Práticas de atenção hospitalar ao recém-nascido saudável no Brasil Clinical practices in the hospital care of healthy newborn infant in Brazil

The aim of this study was to evaluate the care of healthy full-term newborns and to identify variations in childbirth care and practices in the first hour of life. We used data from the Birth in Brazil survey. Unadjusted and adjusted odds ratio (OR) of hospital-delivered care for the mother and during childbirth were estimated for the following outcomes: upper airways and gastric aspiration, use of inhaled oxygen, use of incubator, skin-to-skin contact after birth, rooming-in and breastfeeding in the delivery room and within the first hour of life. We observed wide variations in the care of healthy full-term newborn in the delivery room. Practices considered inadequate, such as use of inhaled oxygen, (9.5%) aspiration of airways (71.1%) and gastric suctioning (39.7%), and the use of incubator (8.8%) were excessively used. Breastfeeding in the delivery room was low (16%), even when the Baby-Friendly Hospital Initiative had been implemented (24%). The results suggest poor knowledge and compliance by health practitioners to good clinical practice. Such noncompliance was probably not due to the differences in resources, since most births take place in hospitals where the necessary resources are available. Newborn Infant; Medical Assistance; Child Care Resumo O objetivo do estudo foi avaliar o cuidado ao recém-nascido saudável a termo e identificar variações nesse cuidado no atendimento ao parto e na primeira hora de vida. Utilizou-se a base de dados da pesquisa Nascer no Brasil. Foram estimadas as razões de produtos cruzados OR brutas e ajustadas entre as características do hospital, maternas e de assistência ao parto com os desfechos: aspiração de vias aéreas e gástrica, uso do oxigênio inalatório, uso de incubadora, contato pele a pele, alojamento conjunto e oferta do seio materno na sala de parto e na primeira hora de vida. Foi observada grande variação das práticas usadas na assistência ao recém-nascido a termo na sala de parto. Práticas consideradas inadequadas como uso de oxigênio inalatório (9,5%), aspiração de vias aéreas (71,1%) e gástrica (39,7%) e uso de incubadora (8,8%) foram excessivamente usadas. A ida ao seio na sala de parto foi considerada baixa (16,1%), mesmo nos hospitais com título de Hospital Amigo da Criança (24%). Esses resultados sugerem baixos níveis de conhecimento e aderência às boas práticas clínicas. Recém-Nascidos; Assistência Médica; Cuidado da Criança http://dx.doi.org/10.1590/0102-311X00145213 Cad. Saúde Pública, Rio de Janeiro, 30 Sup:S128-S139, 2014 ARTIGO ARTICLE S128 PRÁTICAS NA ATENÇÃO AO RECÉM-NASCIDO S129 Cad. Saúde Pública, Rio de Janeiro, 30 Sup:S128-S139, 2014 Introdução O atendimento ao recém-nascido a termo saudável na sala de parto tem sofrido uma série de modificações nos últimos anos e um dos objetivos é diminuir o excesso de intervenções a ele aplicadas logo após o nascimento. Há evidências de que a maioria das intervenções praticadas é desnecessária e algumas delas podem ser, até mesmo, prejudiciais 1,2,3. Além disto, sabe-se que o excesso de intervenções na sala de parto afeta a interação mãe/ bebê. Tal interação precoce influencia a fisiologia e o comportamento de ambos. O contato corporal mãe/bebê ajuda a regular a temperatura do recém-nascido, a manutenção do equilíbrio acidobásico, o ajuste da respiração e do choro, e promove o comportamento cuidador materno 4. Da mesma forma, o bebê pode regular – ou seja, aumentar – a atenção da mãe para suas necessidades, influenciar o início e a manutenção do aleitamento materno e estimular a liberação de hormônios do trato gastrointestinal, resultando em melhor aproveitamento de calorias ingeridas. Os efeitos de algumas dessas situações podem ainda ser detectadas meses mais tarde 4,5. Portanto, tem sido recomendado que práticas dirigidas a outras questões não ligadas exclusivamente à sobrevivência precisam ser incorporadas aos cuidados de rotina do recém-nascido 7. Essas práticas incluem clampeamento tardio do cordão umbilical, contato pele a pele imediato mãe-bebê, início precoce do aleitamento materno, método canguru, entre outras. Entretanto, a utilização dessas práticas na assistência ao parto e ao recém-nascido sofre grande variação entre unidades de saúde no mundo, coexistindo ainda com desigualdades geográficas e sociais, diferenças na expertise médica e características demográficas dos pacientes 8,9,10. Tais práticas podem afetar a qualidade do cuidado ofertado, alterando os resultados neonatais imediatos e ao longo da vida 11,12. A variação nas práticas adotadas no cuidado ao recém-nascido saudável tem sido discutida em várias situações, com diferentes impactos sobre a saúde do recém-nascido. Dados publicados relatam uma enorme disparidade entre a base da evidência científica atual e a prática clínica. Além disso, os clínicos podem não responder à evidência por meio de implementação de mudanças nas políticas e práticas. Isso pode ocorrer por causa da falta de conhecimento ou desacordo com as recomendações, mas especialmente por causa de barreiras locais à implantação 10,13,14,15,16,17. O objetivo deste estudo foi avaliar as práticas adotadas e identificar se houve variações no atendimento ao recém-nascido saudável a termo na sala de parto e na primeira hora de vida.


