Abstract

The aim of this cross-sectional hospital-based study of 7,845 pregnancies was to analyze deaths of women hospitalized for childbirth and abortion, and fetal and neonatal deaths, in public hospitals in the cities of São Paulo, Rio de Janeiro and Niteroi (RJ), Brazil, in 2011. Outcomes of the pregnancies were: one maternal death, 498 abortions, 65 fetal deaths, 44 neonatal deaths and 7,291 infant survivors. Data were collected through interviews, medical records and the women's pregnancy records, and from the Mortality Information System (SIM). The study population was described and kappa coefficients of causes of death (from the SIM, and certified by research) and mortality health indicators were estimated. The maternal mortality ratio was 13.6 per 100,000 live births (LB), the fetal death rate was 8.8‰ births and the neonatal mortality rate was 6.0‰ LB. The drug most used to induce abortion was Misoprostol. The main causes of fetal and neonatal deaths were respiratory disorders and maternal factors. Congenital syphilis, diabetes and fetal death of unspecified cause were under-reported in the SIM. Kappa coefficients by chapter were 0.70 (neonatal deaths) and 0.54 (stillbirths). Good quality care in reproductive planning, prenatal care, during labor and at birth will result in prevention of deaths.

Key words
Abortion; Fetal death; Neonatal mortality; Maternal mortality; Cause of death

Introduction

Women and their concepti are a group with higher vulnerability, expressed by the risk of illness and death¹1. World Health Organization (WHO). Global Strategy for Women's and Children's health. Geneva: WHO; 2010. [acessado 2016 Jun 27]. Disponível em: http://www.who.int/pmnch/topics/maternal/20100914_gswch_en.pdf?ua=1
http://www.who.int/pmnch/topics/maternal... , the biological determinants of which are exacerbated by socio-economic characteristics and factors in healthcare²2. United Nations Development Programme (UNDP). A social determinants approach to maternal health. [acessado 2016 Jan 14]. Disponível em: http://www.undp.org/content/dam/undp/library/Democratic%20Governance/Discussion%20Paper%20MaternalHealth.pdf
http://www.undp.org/content/dam/undp/lib... .

Maternal death is a rare, tragic and avoidable event³3. Cecatt, JG. Saúde da mulher: enfoque da evidência científica para a prevenção da morbidade e mortalidade materna. Rev Bras Saude Mater Infant 2005; 5(1):9-11.. In 2013, in spite of the progress in maternal-infant health in Brazil, the maternal mortality ratio was 58.1 per 100,000 live births⁴4. Brasil. Ministério da Saúde (MS). Secretaria de Vigilância em Saúde. DATASUS. [acessado 2016 Jan 5]. Disponível em: http://www2.datasus.gov.br/DATASUS/index.php?area=0205
http://www2.datasus.gov.br/DATASUS/index... , not achieving the 75% reduction that is the fifth target in the Millennium Development Goals⁵5. World Health Organization (WHO). Trends in maternal mortality: 1990 to 2015. Estimates by WHO, UNICEF, INFPA, The World Bank and the United Nations Population Division. Geneva: WHO; 2015..

For the conceptus, the risk of death is greater at the beginning of the pregnancy, when the pregnancy itself has not even been recognized or diagnosed⁶6. Adams MM, Alexander GR, Kirby RS, Wingate MS. The continuum of Reproductive Loss from pregnancy through infancy. In: Adams MM, Alexander GR, Kirby RS, Wingate MS. Perinatal Epidemiology for Public Health. São Paulo: Spring Ed.; 2009. p. 155-224.. These miscarriages, frequently reported as irregularities of menstruation, are not usually reported⁶6. Adams MM, Alexander GR, Kirby RS, Wingate MS. The continuum of Reproductive Loss from pregnancy through infancy. In: Adams MM, Alexander GR, Kirby RS, Wingate MS. Perinatal Epidemiology for Public Health. São Paulo: Spring Ed.; 2009. p. 155-224..

The frequency of abortions is also imprecise, even in countries where it is legalized⁷7. Sedgh G, Henshaw S, Singh S, Ahman E, Shah IH. Induced 2. abortion: estimated rates and trends worldwide. Lancet 2007; 370(9595):1338-1345.. In 2010, the National Abortion Survey estimated that 15% of Brazilian women, by the completion of their reproductive life, had had an abortion, requiring hospitalization in 50% of those cases⁸8. Diniz D, Medeiros A. Aborto no Brasil: uma pesquisa domiciliar com técnica de urna. Cien Saude Colet 2010; 15(Supl. 1):959-966.. In spite of the reduction of 26% in the indirect national estimates of induced abortions in women of fertile age from 1995 to 2013, their magnitude is a source of concern⁹9. Monteiro MFG, Adesse L, Drezett J. Atualização das estimativas da magnitude do aborto induzido, taxas por mil mulheres e razões por 100 nascimentos vivos do aborto induzido por faixa etária e grandes regiões. Brasil, 1995 a 2013. Reprod & Clim 2015; 30(1):11-18..

What defines whether losses of the fetus are abortions or fetal deaths are the limits established for analysis in relation to birth weight, gestation age or length of the conceptus¹⁰10. Organização Mundial de Saúde (OMS). Classificação Estatística Internacional de Doenças e Problemas Relacionados à Saúde – 10a Revisão. São Paulo: CBCD/EDUSP; 1995. v I e II.. Brazil adopts the definition of the International Statistical Classification of Diseases and Related Health Problems, 10th Revision (ICD), which proposes the following cut-off values for fetal death: weight at birth, 500 grams; gestation age, 22 weeks; and length, 25 cm¹⁰10. Organização Mundial de Saúde (OMS). Classificação Estatística Internacional de Doenças e Problemas Relacionados à Saúde – 10a Revisão. São Paulo: CBCD/EDUSP; 1995. v I e II.. Lawn et al.¹¹11. Lawn JE, Blencowe H, Waiswa P, Amouzo A, Mathers C, Hogan D, Flenady V, Frøen JF, Qureshi ZU, Calder-wood C, Shiekh S, Jassi FB, You D, McClure M, Matha Mi, Cousens S, for The Lancet Ending Preventable Stillbirths Series study group* with The Lancet Stillbirth Epidemiology investigator group. Stillbirths: rates, risk factors, and acceleration towards 2030. Lancet 2016; 18. recommend that the definition of fetal death, for international comparisons, should use gestation age (28 or more weeks), and propose this change in the 11^th ICD. Fetal death is little understood and studied; it was not included in the Millennium Development Goals, and is practically invisible in world political agendas¹¹11. Lawn JE, Blencowe H, Waiswa P, Amouzo A, Mathers C, Hogan D, Flenady V, Frøen JF, Qureshi ZU, Calder-wood C, Shiekh S, Jassi FB, You D, McClure M, Matha Mi, Cousens S, for The Lancet Ending Preventable Stillbirths Series study group* with The Lancet Stillbirth Epidemiology investigator group. Stillbirths: rates, risk factors, and acceleration towards 2030. Lancet 2016; 18.. However, the rates continue to be high, especially in the countries of low and medium income: the average rate in these countries is estimated to be 28.9 fetal deaths per 1,000 births, varying from 13.6 in Argentina to 56.5, in Pakistan¹²12. McClure EM, Pasha O, Goudar SS, Chomba E, Garces A, Tshefu A, Althabe F, Esamai F, Patel A, Wright LL, Moore J, Kodkany BS, Belizan JM, Saleem S, Derman RJ, Carlo WA, Hambidge KM, Buekens P, Liechty EA, Bose C, Koso-Thomas M, Jobe AH, Goldenberg RL; Global Network Investigators. Epidemiology of stillbirth in low-middle income countries: a Global Network Study. Acta Obstet Gynecol Scand 2011; 90(12):1379-1385.. In 2014, the 67^th World Health Assembly established as a target the reduction of fetal death to 12 or less per 1,000 births in 2030¹¹11. Lawn JE, Blencowe H, Waiswa P, Amouzo A, Mathers C, Hogan D, Flenady V, Frøen JF, Qureshi ZU, Calder-wood C, Shiekh S, Jassi FB, You D, McClure M, Matha Mi, Cousens S, for The Lancet Ending Preventable Stillbirths Series study group* with The Lancet Stillbirth Epidemiology investigator group. Stillbirths: rates, risk factors, and acceleration towards 2030. Lancet 2016; 18..

