	1
Yes	–	–	–	–	1.58 (1.28-1.96) ^f
Contextual level—primary health-care unit
EAAB certification criteria fulfilled
6 criteria	1	–	–	–	–
4-5 criteria	1.12 (0.87-1.44)	–	–	–	–
3- 2 criteria	1.14 (0.85-1.52)	–	–	–	–
0-1 criteria	0.94 (0,67-1.33)	–	–	–	–
Random effect EAAB certification criteria fulfilled—constant	2.42 E-37	1.94 e-36	4.61 E-34	5.17 E-37	1.06 e-34
−2loglikelihood	1348.0268	1299.39514	1313.33646	1225.77142	1180.98198

Abbreviation: EAAb, Estratégia Amamenta e Alimenta Brasil.

^aInfants under 6 months, n = 794.

^bIndividual data from the 2 phases of the study.

^cModel 1: number of certification criteria and infant age.

^dModel 2: Model 1 + maternal education.

^eModel 3: Model 2 + intervention and PHU attends without appointment.

^fModel 4: Model 3 + difficulties to breastfeed, lives with the child’s father, maternal daily dedication to child care.

^gModel 5: Model 4 + pacifier use.

^fP < .001.

Table 4.

Multilevel Hierarchical Regression Models to Estimate the Prevalence Ratio of Lack of Minimum Food Diversity (MFD) and Lack of Food Adequacy (FA) in Infants from 6 to 11 Months According to Child, Maternal, Family, Care Received and Primary Healthcare Unit (PHU) Characteristics, Adjusted by Infant Age. Embu das Artes, SP, Brazil—2015 to 2016.^a,b

Lack of Minimum Food Diversity	Model 1^c	Model 2^d	Model 3^e	Model 4^f	Model 5^g
Fixed effects—constant	3.51 (1.56-7.91)^h	3.10 (1.15-8.38)^h	6.17 (1.90-19.96)^h	6.09 (1.86-19.93)^h	6.36 (1.83-22.02)^h
Maternal education (years)
≥12	–	1	–	–	–
8-11	–	1.13 (0.62-2.04)	–	–	–
0-7	–	1.09 (0.54-2.17)	–	–	–
Intervention
Before-intervention	–	–	1	–	–
After-intervention	–	–	0.70 (0.50-0,97)^h	–	–
Maternal age (years)
20-34	–	–	–	1	–
<20	–	–	–	1.19 (0.81-1.75)	–
≥35	–	–	–	0.73 (0.43-1.23)	–
Infant gender
Male	–	–	–	–	1
Female	–	–	–	–	0.96 (0.72-1.29)
Contextual level—Primary health-care unit
EAAB certification criteria fulfilled
6 criteria	1	–	–	–	–
4-5 criteria	0.84 (0.57-1.24)	–	–	–	–
3-2 criteria	0.91 (0.57-1.47)	–	–	–	–
0-1 criterion	0.85 (0.53-1.37)	–	–	–	–
Random effect EAAB certification criteria fulfilled - constant	3.47 E-35	2.87 E-35	1.41 e-35	7.23 e-36	2.34 E-35
−2loglikelihood	551.42068	543.94012	539.3975	534.68724	534.63052
Lack of food adequacy	Model 1^c	Model 2^d	Model 3^e	Model 4ⁱ	Model 5^g
Fixed effects—constant	3.03 (1.45-6.31)^h	2.84 (1.17-6.92)^h	5.04 (1.76-14.40)^h	4.89 (1.69-14.13)^h	5.15 (1.69-15.65)^h
Maternal education (years)
≥12	–	1	–	–	–
8-11	–	1.02 (0.61-1.71)	–	–	–
0-7	–	1.00 (0.54-1.84)	–	–	–
Intervention
Before-intervention	–	–	1	–	–
After-intervention	–	–	0.74 (0.55-0.99)^h	–	–
Lives with the child’s father
Yes	–	–	–	1	–
No	–	–	–	0.89 (0.64-1.25)	–
Infant gender
Male	–	–	–	–	1
Female	–	–	–	–	0.96 (0.73-1.26)
Contextual level—Primary health-care unit
EAAB certification criteria fulfilled
6 criteria	1	–	–	–	–
4-5 criteria	0.91 (0.63-1.30)	–	–	–	–
3-2 criteria	1.00 (0.65-1.55)	–	–	–	–
0-1 criterion	0.89 (0.57-1.38)	–	–	–	–
Random effect EAAB certification criteria fulfilled—constant	1.61 E-39	3.84 E-37	2.79 E-36	1.56 E-35	1.45 e-36
−2loglikelihood	581.49458	572.30522	568.30406	565.72786	565.64044

^aInfants from 6 to 11 months and 29 days, n = 322.