Abstract
The aim of this study was to evaluate the care of healthy full-term newborns and to identify variations in childbirth care and practices in the first hour of life.We used data from the Birth in Brazil survey.Unadjusted and adjusted odds ratio (OR) of hospital-delivered care for the mother and during childbirth were estimated for the following outcomes: upper airways and gastric aspiration, use of inhaled oxygen, use of incubator, skin-to-skin contact after birth, rooming-in and breastfeeding in the delivery room and within the first hour of life.We observed wide variations in the care of healthy full-term newborn in the delivery room.Practices considered inadequate, such as use of inhaled oxygen, (9.5%) aspiration of airways (71.1%) and gastric suctioning (39.7%), and the use of incubator (8.8%) were excessively used.Breastfeeding in the delivery room was low (16%), even when the Baby-Friendly Hospital Initiative had been implemented (24%).The results suggest poor knowledge and compliance by health practitioners to good clinical practice.Such noncompliance was probably not due to the differences in resources, since most births take place in hospitals where the necessary resources are available.

Introduction
Medical care of the full-term, healthy newborn in the delivery room has undergone numerous changes over the past few years, one of the purposes being to decrease the excess of interventions performed on the newborn.There are evidences that most of the performed interventions are unnecessary, and some may even be harmful. 1,2,3.
In addition, it is known that the excess of interventions in the delivery room affects the mother/baby interaction.Such early interaction influences the physiology and behavior of both of them.The mother/baby body contact helps regulate the body temperature of the newborn, the maintenance of the acid-base balance, the adjustments for respiration and crying, and promotes maternal-care behavior 4 .Similarly, the baby may regulate, i.e., increase, the mother's attention for their needs, influence the beginning and maintenance of breastfeeding, and stimulate the release of gastrointestinal hormones, leading to a better use of the calories ingested.The effects of some of these situations may still be seen months later 4,5 .
It has thus been recommended that practices addressing issues other than just survival should be incorporated into the routine care of the newborn 7 .Such practices include late umbilical cord-clamping, immediate skin-to-skin contact between mother and baby, early initiation of breastfeeding, kangaroo care, among others.The use of such practices in birthcare, however, varies greatly among health-care facilities around the world, along with geographic and social inequalities, differences in medical expertise, and demographic characteristics of the patients 8,9,10 .These practices may affect the quality of the care provided, and change immediate and life-long outcomes 11,12 .
Variations in the practices used for the care of healthy newborns have been addressed in different situations, with different impacts on their health.Published data reports huge discrepancies between the existing scientific evidences and the medical practices.Furthermore, practitioners may not respond to evidences by implementing changes in policies and practices.This may be due to lack of knowledge or disagreement with the recommendations, but particularly due to local barriers for implementation. 10,13,14,15,16,17.
The goal of this study was to assess the practices performed, and to identify variations in the delivery of care for the full-term, healthy newborn in the delivery room and in the first hour of life.

Methodology
Birth in Brazil is a national hospital-based study with postpartum women and their newborns, carried out between February 2011 and October 2012 18 .The sample was selected in three stages.The first included hospitals with 500 or more deliveries/year, stratified by the five macro-regions of Brazil, location (capital city or not), and type of hospital (private, public, and mixed).The second included days (at least seven days for each hospital), and the third included postpartum women.In each of the 266 hospitals of the sample, 90 postpartum women were interviewed, a total of 23,940 subjects.Further information of the sample design is presented on Vasconcellos et al. 19 .
In the first stage of the study, face-to-face interviews were held with the postpartum women during their hospital stay, data about the woman and the newborn were collected from their medical chart, and pictures of the pre-natal cards of the woman were taken.Interviews over the phone were held before six months and at twelve months after delivery, to collect information about maternal and newborn outcomes.Detailed information on data collection is reported in do Carmo Leal et al. 18 .