For live births, the risk of infant death is greatest immediately after the transition to extra-uterine life, and decreases markedly during the first week of life⁵5. World Health Organization (WHO). Trends in maternal mortality: 1990 to 2015. Estimates by WHO, UNICEF, INFPA, The World Bank and the United Nations Population Division. Geneva: WHO; 2015.^,66. Adams MM, Alexander GR, Kirby RS, Wingate MS. The continuum of Reproductive Loss from pregnancy through infancy. In: Adams MM, Alexander GR, Kirby RS, Wingate MS. Perinatal Epidemiology for Public Health. São Paulo: Spring Ed.; 2009. p. 155-224.^,1313. Lansky S, Friche AAL, Silva AAM, Campos D, Bittencourt SDA, Carvalho ML, Frias PG, Cavalcante RS, Cunha AAL. Pesquisa Nascer no Brasil: perfil da mortalidade neonatal e avaliação da assistência à gestante e ao recém-nascido. Cad Saude Publica 2014; 30(Supl. 1):S192-S207.. Considering the similarity of the causes of death and the risk factors in fetal deaths and early neonatal deaths, a variation in the quality of statement in the death certificate is not expected, but systematic errors generate a lesser quality of information for fetal deaths¹⁴14. Lee EJ, Gambatese M, Begier E, Soto A, Das T, Madsen A. Understanding Perinatal Death: A Systematic Analysis of New York City Fetal and Neonatal Death Vital Record Data and Implications for Improvement, 2007–2011. Matern Child Health J 2014; 18(8):1945-1954.. Incompleteness, and low reliability and validity of the causes of death for neonatal and fetal deaths, principally, have been reported in Brazilian studies¹⁵15. Barbeiro FMS, Fonseca SC, Tauffer MG, Ferreira MSS, Silva FP, Ventura PM, Quadros JI. Óbitos fetais no Brasil: revisão sistemática. Rev Saude Publica 2015; 49:22.

16. Barbuscia DM, Rodrigues-Júnior AL. Completude da informação nas declarações de nascido vivo e nas declarações de óbito, neonatal precoce e fetal, da região de Ribeirão Preto, São Paulo, Brasil, 2000-2007. Cad Saude Publica 2011; 27(6):1192-1200.

17. Almeida MF, Alencar GP, Schoeps D, Minuci EG, Silva ZP, Ortiz LP, Novaes HMD, Alencar AP, Raspatini PR, Santos PC. Qualidade das informações registradas nas declarações de óbito fetal em São Paulo, SP. Rev Saude Publica 2011; 45(5):845-53.^-1818. Pedrosa LCDO, Sarinho SW, Ordonha MR. Análise da qualidade da informação sobre causa básica de óbitos neonatais registrados no Sistema de Informações sobre Mortalidade: um estudo para Maceió, Alagoas, Brasil, 2001-2002. Cad Saude Publica, 2007; 23(10):2385-2395..

On the one hand technological advances in healthcare for pregnancy and the neonatal period have brought positive results for perinatal health, on the other hand they have accentuated the inequality in which access to technology is not universal or equitable¹⁹19. Lansky S, França E, Leal MC. Mortalidade perinatal e evitabilidade: revisão da literatura. Rev Saude Publica 2002; 36(6):759-772.. Inequalities in healthcare and socio-demographic inequalities are risk factors common to induced abortions²⁰20. Cecatti JG, Guerra CVQL, Sousa MH, Menezes GMS. Aborto no Brasil: um enfoque demográfico. Rev Bras Ginecol Obstet 2010; 32(3):105-111., fetal deaths¹⁵15. Barbeiro FMS, Fonseca SC, Tauffer MG, Ferreira MSS, Silva FP, Ventura PM, Quadros JI. Óbitos fetais no Brasil: revisão sistemática. Rev Saude Publica 2015; 49:22., neonatal deaths¹³13. Lansky S, Friche AAL, Silva AAM, Campos D, Bittencourt SDA, Carvalho ML, Frias PG, Cavalcante RS, Cunha AAL. Pesquisa Nascer no Brasil: perfil da mortalidade neonatal e avaliação da assistência à gestante e ao recém-nascido. Cad Saude Publica 2014; 30(Supl. 1):S192-S207. and maternal deaths³3. Cecatt, JG. Saúde da mulher: enfoque da evidência científica para a prevenção da morbidade e mortalidade materna. Rev Bras Saude Mater Infant 2005; 5(1):9-11..

Hypertension and hemorrhage have been the most frequent causes of maternal death in Brazil, and analysis of the causes of neonatal and fetal deaths shows maternal morbidities as a growing cause⁴4. Brasil. Ministério da Saúde (MS). Secretaria de Vigilância em Saúde. DATASUS. [acessado 2016 Jan 5]. Disponível em: http://www2.datasus.gov.br/DATASUS/index.php?area=0205
http://www2.datasus.gov.br/DATASUS/index... ^,1818. Pedrosa LCDO, Sarinho SW, Ordonha MR. Análise da qualidade da informação sobre causa básica de óbitos neonatais registrados no Sistema de Informações sobre Mortalidade: um estudo para Maceió, Alagoas, Brasil, 2001-2002. Cad Saude Publica, 2007; 23(10):2385-2395., indicating the interrelationship between these deaths and the interventions necessary for their reduction. During the prenatal period, deaths of mothers and their concepti²¹21. Chaves SC, Cecatti JG, Carroli G, Lumbiganon P, Hogue CJ, Mori R, Zhang J, Jayaratne K, Togoobaatar G, Pileggi-Castro C, Bohren M, Vogel JP, Tunçalp Ö, Oladapo OT, Gülmezoglu AM, Temmerman M, Souza JP. Obstetric transition in the World Health Organization Multicountry Survey on Maternal and Newborn Health: exploring pathways for maternal mortality reduction. Rev Panam Salud Publica 2015; 37(4/5):203-210. can be prevented by screening for and treatment of morbidities such as syphilis, diabetes, high blood pressure²²22. Goldemberg RL, MacClure EM, Bhutta ZGA, Belizán JM, Reddy UM, Rubens GE, Mabeya H, Flenady V, Darmstadt GL; for The Lancet's Stillbirths Series steering committee. Stillbirths: the vision for 2020. Lancet 2011; 377(9779):1798-1805., and by the ability to detect and handle obstructed labor, and reduction of medicalization at the moment of birth.