^bIndividual data from the 2 phases of the study.

^cModel 1: number of certification criteria and infant age.

^dModel 2: model 1 + maternal education.

^eModel 3: model 2 + intervention.

^fModel 4: model 3 + maternal age.

^gModel 5: model 4 + infant gender.

^hP < .05.

ⁱModel 4: model 3 + lives with the child’s father.

Discussion

Our findings show that the educational intervention delivered by EAAB tutors using the manual to guide their work with a health-care team can effectively improve the quality of CF in infants aged between 6 and 12 months. The Manual to Support EAAB Tutors proved to be useful for continuing education and capable of improving infant feeding practices.

In this study, we observed a prevalence of 42% EBF before the intervention and 46% after the intervention. While our results indicated a lower prevalence of EBF cessation after intervention, it did not reach statistical significance. Indeed, the intervention was not associated with changes in EBF prevalence. In Embu das Artes, EBF rates increased gradually from 16% in 2001 to 42% in 2010 and decreased to 34% in 2012.¹⁴ Perhaps no additional benefit due to our intervention was observed because historically the municipality has been developing regular actions to promote breastfeeding and has reached higher prevalence in comparison to national data which currently is 37% for EBF.¹⁹ Maternal daily dedication, breastfeeding problems, and pacifier use were independently associated with the early cessation of EBF. These results are in accordance with recently systematic reviews addressing determinants and barriers to EBF^2,10,20 and might be a clue to develop focused interventions to improve EBF in Embu das Artes.

The intervention was an independent factor associated with lack of MFD and with lack of FA confirming our hypothesis that the use of the manual can improve CF indicators. These results are consistent with a previous study that evaluated a program to promote appropriate CF and found that the intervention program resulted in lower consumption of food considered unhealthy.²¹ Our findings also support the conclusions of a systematic review about the effectiveness of educational interventions on CF indicators that found nutrition training for health workers can improve feeding frequency, energy intake and dietary diversity in infants from 6 months to 2 years.²²

Moreover, simple, valid, and reliable indicators are crucial to track progress and guide investment to improve nutrition and health during the first 2 years of life.²³ In this study, we evaluated important indicators widely used in many studies, such as EBF under 6 months, as recommended by the WHO.²⁴ The CF indicators selected based on the Food Intake Markers proposed by the Brazilian Ministry of Health that have shown high sensibility to assessing infant feeding practices in contexts where malnutrition rates are low, and the prevalence of overweight and obese children is increasing, such as in Brazil.²⁵

In the comparative analyses of the study population, the number of EAAB certification criteria fulfilled (considered to be a quality indicator of EAAB implementation) by PHU was expressively higher in the after-intervention group. This variable reflects the immediate effect of the intervention on health-care team’s practices as well as the level of EAAB implementation. Reinforcing the value of this result, 5 PHU from Embu das Artes received the EAAB certification soon after the study ended. Around the country, despite having many tutors already trained by the Ministry of Health, the number of PHU that have been tutored is very small (there are almost 3 times more tutors than the number of PHU being followed, according to data reported by the Ministry of Health), which makes clear the need to orient/offer guidance to the tutor’s work. In addition, the very small number of PHU certified shows that even if the tutors are performing their work, the certification is not being reached. Thus, our results answer the need presented by the Ministry of Health of an effective manual to provide direction to tutor’s work. It is important to highlight that despite the increasing in the number of EAAB certification criteria fulfilled by PHU and the improvement found in the infant practices, these 2 variables did not present association. This unexpected result leads us to think that the certification criteria should be interpreted as a normative indicator of EAAB implementation but may be not a good marker of the impact of EAAB on population outcomes.

The design can be seen as a weakness in this study. It was not possible to conduct a randomized controlled trial because the EAAB local coordinator decided to implement the EAAB using the manual in all PHU. However, plausibility evaluations design has been considered a good alternative when studying the feasibility of a material such as described in this study.²⁶ Some limitations have to be considered while generalizing the findings presented: (1) mothers have been interviewed at the PHU and this context could have influenced their answers, and (2) the instrument of data collection did not allow to characterize which changes in health workers’ practice contributed to changes in food behavior. On the other hand, it is important to consider that some factors are likely to have underestimated the intervention effects: (1) the short period between the intervention and the evaluation; (2) in 2 PHU there were no Cas implemented according to the manual recommendation; (3) only 7 PHU conducted all the 5 CAs, and (iv) the small number of health-care workers involved in the CAs.