The study included only variations in care of full-term births (≥ 37 weeks of pregnancy), with birth weight ≥ 2,500g, 1-minute Apgar score ≥ 7, and of single pregnancy.Exclusion criteria included HIV-positive mothers, newborns with congenital malformations, (reported by the mother or recorded in the maternal or the newborn chart), and need of positive pressure ventilation in the first minutes of life.With inclusion and exclusion criteria applied, the actual sample was of 18,639 newborns.All analyses considered took into consideration the design of the sample, and the figures presented on the Tables are adjusted by the sample weight 18 .
As the number of newborns eligible for this analysis was smaller than the total of the sample, post-hoc sample size calculations were made.Considering a 50% prevalence of newborns who received an intervention during birth and a 5% significance level, the smallest sample used in the article had a 90% power to detect differences of at least 2%.
Two groups of outcome variables of the care provided to the healthy newborn were considered: the first group included improper practices (aspiration of airways and gastric suctioning, use of inhaled oxygen, and use of incubator 1,2,3 ), and the second group considered the good medical practices (skin-to-skin contact immediately after birth, offering the breast to the newborn in the delivery room, breastfeeding within the first hour of life, and rooming-in straight from the delivery room 6,7,10 ).Breastfeeding in the first hour of life, in this study, means the mother offering her breast until 59 minutes after delivery.
The independent variables used to assess different medical practices in the care of the healthy newborn included: region of the country (North, Northeast, Central, Southeast or South), location (capital city or not), and Baby-Friendly accreditation (yes or no) as a characteristic of the hospital; maternal age (< 20 years, 20-34, ≥ 35), schooling (incomplete primary education, complete primary education, complete secondary education, complete higher education), skin color as mentioned by the woman (white, black, brown, yellow/indians), and number of previous deliveries (primiparus, 1 to 2 previous deliveries, three and more previous deliveries) as maternal characteristics; presence of a companion during delivery (yes, no), type of delivery (vaginal, c-section), and source of payment for childbirth (public or private) as birthcare characteristics.It was considered of private funding the childbirth that occurred in private-only facilities, paid by private health plans or out-of-pocket, or those that took place in a public or mixed facility and were paid by a health plan.The remainder was classified as public-funded birthcare.
All the selected outcome and independent variables were collected from the information recorded on hospital charts and validated by a questionnaire applied to the women.
Bivariate analyses were performed with the independent and the outcome variables.To identify variations, the proportions and their respective 95% confidence intervals (95%CI) for the interventions and good practices were calculated according to the independent variables.For the multivariate analysis, logistic regression techniques were used.All independent variables were included in the multivariate model.The adjusted odds ratio (OR) (and respective 95%CI) were estimated for all exposure variables.Interactions among the investigated variables were tested.
The study was approved by the Ethics Research Committee of the Oswaldo Cruz Foundation, and by the ethics committees of the participating institutions.All postpartum women who were subjects in this study signed the Informed Consent form.

Results
There was a high proportion of aspiration of the upper airways, ranging from 62.5% in the Northeast to 76.8% in the Southeast, as well as gastric suctioning, which ranged from 33.8% in the Northeast to 47.8% in the Central Region.The use of inhaled oxygen (8.8%) and incubator (8.7%) was also high, given the low risk presented by these newborns.The proportion of interventions was higher in hospitals not accredited by the Baby-Friendly, in childbirths paid by the private sector, or performed in women who had higher level of education, were white, primiparous, and in those whom a c-section was performed (Table 1).
The skin-to-skin contact of the mother with the newborn immediately after birth was more frequent in the Southern Region (32.5%), as well as offering the breast in the delivery room (22.4%).However, the proportion of childbirths in which the breast is offered in the delivery room is still low in all regions of Brazil (16.1%), with the lowest proportion found in the Northeastern region (11.5%).In hospitals with Baby-Friendly accreditation, the offering of the breast in the delivery room was significantly higher, but still in a low proportion (24%).Separating the baby from the mother also varied significantly among the regions of Brazil.In the Northern Region, 87.3% of the newborns roomed-in with the mothers, whereas in the Southeastern Region this proportion was of 61.4% only.Newborns from vaginal delivery and in hospitals of the Brazilian Unified National Health System (SUS) had a significantly smaller chance of being placed away from the mother after delivery; the same goes for adolescent, lower educated, Indian, and multiparous women, and in deliveries in hospitals located in capital cities, or accredited with the Baby-Friendly Initiative (Table 2).
In the adjusted analysis, the Northeastern Region was a protective variable for aspiration of the upper airways.Gastric suction, in turn, was less frequent in cities of the countryside.The use of inhaled O2 was less frequent in the Northern and Northeastern regions, and in newborns from vaginal deliveries.Lower use of incubators was associated to deliveries in the Central Region, vaginal delivery, and public funding of childbirth (Table 3).