The objective of this study was to analyze, in maternity facilities of Brazil's Single Health System (Sistema Único de Saúde), maternal deaths, abortions, and fetal and neonatal deaths, in the cities of São Paulo, Rio de Janeiro and Niteroi (Rio de Janeiro State) in 2011, with an emphasis on the causes of deaths.

Methodology

A sectional study was carried out, over 3 months in the second half of 2011, in maternity facilities in Brazil's Single Health System with large numbers of births: four facilities in the city of São Paulo, in São Paulo State (‘the SP center’); and in Rio de Janeiro State (‘the RJ center’), one in Niteroi and one in Rio de Janeiro city. The deaths in hospital that took place during pregnancy, in birth and until discharge from hospital were identified, first, in the fieldwork, and subsequently, by probabilistic record linkage of the research database to the Brazilian Mortality Information System (SIM): (i) in São Paulo, by the Program for Improvement of Mortality Data (PRO-AIM) of the Prefecture of the municipality of São Paulo; and (ii) in Rio de Janeiro and Niteroi, by the Rio de Janeiro State Health Department. The computer program RecLink3²³23. Camargo Júnior KR, Coeli CM. RecLink 3: nova versão do programa que implementa a técnica de associação probabilística de registros (probabilistic record linkage). Cad Saúde Coletiva 2006; 14(2):399-404. was used. Losses in the study were less than 5%.

Interviews were held with the women 12 hours after birth, and 5 hours after abortion, and hospital medical records, the pregnant mother's records and the records of the delivery room – made by nurses and previously trained healthcare students – were consulted.

Study population

Figure 1 shows the flow diagram of the population of the study in relation to the situations of the women and their concepti.

7,845 women were interviewed (5,955 in São Paulo center and 1,890 in Rio de Janeiro center). There were four deaths of women in the SP center, one of a mother after transfer to another facility, and three after discharge, in the late puerperium for reasons not related to the maternity cycle, and not computed as maternal deaths. These deaths were investigated by the Municipal Maternal, Fetal and Infant Death Committee (CMMIF) of São Paulo. No women died in the RJ center.

Fetal deaths and neonatal deaths were analyzed as to time of occurrence: abortions and fetal deaths as to < 22 and ≥ 22 weeks of gestation, respectively, and infant deaths from 0-27 days¹⁰10. Organização Mundial de Saúde (OMS). Classificação Estatística Internacional de Doenças e Problemas Relacionados à Saúde – 10a Revisão. São Paulo: CBCD/EDUSP; 1995. v I e II.. In spite of the definition of fetal deaths (death of a product of conception, before complete expulsion or extraction from the body of the mother, whatever the duration of pregnancy)¹⁰10. Organização Mundial de Saúde (OMS). Classificação Estatística Internacional de Doenças e Problemas Relacionados à Saúde – 10a Revisão. São Paulo: CBCD/EDUSP; 1995. v I e II., we adopted fetal death to represent the fetal component of perinatal mortality¹⁰10. Organização Mundial de Saúde (OMS). Classificação Estatística Internacional de Doenças e Problemas Relacionados à Saúde – 10a Revisão. São Paulo: CBCD/EDUSP; 1995. v I e II., for the purpose of comparison with other studies. An abortion was reclassified as a stillbirth. From visual sources it was possible to identify the infant deaths that took place after the survey, and thus the survivors of the infancy period (364 complete days of life). 26 infant deaths that took place after the discharge from hospital were not analyzed, due to the impossibility of access to the medical records for certification of a new Death Certificate. These deaths were excluded from the analysis of the group of survivors. Of the total of 7,898 concepti, 563 fetal losses were analyzed (498 abortions and 65 fetal deaths), 44 neonatal deaths (33 early and 11 late), and 7,291 surviving infants.

The frequencies of deaths of the concepti were presented according to the moment of death. Abortions were described according to the type (without expulsion of the fetus; spontaneous; intentional, under law or court order; and induced) and by the means used for induced abortions. Fetal and neonatal deaths informed by the SIM were described according to the confirmation by the researchers.

Concepti were described according to the center where the birth or abortion occurred (SP center; RJ center), and maternal socio-demographic characteristics, including age range (< 20; 20-34; ≥ 35 years), schooling (< 8; ≥ 8 years of schooling) and self-reported skin color (‘white’; ‘brown; ‘black’; others), characteristics of the reproductive and pregnancy history, such as number of previous births (none, 1-2 and 3 or more); previous abortion (yes; no; not applicable), previous stillbirths (yes; no; not applicable); and, in binary form (yes; no): self-reported morbidities during pregnancy (high blood pressure, hemorrhage, syphilis); risk behavior during pregnancy (use of tobacco, alcohol, unlawful drugs); presence of the partner; and realization of prenatal care. Additionally, the maternal death was described by nationality, obstetric assistance provided, cause of death and outcome of the product of the pregnancy. Birth weight (BW, in grams) and gestation age (GA, in weeks) were described for the subgroups of the population of the study, except abortion, based on the indicators: proportion of extremely low BW (< 1,000g), very low BW (< 1,500g), low BW (< 2,500g) and early pre-term (< 32 weeks) and pre-term (< 37 weeks). The following records were excluded from the analysis of gestation age (GA) and birth weight (BW): records with inconsistency between the following information²⁴24. Rio de Janeiro (Município). Secretaria de Saúde e Defesa Civil do Município do Rio de Janeiro. Roteiro para crítica dos sistemas de informação SINASC e SIM. Rio de Janeiro: [s.n.]; 2008. - GA< 22 weeks and BW>600g; GA of 22-27 weeks and BW > 1,500g; GA de 37-41 weeks or GA ≥ 42 weeks and BW < 1,500g or > 5,500g. Frequencies of live births (LB) with 5-minute Apgar < 7 were described only for neonatal deaths and surviving births.

Fetal and neonatal deaths were classified by specific causes (three- and fourcharacter categories), groupings and chapters of the ICD¹⁰10. Organização Mundial de Saúde (OMS). Classificação Estatística Internacional de Doenças e Problemas Relacionados à Saúde – 10a Revisão. São Paulo: CBCD/EDUSP; 1995. v I e II.. Where there has been investigation of fetal and neonatal deaths, causes of death refer to those post-investigations recorded in the SIM. Additionally, the investigators, not knowing such information, filled in a new death certificate (in São Paulo, by the general coordinator of the survey, and in Rio de Janeiro State, by the third author of the article), codified the causes of death and selected the underlying cause (fourth author of the article).