Despite these limitations, several strengths reinforce our results. It can be characterized as a plausibility evaluation because the data allow to evaluate whether the observed effect was due to the intervention or to external factors.²⁷ By considering the conceptual framework built for each outcome, the effect of the manual on the indicators was adjusted for the characteristics as described. In addition, the multilevel analysis is based on the concept that individuals interact in social contexts in which they are involved and the characteristics of these social groups are influenced by the individuals that compose them.²⁸ Individual variables were included in the model following a hierarchical modeling strategy. In this approach, variables are placed into hierarchies according to a previously established conceptual framework based on the influence that they would have on the outcome of interests, then the choice of criteria for selecting variables stretches beyond purely statistical considerations.²⁹

Conclusions

This article has highlighted the importance of the use of a continuing education strategy grounded on a problem-based learning methodology in order to produce improvements in CF practices in primary health care. Despite innumerable interventions attempting to improve the quality of infant feeding, the small advances or even any stagnation in epidemiological scenarios makes clear the challenges of scaling-up effective interventions. In this perspective, the EAAB differs from other programs as it is not a punctual or formal course, but a strategy of continuing education focused on changes in health workers’ practice to promote breastfeeding and CF. To contribute in this process, the manual consists of a tool that has the potential to be adapted for different contexts and easy to disclose to all EAAB tutors in the country. This feature reinforces the strong potential of this material to strengthen the EAAB implementation.

Footnotes

Acknowledgments

This study was fully supported by State Health Department of Mato Grosso, Brazil, and partially supported by Capes Foundation, Ministry of Education, Brazil. We wish to thank the authorities at the Health Department of Embu das Artes and the staff of the Brazilian Ministry of Health by supporting the research project. Special thanks to Lucimeire de Sales Magalhães Brockveld for her unconditional support with the data collection and to Regina Tomie Ivata Bernal for her assistance with the sample size calculation.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Gláubia Rocha Barbosa Relvas, PhD

References

Langley-Evans

. Nutrition in early life and the programming of adult disease: a review. J Hum Nutr Diet. 2015;28(s1):1–14. doi:10.1111/jhn.12212.

Rollins

Bhandari

Hajeebhoy

, et al.

Why invest, and what it will take to improve breastfeeding practices?

Lancet. 2016;387(10017):491–504. doi:10.1016/S0140-6736(15)01044-2.

World Health Organization [WHO]. Infant And Young Child Nutrition. Geneva, Switzerland: World Health Organization [WHO]; 2001.

Horta

Bahl

Martines

Victora

. Evidence On The Long-Term Effects Of Breastfeeding: Systematic Reviews And Meta-Analyses. Geneva, Switzerland: World Health Organization; 2007.

Mameli

Mazzantini

Zuccotti

. Nutrition in the first 1000 days: the origin of childhood obesity. Int J Environ Res Public Health. 2016;13(9):E838. doi:10.3390/ijerph13090838.

Nicklaus

Remy

. Early origins of overeating: tracking between early food habits and later eating patterns. Curr Obes Rep. 2013;2(2):179–184. doi:10.1007/s13679-013-0055-x.

Pearce

Taylor

Langley-Evans

. Timing of the introduction of complementary feeding and risk of childhood obesity: a systematic review. Int J Obes. 2013;37:1295–1306. doi:10.1038/ijo.2013.99.

Bortolini

Gubert

Santos

LMP

. Food consumption Brazilian children by 6 to 59 months of age. Cad Saude Publica. 2012;28(9):1759–1771.

Issaka

Agho

Burns

Page

Dibley

. Determinants of inadequate complementary feeding practices among children aged 6–23 months in Ghana. Public Health Nutr. 2015;18(04):669–678. doi:10.1017/S1368980014000834.

10.

Kavle

LaCroix

Dau

Engmann

. Addressing barriers to exclusive breast-feeding in low- and middle-income countries: a systematic review and programmatic implications. Public Health Nutr. 2017;(1):1–15. doi:10.1017/S1368980017002531.

11.

Victora

Bahl

Barros

AJD

, et al. Breastfeeding in the 21st century: epidemiology, mechanisms, and lifelong effect. Lancet. 2016;387(10017):475–490. doi:10.1016/S0140-6736(15)01024-7.