After adjustment for confounders, the variables that were more significantly associated to skin-to-skin contact between mother and baby immediately after birth included delivery in a Baby-Friendly accredited hospital, the presence of a companion during delivery, and vaginal delivery.Women whose delivery was in the countryside, of lower education, and whose birthcare was paid by the public sector had less chance of early skin-to-skin contact.The offering of the maternal breast in the delivery room was associated to being born in a Baby-Friendly accredited hospital, the presence of a companion during delivery, and   vaginal delivery.Women in the Northeastern Region were the ones who less offered their breast to the newborns in the delivery room.In terms of rooming-in, it was more frequent in the Northern Region of the country.In addition, being born in a hospital where the Baby-Friendly Initiative was already implemented, the presence of a companion during delivery, vaginal delivery, and public funding of childbirth were also associated to this outcome (Table 4).The variables associated to higher chance of breastfeeding within the first hour of life of the baby were being born in the Northern or Central region, in Baby-Friendly accredited hospitals, the presence of a companion during delivery, vaginal delivery, and childbirth funded by the public health system.On the other hand, the proportion of breastfeeding in the first hour of life was lower among women age 35 or more, or primiparous (Table 4).
Interactions among the variables were tested, and the only interaction found in outcomes was Baby-Friendly accreditation in childbirth paid by the private sector for the variables breastfeeding in the first hour of life, and early skin-toskin contact.

Discussion
There was high variation among the practices performed in the care of healthy newborns, in the delivery room, in Brazil.Practices considered unsuitable were still broadly performed, whereas some of the considered good practices were left aside.Variations in medical practices have been described since 1938, when Glover 20 reported differences in tonsillectomy rates in different geographic areas of the United Kingdom.Studies initiated by Wennberg & Gittelson in the 1970s observed the presence of variations in medical practices for different surgical procedures, and, even after adjustments for age, income, prevalence of diseases, and demographic characteristics of patients, such variations persisted for a number of practices and therapies 15,16,17 .In neonatal care, the presence of variations in delivery is seen in the more common practices 21,22,23,24 , in addition to those analyzed in this study.
Since 2010, the use of oxygen in healthy newborns, in the delivery room, is being considered unsuitable, and yet this is still done.The same goes for aspiration of upper airways and gastric suction.The medical protocols based on the best available evidences recommend that healthy newborns should be assisted close to their mothers, and do not require being submitted to such procedures 1 .Notwithstanding, in this study, there was a high proportion of these practices (use of oxygen, upper airways and gastric aspiration) performed in a group of newborns that did not need them.Such high proportions of unwarranted procedures occur in all regions of the country, regardless of the source of payment.In the adjusted analysis, delivery in the Northeastern Region is a protective factor for upper airways aspiration; delivery in a non-capital city protects from gastric suctioning; and being in the Northern and Northeastern regions, and having a vaginal delivery are protective factors from the use of inhaled oxygen.The other exposure factors were not significantly associated to these practices, which shows that these are largely disseminated, and that new available evidences are still unknown by most practitioners.
For many years, training of human resources for the provision of care in the delivery room recommended the use of these practices, which still are largely employed in birthcare.To move away from prevailing techniques and technology is difficult, and more so to change behaviors and practices of health practitioners 16,17 .The reasons for non-compliance with good practices were not the subject of this study.However, non-compliance was high, and not related to structural reasons, differently from what is found in most studies on the quality of care 25,26 .
Variations in the medical practices have been attributed to a number of factors, among them the option of the individual practitioner, which is known as "practice style" 27 .Other factors can, however, contribute to these variations, such as characteristics of the patients, socioeconomic aspects, cultural or leadership issues, among others 28,29 .In this study, for instance, vaginal delivery is significantly protects from the use of inhaled oxygen,; the same goes for births in the Northern and Northeastern regions.Thus, the "practice style" alone would not account for the variations.There was also a high proportion of the use of incubator, which implies separating the mother from the baby, particularly in wealthier areas and populations.This reflects a contradiction between access to best practices and social and economic class: the wealthier the population, the more unwarranted practices were performed.The type of delivery also contributed significantly for separating mother from baby, with vaginal delivery showing a protective effect from separation.