The analyses of causes of death were carried out separately for the fetal deaths and the neonatal deaths, and by a certifying source (SIM, and research). The underlying cause of death (CD) was classified by chapter and grouping of the ICD and by specific causes (three- and four-character categories) only for the principal groupings. The changes in the frequency of deaths by chapter and grouping informed by the SIM and certified by the survey by moment of death (fetal/neonatal) were calculated: positive values expressed increases and negative values reduction. The frequency of diagnoses other than the underlying CD was described as accounted, independently of its location in the Death Certificate (same line, or not, of Part I or in Part II), and the reliability of the underlying CD informed by the SIM and certified by researchers was analyzed.

Statistical analyses

The following indicators for hospital death were calculated: maternal death ratio per 100,000 LB, risk of fetal death, fetal deaths per 1,000 births, and risk of neonatal death per 1,000 LB¹⁰10. Organização Mundial de Saúde (OMS). Classificação Estatística Internacional de Doenças e Problemas Relacionados à Saúde – 10a Revisão. São Paulo: CBCD/EDUSP; 1995. v I e II..

The absolute and percentage distributions of the concepti according to socio-economic and demographic characteristics, and reproductive, pregnancy and birth history, were described. To test the hypothesis of homogeneity of proportion a Pearson chi-squared test or a Fisher's exact test were carried out. For ordinal categorical variables, the chi-squared test for linear trend was used.

The frequency distribution and position of the underlying CD by ICD chapter and grouping and by specific cause (three- and four-character) were described only for the principal groupings. The mean, and the absolute frequency, of diagnoses other than the declared underlying CD (SIM and research) were calculated. The McNemar test was carried out to investigate difference between the proportions of death by cause, by ICD chapter and grouping informed by the SIM, and those informed by the researchers. For evaluation of the reliability of the underlying CD (dichotomous variables for the specific cause – yes/no – informed by the SIM and by the research) the percentage agreement was calculated (< 25% - bad; de 25% to 49% - regular; 50% to 74% - good and ≥ 75% - very good)²⁵25. Mello Jorge MHP, Laurenti R, Gotlieb SLD. Avaliação dos Sistemas de Informações em Saúde no Brasil. Cad Saúde Coletiva 2010; 18(1):7-18. and the kappa coefficient (< 0.00 – poor; 0.00-0.20 – superficial; 0.21-0.40 – reasonable; 0.41-0.60 – moderate; 0.61-0.80 – substantial; 0.81-1.00 – almost perfect)²⁶26. Szklo M, Nieto F. Epidemiology, Beyond the Basics. 2nd ed. Boston: Jones and Barllett Publishers, 2007..

Ethical considerations

This study was approved by the Research Ethics Committee of the Collective Health Studies Institute of the Federal University of Rio de Janeiro, the Rio de Janeiro Municipal Health and Civil Defense Department, the Study Centers of each Maternity Unit of Rio de Janeiro State, and the research Ethics Committee of the Public Health Faculty of São Paulo University, and participating hospitals of São Paulo. Informed consent was requested from all the eligible pregnant women, or for those under age 18, their guardians.

Results

The hospital maternal mortality ratio was 13.6 per 100,000 LB and the risk of fetal death was 8.8 per 1,000 births (13.9 per 1,000 in the RJ centers, and 7.1 per 1,000 in the SP centers). The neonatal mortality ratio was 6.0 per 1,000 LB (10.1‰ in RJ centers, 4.7‰ in SP centers).

Study population

The only maternal death was of a woman resident in the municipality of São Paulo, born in Bolivia, age 34, color described as ‘yellow’ and with low level of schooling. She did not live with a partner, her profession was stated as homemaker, she had had five previous pregnancies (one abortion and four live births) and the present pregnancy resulted in fetal death (560 grams and 23 weeks). She had had no prenatal care and reported no morbidity or risk behavior during the pregnancy. The death took place in the puerperium immediately after transfer to a hospital of greater complexity. The underlying cause of the death was other hemorrhages immediately postpartum (ICD code O72) and the associated cause recorded was pulmonary tuberculosis (ICD code A162).

Of the 7,898 concepti, there were 498 abortions. Of these, 29.1% were without expulsion of the embryo; 64.5% spontaneous; 0.4% induced, specified by law or by court of law; 3.2% induced; and 2.8%, unknown. The method most used to directly cause abortion was use of the drug Misoprostol. The proportion of induced abortions in the RJ center (11.9%) was approximately 12 times the proportion for the SP center (1%), while for spontaneous abortion it was 30% higher in the SP center (68.1%) than in the RJ center (50.5%). Of the births, 65% were fetal deaths, and 7,335 live births were followed by 44 hospital neonatal deaths.

Table 1 shows the distribution of the concepti by characteristics of social-demographic, reproductive and pregnancy history, risk behavior, healthcare, and characteristics of newborns.

Concepti were heterogeneous in all the maternal socio-demographic characteristics (p value < 0.05). In the outer ends of the age ranges, 34% of the fetal deaths were to adolescent mothers, and 20% of the abortions were in women over 35. Women with black skin color were a higher proportion among fetal and neonatal deaths than among mothers of survivors (13.5%), reaching 30.2% of neonatal deaths. There was a higher percentage of mothers with low schooling among the fetal deaths (32.3%). There was a predominance of women with a partner among the neonatal deaths and survivors, and absence of partner among abortions and fetal deaths.

Except for previous stillbirths, all the other reproductive characteristics differed by the type of outcome (p value < 0.0001). Among mothers that had abortions, 11.1% had a history of four or more births. Prior abortion was reported more frequently among the subgroups of death than the subgroup of survival.

Maternal high blood pressure and syphilis were more frequent except among the abortions, when hemorrhage prevailed. Use of tobacco, alcohol and drugs were more common among the mothers with neonatal deaths. In this group, no mother stated that she had suffered violence during pregnancy. Among the morbidities and risk behaviors, only high blood pressure and use of alcohol were statistically significant (p < 0.05). Not having prenatal care was most frequent among the cases of abortion (63%). Low birth weight and premature birth were more frequent among fetal and neonatal deaths (p < 0.0001).

As to the moment of occurrence of fetal and neonatal death, reduction in frequency was observed as pregnancy age and chronological age increased (Graph 1). There was relative decline (approximately 87%) in the number of deaths, from highest at the start of pregnancy (498 abortions), to 65 fetal deaths at the twenty-second week of pregnancy, and by 72% from this point to the first day of incomplete life (18 neonatal deaths).

Causes of deaths of the conceptus

The investigation of deaths reported by SIM occurred in 83.1% of the fetal deaths and 84.1% of the neonatal deaths. The principal underlying causes of deaths by ICD chapter were perinatal causes (XVI) followed by congenital malformations (XVII) and infectious and parasitic diseases (I), both among those informed by the SIM and those certified by the researchers. A neonatal death was certified by the researchers as external causes (XX). For the fetal deaths, in the certification by the researchers, as compared to the information from the SIM, the frequency of infectious and parasitic diseases and congenital malformations was higher; and the frequency of perinatal causes was lower; for the neonatal deaths, the frequency of congenital malformations was lower and the frequency of other causes was higher.

Table 2 shows the distribution of fetal and neonatal deaths by underlying CD and number of diagnoses, and also underlying CD by ICD grouping, informed by the SIM and certified by the researchers.