12.

Brasil. Estratégia Nacional Para Promoção Do Aleitamento Materno E Alimentação Complementar Saudável No Sistema Único De Saúde: Manual De Implementação. Brasília: Ministério da Saúde; 2015.

13.

Rogers

. Program theory: Not whether programs work but how they work. In E. Models (Edd.), S.D. I., M.F.G. & K. T. Boston, MA: Kluwer Academic; 2000.

14.

Brockveld L de

. Promoção, Proteção E Apoio Ao Aleitamento Materno Na Última Década (2002-2012) No Município De Embu Das Artes, SP: Um Estudo De Caso [Dissertation]. . São Paulo, Brazil: Secretaria de Estado de Saúde de São Paulo - Instituto de Saúde; 2013.

15.

Woodward

M (Mark)

. Epidemiology: Study Design And Data Analysis. Boca Raton, FL: Chapman & Hall/CRC; 2005.

16.

Brasil. Orientações Para Avaliação De Marcadores De Consumo Alimentar Na Atenção Básica. Brasília: Ministério da Saúde; 2015.

17.

Barros

Hirakata

. Alternatives for logistic regression in cross-sectional studies: an empirical comparison of models that directly estimate the prevalence ratio. BMC Med Res Methodol. 2003;3:21.

18.

Snijders

Bosker

. Multilevel Analysis: An Introduction To Basic And Advanced Multilevel Modeling. 2nd ed. London, England: Sage; 2012.

19.

Boccolini

Boccolini P de

Monteiro

Venancio

Giugliani

ERJ

. Breastfeeding indicators trends in Brazil for three decades. Rev Saude Publica. 2017;51:108. doi:10.11606/S1518-8787.2017051000029.

20.

Buccini G dos

Pérez-Escamilla

Paulino

Araújo

Venancio

. Pacifier use and interruption of exclusive breastfeeding: systematic review and meta-analysis. Matern Child Nutr. 2017;13(3):1–19. doi:10.1111/mcn.12384.

21.

Baldissera

Issler

RMS

Giugliani

ERJ

. Efetividade da Estratégia Nacional para Alimentação Complementar Saudável na melhoria da alimentação complementar de lactentes em um município do Sul do Brasil. Cad Saude Publica. 2016;32(9). doi:10.1590/0102-311X00101315.

22.

Sunguya

Poudel

Mlunde

, et al. Effectiveness of nutrition training of health workers toward improving caregivers’ feeding practices for children aged six months to two years: a systematic review. Nutr J. 2013;12:1. doi:10.1186/1475-2891-12-66.

23.

Arimond

Daelmans

Dewey

, Steering Team of the Working Group on Infant and Young Child Feeding Indicators. Indicators for feeding practices in children. Lancet. 2008;371(9612):541–542.

24.

World Health Organization. Indicators For Assessing Infant And Young Child Feeding Practices: Conclusions Of A Consensus Meeting Held 6. Washington, DC: World Health Organization; 2008.

25.

Saldan

Venancio

Saldiva

SRDM

de Mello

. Proposal of indicators to evaluate complementary feeding based on World Health Organization indicators. Nurs Health Sci. 2016;18(3):334–341. doi:10.1111/nhs.12273.

26.

Victora

Habicht

Bryce

. Evidence-based public health: moving beyond randomized trials. Am J Public Heal J Public Heal. 2004;94(3):400–405.

27.

Habicht

Victora

Vaughan

. Evaluation designs for adequacy, plausibility and probability of public health programme. Int J Epidemiol. 1999;28(1):10–18.

28.

Diez-Roux

. Bringing context back into epidemiology: variables and fallacies in multilevel analysis. Am J Public Health. 1998. doi:10.2105/AJPH.88.2.216.

29.

Victora

Huttly

Fuchs

Olinto

. The role of conceptual frameworks in epidemiological analysis: a hierarchical approach. Int J Epidemiol. 1997;26(1):224–227. doi:10.1093/ije/26.1.224.

Effectiveness of an Educational Manual to Promote Infant Feeding Practices in Primary Health Care

Abstract

Objective:

Methods:

Results:

Conclusions:

Keywords

Introduction

Methods

Design and Study Setting

Intervention

Participants

Data Collection

Study Variables

Outcomes

Independent variable

Covariables

Data Analysis

Results

Discussion

Conclusions

Footnotes

Acknowledgments

Declaration of Conflicting Interests

Funding

ORCID iD

References