Another important aspect considered in offering the newborn the maternal breast in the delivery room, and the mother having contact with the baby in its first hour of life.Once again, the type of delivery was instrumental for this practice.Newborns from vaginal delivery had a higher chance of being breastfed in the deliv- ery room; the same goes for newborns whose mothers had a companion present in the delivery room, or in hospitals with Baby-Friendly accreditation.Variations in the breastfeeding-inthe-delivery-room practice ranged from 9.8% to 22.8% for the type of birth, and from 11.9% to 24.3% for the presence of a companion in the delivery room.Maternal age, number of previous children, schooling, and source of payment did not influence such practice.Other authors also found a strong association between type of delivery and offering the maternal breast to the baby in the delivery room 30,31 .However, differently from the study by Boccolini et al. 30 , in this investigation there were no differences for this variable in relation to the source of payment (public or private).There is a 10-year gap between the data collection of our study and Boccolini's 31 .The improvement observed in Brazil regarding this practice is probably due to compliance with the 4 th -step recommendations of the Baby-Friendly Hospital Initiative 33 .The 4 th step of the Baby-Friendly Hospital Initiative implementation is fulfilled when skin-toskin contact between newborn and mother occurs within thirty minutes from birth, and breastfeeding takes place within one hour after delivery 32 .In this study, even though the skin-to-skin contact reported was around 28% in the whole country, breastfeeding in the delivery room rates were low, ranging from 11.5% to 22.4% among the regions Brazil.The highest breastfeeding-in-thedelivery-room rates are found in the Southern region (22,4%).The Northeast of the country is the area with the highest number of obstetric beds in hospitals of the public health system accredited by the Baby-Friendly Initiative, 32 and was associated to non-breastfeeding in the delivery room.This was an unexpected finding of our study, as birth in a Baby-Friendly accredited hospital was significantly associated to breastfeeding in the delivery room.However, in the Northeast of the country, only 41.3% of healthy newborns were breastfed in their first hour of life; this means that not only newborns were not breastfed in the delivery room, they were deprived from the maternal breast in their first hour of life.These results differ from those of the 2006 National Survey on Demographics and Health (PNDS 2006) 33 .Even though the methodology of data collection was different, the PNDS found that 43% of newborns were breastfed in the first hour of life, particularly in the countryside, and in the Northern and Northeastern regions of the country 33 .
The variables that were significantly associated to good practices in the delivery room, after adjustment for confounders, were vaginal delivery, birth in a Baby-Friendly accredited hospital, and presence of a companion in the delivery room.Our study found that vaginal delivery warranted most of the good practices investigated: early skin-to-skin contact, offering of maternal breast in the delivery room, breastfeeding in the first hour of life, and rooming-in with the mother.Birth in a Baby-Friendly accredited hospital also favored good practices in relation to skin-toskin contact and breastfeeding, in the adjusted analysis.In 2009, the set of recommendations of the Baby-Friendly Initiative was reviewed by the World Health Organization, and some more comprehensive approaches on good practices that targeted the mother were included 34 .Some studies have shown an association between Baby-Friendly accreditation and lower proportion of c-sections 35,36 .Therefore, the type of delivery (vaginal), and birth in a Baby-Friendly accredited hospital should, together, have a protective effect for the newborn being exposed to good practices.Public payment of birth also favors rooming-in and breast offering in the first hour of life, but not for skin-to-skin contact and breast offering in the delivery room.
One of the limitations of this study is that it did not allow the assessment of the reasons for the non-compliance to good practices or for all the variations in medical practices.Another limitation is that, despite the interaction between delivery in a Baby-Friendly accredited hospital and private source of payment, among all hospitals of the sample only one presented these variables with a distinct care model, with goals associated to the use of good practices in childbirth care; in this case, the environment likely fostered the use of good practices.
To conclude, in Brazil, where most deliveries take place in a hospital, there are great variations in the routine care of healthy newborns immediately after birth.The results of this study show that new guidelines for the care of healthy newborns were not incorporated in the medical practice.Non-compliance is not due to differences in the resources, as the deliveries took place in hospitals where the necessary resources were available.
Recién Nacido; Asistencia Médica; Cuidado da Criança ContributorsM.E. L. Moreira, S. G. N. Gama, A. A. M. Silva, S. Lansky and M. C. Leal participated in data design, analysis and interpretation, and in the writing and approval of the final version of the article.A. P. E. Pereira and R. S. Pinheiro contributed to data analysis and interpretation, and in the writing of the article and approval of its final version. A. C. Gonçalves contributed to data interpretation, and in the writing and approval of the final version of the article.

Table 1
Interventions considered unsuitable performed in healthy newborns according to sociodemographic characteristics: bivariate analysis.Brazil, 2011-2012.
* Figures corrected by the sample weight.

Table 4
Multivariate logistic regression for good practices performed in healthy newborns.Brazil, 2011-2012.