Among fetal deaths, 78% of the underlying CDs informed by SIM were due to respiratory and cardiovascular disorders specific to the perinatal period (P20-P29), exclusively due to intrauterine hypoxia, unspecified (P20.9), and newborn affected by maternal factors and by complications of pregnancy, labor, and delivery (P00-P004), highlighting, in this group, newborn affected by maternal hypertensive disorders (P000) and other forms of placental separation and hemorrhage (P021). After certification by the researchers, a different group was more dominant: other disorders originating in the perinatal period (P90-P97), exclusively due to fetal deaths from unspecified causes (P95), corresponding to undefined causes for stillbirths (Table 2). The number of diagnoses, as well as the underlying CD, in death certificates of the fetal deaths varied between zero and three; in approximately 60% (SIM) and 62.5% (researchers) there was a record only of the underlying CD, principally when classified in the group respiratory and cardiovascular disorders (SIM) and other disorders originated in the perinatal period (researchers) (Table 2). The average of diagnoses other than the underlying CD certified by the researchers was 1.9, and for the SIM, 1.3.

More than half of the causes of neonatal deaths at the level of grouping were also due to specific respiratory and cardiovascular disorders of the perinatal period (P20-P29) and maternal factors and complications of pregnancy, labor, and delivery (P00-P004) when informed in the SIM (54.5%), and 56.8% when certified by the researchers, with only the first and second positions of greatest frequency alternating. Among the neonatal deaths informed by the SIM, all showed at least one diagnosis other than the underlying CD, and in five records this number exceeded four. In the research, 11 neonatal deaths presented only the underlying CD in the death certificate, and the maximum number of diagnoses other than the CD was four. The main groupings, P20-P29 among the neonatal deaths informed by the SIM, and P00-P004 certified by the researchers, presented a larger number of diagnoses other than the underlying CD (Table 2). The average number of diagnoses other than the underlying CD of the neonatal deaths were greater than that for the stillbirths, and the number for the researchers (3.0) was also greater than the number from the SIM (2.7).

Changes in the basic cause of death of concepti

The principal change of specific underlying CD of the fetal deaths was from unspecified intrauterine hypoxia (P209), informed by the SIM (30), to fetal deaths for unspecified causes (P95), certified by the researchers (16). In all the situations in which this change occurred there was no diagnosis other than the basic cause (SIM and researchers). For eight fetal deaths of which the underlying CD informed by the SIM was P209, there was a diagnosis other than underlying CD, six in the grouping maternal factors and complications (P00-P04) and two transitory and specific endocrine and metabolic disorders (P70-74), which became the underlying CD adopted by the researchers. For a fetal death, the underlying CD unspecified intrauterine hypoxia in the SIM changed to early congenital syphilis, symptomatic (A500) by the researchers. Intrauterine hypoxia was not maintained as underlying cause for any of the fetal deaths after certification by the researchers.

The specific changes of causes of neonatal deaths between the SIM and the researchers were more distributed between the ICD codes. However, two changes of underlying CD of neonatal deaths reported in the SIM and certified in the research attracted attention. In one death, unspecific bacterial septicemia (P369) was informed by the SIM and early congenital syphilis, symptomatic (A500) was certified by the researchers. The second change was from the natural cause preterm newborn (other) (P073) informed by the SIM to the unnatural cause, other slipping, tripping and stumbling and falls (W18), certified by the researchers.

Statistically significant changes of underlying CD informed by the SIM and by the researchers occurred only by grouping (P20-P29 and P90-P97) for the fetal deaths (p < 0.01) (Table 2).

Reliability of the basic cause of death of the concepti

The reliability of the neonatal CDs was higher than that of the fetal deaths, it being found that, since the causes of death are more specified, the agreement between the SIM and the researchers diminishes (Table 3). The percentage agreement per chapter of the ICD was very good both for stillborns and for neonatal deaths; the kappa coefficient presented a substantial agreement (0.70) and moderate agreement (0.54) for the neonatal and stillborn deaths, respectively. The neonatal deaths showed moderate agreement (kappa = 0.40) and reasonable agreement at the three-character level (kappa = 0.23) and four-character level (kappa = 0.5). For the fetal deaths, the agreement by grouping and specific cause was superficial (Table 3).

Discussion

This study made it possible to evaluate the whole spectrum of deaths dependent on the occurrence of pregnancy.

There was one maternal death, for 7,335 live births in this study. In absolute numbers, maternal death is a rare event, and possibly the occurrence in the SP center was due more to the frequency of live births (75.8%) than to the demand of risk pregnancy and/or worse obstetric care. In the multi-centric cross-sectional study in 2010/11, in seven participating Brazilian maternity wards, there was one maternal death and the ratio of maternal death was 14.2/100,000 LB²¹21. Chaves SC, Cecatti JG, Carroli G, Lumbiganon P, Hogue CJ, Mori R, Zhang J, Jayaratne K, Togoobaatar G, Pileggi-Castro C, Bohren M, Vogel JP, Tunçalp Ö, Oladapo OT, Gülmezoglu AM, Temmerman M, Souza JP. Obstetric transition in the World Health Organization Multicountry Survey on Maternal and Newborn Health: exploring pathways for maternal mortality reduction. Rev Panam Salud Publica 2015; 37(4/5):203-210., a similar value to that of this study.

The worst scenario for fetal and neonatal death was in the RJ center. Except for one maternity ward in São Paulo, the others are reference facilities for risk pregnancies. It would be taken into account that as well as the positive differences in quality of obstetric and neonatal care between the RJ and SP centers, in the maternity facility of the city of Niterói, which is reference for the II Metropolitan Region of Rio de Janeiro State, a large portion of the mothers admitted are at-risk pregnancies, and come from the most improvised municipalities, increasing the risks of death of women and concepti. The risk of fetal death estimated for this study gave priority to birth weight following the definition of the ICD¹⁰10. Organização Mundial de Saúde (OMS). Classificação Estatística Internacional de Doenças e Problemas Relacionados à Saúde – 10a Revisão. São Paulo: CBCD/EDUSP; 1995. v I e II. and, thus, may be underestimated²⁷27. Froen JF, Gordijn SJ, Abdel-Aleem H, Bergsjo P, Betran A, Duke CW, Fauveau V, Flenady V, Hinderaker SG, Hofmeyr GJ, Jokhio AH, Lawn J, Lumbiganon P, Merialdi M, Pattinson R, Shankar A. Making stillbirths count, making numbers talk - Issues in data collection for stillbirths. BMC Pregnancy and Childbirth 2009; 9:58.. This underestimation may be greater when the frequency of births with restriction of intrauterine growth is high²⁷27. Froen JF, Gordijn SJ, Abdel-Aleem H, Bergsjo P, Betran A, Duke CW, Fauveau V, Flenady V, Hinderaker SG, Hofmeyr GJ, Jokhio AH, Lawn J, Lumbiganon P, Merialdi M, Pattinson R, Shankar A. Making stillbirths count, making numbers talk - Issues in data collection for stillbirths. BMC Pregnancy and Childbirth 2009; 9:58..

The socio-demographic characteristics of mothers who died in this study, as well as the associated cause of death (pulmonary tuberculosis), express conditions of vulnerability²2. United Nations Development Programme (UNDP). A social determinants approach to maternal health. [acessado 2016 Jan 14]. Disponível em: http://www.undp.org/content/dam/undp/library/Democratic%20Governance/Discussion%20Paper%20MaternalHealth.pdf
http://www.undp.org/content/dam/undp/lib... . The certified causes of death, hemorrhage and obstruction of the birth, are the most frequent complications at the moment of birth and are common to maternal, neonatal and fetal death²²22. Goldemberg RL, MacClure EM, Bhutta ZGA, Belizán JM, Reddy UM, Rubens GE, Mabeya H, Flenady V, Darmstadt GL; for The Lancet's Stillbirths Series steering committee. Stillbirths: the vision for 2020. Lancet 2011; 377(9779):1798-1805.. This tragedy resulted in two deaths, the mother and the conceptus, and four orphans. Only improvements in professional and institution care at birth, independently of women's conditions, are capable of significantly reducing the occurrence of maternal deaths³3. Cecatt, JG. Saúde da mulher: enfoque da evidência científica para a prevenção da morbidade e mortalidade materna. Rev Bras Saude Mater Infant 2005; 5(1):9-11..

As pointed out in other studies⁸8. Diniz D, Medeiros A. Aborto no Brasil: uma pesquisa domiciliar com técnica de urna. Cien Saude Colet 2010; 15(Supl. 1):959-966.^,99. Monteiro MFG, Adesse L, Drezett J. Atualização das estimativas da magnitude do aborto induzido, taxas por mil mulheres e razões por 100 nascimentos vivos do aborto induzido por faixa etária e grandes regiões. Brasil, 1995 a 2013. Reprod & Clim 2015; 30(1):11-18.^,2020. Cecatti JG, Guerra CVQL, Sousa MH, Menezes GMS. Aborto no Brasil: um enfoque demográfico. Rev Bras Ginecol Obstet 2010; 32(3):105-111., errors of classification of the type of abortion, due to illegality, result in underestimation of induced abortions, which has compromised knowledge about this type of abortion. In the RJ center, the proportion of induced abortions was greater than in SP, a result that is compatible with the national survey²⁰20. Cecatti JG, Guerra CVQL, Sousa MH, Menezes GMS. Aborto no Brasil: um enfoque demográfico. Rev Bras Ginecol Obstet 2010; 32(3):105-111.. In this study, among the induced abortions, the use of Misoprostol was recorded by almost all the respondents, in spite of this drug being restricted to qualifying hospitals. Illegal selling of Misoprostol favors its adulteration and the risk of sale of presentations with sub-doses or without the active principle compromises its efficacy²⁸28. Diniz D, Medeiros M. Cytotec e Aborto: a polícia, os vendedores e as mulheres. Cien Saude Colet 2012; 17(7):1795-1804., exposing women to risk situations, either due to the access to the medication or to success not being achieved²⁹29. Arilha MM. Misoprostol: percursos, mediações e redes sociais para o acesso ao aborto medicamentoso em contextos de ilegalidade no Estado de São Paulo. Cien Saude Colet 2012; 17(7):1785-1794.. Abortion is an important cause of maternal morbimortality, principally when carried out in unsafe conditions³⁰30. Domingos SRF, Merigui MAB. O aborto como causa de morte materna: um pensar para o cuidado da enfermagem. Esc Anna Nery Rev Enferm 2010; 14 (1):177-181..

Abortion, fetal death, neonatal death and infant survivors were distinct groups, corroborating results of studies on risk factors and mortality¹¹11. Lawn JE, Blencowe H, Waiswa P, Amouzo A, Mathers C, Hogan D, Flenady V, Frøen JF, Qureshi ZU, Calder-wood C, Shiekh S, Jassi FB, You D, McClure M, Matha Mi, Cousens S, for The Lancet Ending Preventable Stillbirths Series study group* with The Lancet Stillbirth Epidemiology investigator group. Stillbirths: rates, risk factors, and acceleration towards 2030. Lancet 2016; 18.

12. McClure EM, Pasha O, Goudar SS, Chomba E, Garces A, Tshefu A, Althabe F, Esamai F, Patel A, Wright LL, Moore J, Kodkany BS, Belizan JM, Saleem S, Derman RJ, Carlo WA, Hambidge KM, Buekens P, Liechty EA, Bose C, Koso-Thomas M, Jobe AH, Goldenberg RL; Global Network Investigators. Epidemiology of stillbirth in low-middle income countries: a Global Network Study. Acta Obstet Gynecol Scand 2011; 90(12):1379-1385.^-1313. Lansky S, Friche AAL, Silva AAM, Campos D, Bittencourt SDA, Carvalho ML, Frias PG, Cavalcante RS, Cunha AAL. Pesquisa Nascer no Brasil: perfil da mortalidade neonatal e avaliação da assistência à gestante e ao recém-nascido. Cad Saude Publica 2014; 30(Supl. 1):S192-S207.^,1515. Barbeiro FMS, Fonseca SC, Tauffer MG, Ferreira MSS, Silva FP, Ventura PM, Quadros JI. Óbitos fetais no Brasil: revisão sistemática. Rev Saude Publica 2015; 49:22.^,2020. Cecatti JG, Guerra CVQL, Sousa MH, Menezes GMS. Aborto no Brasil: um enfoque demográfico. Rev Bras Ginecol Obstet 2010; 32(3):105-111., except for some peculiarities. Concepti show themselves to be homogeneous in relation to the variables prior stillbirths and use of tobacco (p value > 0.20). Prior stillbirth was described as an independent risk factor for fetal deaths¹⁵15. Barbeiro FMS, Fonseca SC, Tauffer MG, Ferreira MSS, Silva FP, Ventura PM, Quadros JI. Óbitos fetais no Brasil: revisão sistemática. Rev Saude Publica 2015; 49:22. and neonatal deaths¹³13. Lansky S, Friche AAL, Silva AAM, Campos D, Bittencourt SDA, Carvalho ML, Frias PG, Cavalcante RS, Cunha AAL. Pesquisa Nascer no Brasil: perfil da mortalidade neonatal e avaliação da assistência à gestante e ao recém-nascido. Cad Saude Publica 2014; 30(Supl. 1):S192-S207. and, not only use of tobacco, but also use of alcohol and drugs, were not associated with fetal deaths¹⁵15. Barbeiro FMS, Fonseca SC, Tauffer MG, Ferreira MSS, Silva FP, Ventura PM, Quadros JI. Óbitos fetais no Brasil: revisão sistemática. Rev Saude Publica 2015; 49:22.. The low gestation age of the abortions could have influenced the access to pre-natal care.

Certain socio-demographic characteristics stood out for their high frequency: examples are adolescence, among fetal deaths; black maternal skin color among fetal and neonatal deaths¹³13. Lansky S, Friche AAL, Silva AAM, Campos D, Bittencourt SDA, Carvalho ML, Frias PG, Cavalcante RS, Cunha AAL. Pesquisa Nascer no Brasil: perfil da mortalidade neonatal e avaliação da assistência à gestante e ao recém-nascido. Cad Saude Publica 2014; 30(Supl. 1):S192-S207. and absence of partner in the cases of abortion or fetal death¹⁵15. Barbeiro FMS, Fonseca SC, Tauffer MG, Ferreira MSS, Silva FP, Ventura PM, Quadros JI. Óbitos fetais no Brasil: revisão sistemática. Rev Saude Publica 2015; 49:22.. These characteristics suggest the influence of social inequality in care and in health outcomes²2. United Nations Development Programme (UNDP). A social determinants approach to maternal health. [acessado 2016 Jan 14]. Disponível em: http://www.undp.org/content/dam/undp/library/Democratic%20Governance/Discussion%20Paper%20MaternalHealth.pdf
http://www.undp.org/content/dam/undp/lib... .

Among the perinatal deaths, fetal deaths predominated (66.3%); neonatal deaths occurred primarily in the first 48 hours, in line with the literature¹1. World Health Organization (WHO). Global Strategy for Women's and Children's health. Geneva: WHO; 2010. [acessado 2016 Jun 27]. Disponível em: http://www.who.int/pmnch/topics/maternal/20100914_gswch_en.pdf?ua=1
http://www.who.int/pmnch/topics/maternal... ^,66. Adams MM, Alexander GR, Kirby RS, Wingate MS. The continuum of Reproductive Loss from pregnancy through infancy. In: Adams MM, Alexander GR, Kirby RS, Wingate MS. Perinatal Epidemiology for Public Health. São Paulo: Spring Ed.; 2009. p. 155-224.^,1313. Lansky S, Friche AAL, Silva AAM, Campos D, Bittencourt SDA, Carvalho ML, Frias PG, Cavalcante RS, Cunha AAL. Pesquisa Nascer no Brasil: perfil da mortalidade neonatal e avaliação da assistência à gestante e ao recém-nascido. Cad Saude Publica 2014; 30(Supl. 1):S192-S207..

In spite of it being obligatory³¹31. Brasil, Diário Oficial da União. N° 10, sexta-feira, 15 de janeiro de 2010. [acessado 2016 Jan 10]. Disponível em: http://pesquisa.in.gov.br/imprensa/jsp/visualiza/index.jsp?jornal=1&pagina=67&data=15/01/2010
http://pesquisa.in.gov.br/imprensa/jsp/v... , more than half of the fetal and neonatal deaths had not been investigated. Investigation of a death makes it possible to qualify the certification of the causes of death. In 2011 investigation was not universal and the health departments established criteria of priority³²32. Rio de Janeiro. Diário Oficial do Município do Rio de Janeiro. Secretaria de Saúde e Defesa Civil do Rio de Janeiro. Resolução SMS n° 1.257 de 12 de fevereiro de 2007. [acessado 2015 Dez 16]. Disponível em: http://doweb.rio.rj.gov.br/ler_pdf.php?edi_id=902&page=19
http://doweb.rio.rj.gov.br/ler_pdf.php?e... ^,3333. São Paulo. Diário Oficial do Estado de São Paulo. N° 37 - DOE de 24/02/11 -Seção 1 - p.34. [acessado 2015 Dez 16]. Disponível em: ftp.saude.sp.gov.br/ftpsessp/bibliote/informe_eletronico/2011/iels.fev.11/Iels37/E_DL-CIB-5_230211.pdf
ftp.saude.sp.gov.br/ftpsessp/bibliote/in... . Similarly, it is expected the certification of causes of death by the researchers should be more qualified than the original certification in the majority of cases. The differences between the certifications by the SIM and the researchers resulted in ordering of the CDs and in the agreement, fundamentally on ICD grouping. However, the underestimation of congenital syphilis (Chapter I) among stillbirths and neonatal deaths, and slips, tumbles, etc. (Chapter XX) in neonatal deaths according to the SIM contributed to weakening the agreement also in terms of chapter.

Disagreements on CD informed by the SIM and by the researchers were greater among fetal deaths than among neonatal deaths, and increased with the specificity of cause (by four-character chapter). This shows on one side the lack of qualification of the professionals in recognizing the causal chain that resulted in death¹⁸18. Pedrosa LCDO, Sarinho SW, Ordonha MR. Análise da qualidade da informação sobre causa básica de óbitos neonatais registrados no Sistema de Informações sobre Mortalidade: um estudo para Maceió, Alagoas, Brasil, 2001-2002. Cad Saude Publica, 2007; 23(10):2385-2395.; and on the other hand, the presence of a complex interrelationship between the various causes¹⁸18. Pedrosa LCDO, Sarinho SW, Ordonha MR. Análise da qualidade da informação sobre causa básica de óbitos neonatais registrados no Sistema de Informações sobre Mortalidade: um estudo para Maceió, Alagoas, Brasil, 2001-2002. Cad Saude Publica, 2007; 23(10):2385-2395.. Brazilian studies have shown low reliability of CD for fetal, perinatal and infant deaths¹⁵15. Barbeiro FMS, Fonseca SC, Tauffer MG, Ferreira MSS, Silva FP, Ventura PM, Quadros JI. Óbitos fetais no Brasil: revisão sistemática. Rev Saude Publica 2015; 49:22.^,1717. Almeida MF, Alencar GP, Schoeps D, Minuci EG, Silva ZP, Ortiz LP, Novaes HMD, Alencar AP, Raspatini PR, Santos PC. Qualidade das informações registradas nas declarações de óbito fetal em São Paulo, SP. Rev Saude Publica 2011; 45(5):845-53.^,1818. Pedrosa LCDO, Sarinho SW, Ordonha MR. Análise da qualidade da informação sobre causa básica de óbitos neonatais registrados no Sistema de Informações sobre Mortalidade: um estudo para Maceió, Alagoas, Brasil, 2001-2002. Cad Saude Publica, 2007; 23(10):2385-2395.^,3434. Santos SPC, Lansky S, Ishitani LH, França EB. Óbitos infantis evitáveis em Belo Horizonte: análise de concordância da causa básica, 2010-2011. Rev Bras Saúde Matern Infant 2015; 15(4):389-399.^,3535. Lansky S, França E, Ishitani L, Xavier CC. Confiabilidade da informação sobre óbito perinatal em Belo Horizonte, 1999: causas de óbito e variáveis selecionadas. Cad Saúde Colet 2010, 18(1):63-70..

Among the fetal deaths, the principal change of basic cause between the SIM and the research was from unspecified intrauterine hypoxia to unspecified cause of fetal death. This change has also been highlighted in similar studies of fetal¹⁵15. Barbeiro FMS, Fonseca SC, Tauffer MG, Ferreira MSS, Silva FP, Ventura PM, Quadros JI. Óbitos fetais no Brasil: revisão sistemática. Rev Saude Publica 2015; 49:22. and perinatal³⁵35. Lansky S, França E, Ishitani L, Xavier CC. Confiabilidade da informação sobre óbito perinatal em Belo Horizonte, 1999: causas de óbito e variáveis selecionadas. Cad Saúde Colet 2010, 18(1):63-70. deaths, and may reflect the incompleteness of the medical record for the fetal death, or non-realization of postmortem or lack of information in the diagnoses¹⁷17. Almeida MF, Alencar GP, Schoeps D, Minuci EG, Silva ZP, Ortiz LP, Novaes HMD, Alencar AP, Raspatini PR, Santos PC. Qualidade das informações registradas nas declarações de óbito fetal em São Paulo, SP. Rev Saude Publica 2011; 45(5):845-53.^,3535. Lansky S, França E, Ishitani L, Xavier CC. Confiabilidade da informação sobre óbito perinatal em Belo Horizonte, 1999: causas de óbito e variáveis selecionadas. Cad Saúde Colet 2010, 18(1):63-70.. The absence of clinical information about the pregnancy and the birth and non-inclusion of the placentas in the postmortem were obstacles to precision in the causes of death¹⁷17. Almeida MF, Alencar GP, Schoeps D, Minuci EG, Silva ZP, Ortiz LP, Novaes HMD, Alencar AP, Raspatini PR, Santos PC. Qualidade das informações registradas nas declarações de óbito fetal em São Paulo, SP. Rev Saude Publica 2011; 45(5):845-53..

Intrauterine hypoxia is the terminal stage of the causal chain of death, and thus should not be certified as the cause of death: unspecified causes of fetal death are the ‘unclear causes’ of the perinatal deaths. Both the diagnoses (P209 and P95) are non-specific in terms of leading to prevention measures³⁶36. Assis HM, Siviero PCL, Drumond EF, Machado CJ. Óbitos fetais sob o prisma da evitabilidade: análise preliminar de um estudo para o município de Belo Horizonte Cad Saúde Colet 2014, Rio de Janeiro, 22(3):314-317..

In spite of the small absolute number of fetal deaths, the frequencies of congenital syphilis (A50-A54) and diabetes (P70-P74) were, respectively, 2.5 and 5 times greater when compared to the information of the SIM. These causes of fetal death arise from maternal morbidities that are considered to be modifiable by better obstetric assistance and are frequent in countries of average and low income¹¹11. Lawn JE, Blencowe H, Waiswa P, Amouzo A, Mathers C, Hogan D, Flenady V, Frøen JF, Qureshi ZU, Calder-wood C, Shiekh S, Jassi FB, You D, McClure M, Matha Mi, Cousens S, for The Lancet Ending Preventable Stillbirths Series study group* with The Lancet Stillbirth Epidemiology investigator group. Stillbirths: rates, risk factors, and acceleration towards 2030. Lancet 2016; 18.. In Brazil, the prevalence of syphilis in pregnancy is greater in the absence of prenatal care, and even when this is carried out, the late start can be a barrier to diagnosis and treatment at the recommended fetal age³⁷37. Domingues RMSM, Szwarcwald CL, Souza Junior PRB, Leal MC. Prevalência de sífilis na gestação e testagem pré-natal: Estudo Nascer no Brasil. Rev Saude Publica 2014; 48(5):766-774.. The population fractions attributable to syphilis and pre-pregnancy diabetes in fetal death worldwide have been estimated at approximately 8% and 7%, respectively¹¹11. Lawn JE, Blencowe H, Waiswa P, Amouzo A, Mathers C, Hogan D, Flenady V, Frøen JF, Qureshi ZU, Calder-wood C, Shiekh S, Jassi FB, You D, McClure M, Matha Mi, Cousens S, for The Lancet Ending Preventable Stillbirths Series study group* with The Lancet Stillbirth Epidemiology investigator group. Stillbirths: rates, risk factors, and acceleration towards 2030. Lancet 2016; 18.. Congenital syphilis has also been under-registered as a basic cause of neonatal deaths, as shown in a study in the Brazilian state of Alagoas in the previous decade¹⁸18. Pedrosa LCDO, Sarinho SW, Ordonha MR. Análise da qualidade da informação sobre causa básica de óbitos neonatais registrados no Sistema de Informações sobre Mortalidade: um estudo para Maceió, Alagoas, Brasil, 2001-2002. Cad Saude Publica, 2007; 23(10):2385-2395.. The authors highlight the avoidability of neonatal deaths due to congenital syphilis and diabetes in pregnancy, among other causes.

The qualification of the CD of the fetal and neonatal deaths by the researchers can be evaluated also from the increase in the number of certified diagnoses. Sometimes the doctors omit to inform diagnoses, and also the CD, in the death certificate, because they are already recorded in the medical record³⁸38. Laurenti R. A análise da mortalidade por causas múltiplas. Rev Saude Publica 1974: 8:421-435.. Other factors, particularly for fetal deaths, are issues relating to ICD codes, scarce and incomplete information for the conditions of the placenta, and the model of the death certificate itself, which does not include information on complications in pregnancy and birth²⁷27. Froen JF, Gordijn SJ, Abdel-Aleem H, Bergsjo P, Betran A, Duke CW, Fauveau V, Flenady V, Hinderaker SG, Hofmeyr GJ, Jokhio AH, Lawn J, Lumbiganon P, Merialdi M, Pattinson R, Shankar A. Making stillbirths count, making numbers talk - Issues in data collection for stillbirths. BMC Pregnancy and Childbirth 2009; 9:58..

A limitation of the survey was that (i) there is no Death Verification Service in the cities of Rio de Janeiro State, and (ii) in the city of São Paulo which does have the service, some results were imprecise for qualifying cause of death, as signaled by Almeida et al.¹⁷17. Almeida MF, Alencar GP, Schoeps D, Minuci EG, Silva ZP, Ortiz LP, Novaes HMD, Alencar AP, Raspatini PR, Santos PC. Qualidade das informações registradas nas declarações de óbito fetal em São Paulo, SP. Rev Saude Publica 2011; 45(5):845-53. Additionally, the strategy of covering a larger sample of newborns by choosing maternity facilities with higher numbers of births in the SUS in the cities investigated by the survey, similar to the sampling strategy used in the World Health Organization's multicentric Maternal and Newborn Health survey³⁹39. Souza JP, Cecatti JG, Faundes A, Morais SS, Villar J, Carroli G, Gulmezoglu M, Wojdyla D, Zavaleta N, Donner A, Velazco A, Bataglia V, Valladares E, Kublickas M, Acosta A; WHO 2005 global survey on maternal and perinatal health research group. Maternal near miss and maternal death in the 2005 WHO global survey on maternal and perinatal health. Bull World Health Organ 2010; 88(2):113-119., meant that the sample was not random.

As positive points of the study it can be pointed out that there is practically no error of classification for the fetal losses (abortion and fetal deaths), and for the fetal deaths with live births, the latter being already signaled in a publication with the database of the survey in the city of São Paulo⁴⁰40. Laurenti R, Mello Jorge MHP, Gotlieb SLD, Oliveira BC, Pimentel EC. O estudo do binômio mãe-filho: descrição e resultados gerais. Rev Bras Epidemiol 2015; 18(2):398-412..

To achieve better filling out of the death certificate, and consequent precision on causes of death, there is a need for the attention of doctors and principally the medical students to be drawn to the individual definitions of the deaths involved (maternal, abortion, fetal component of the perinatal death, and infant death, in their various components) and the inclusion of all illnesses in the death certificates³⁸38. Laurenti R. A análise da mortalidade por causas múltiplas. Rev Saude Publica 1974: 8:421-435., those of the newborn and those consequent upon maternal morbidities. This would have a beneficial effect on the quality of care.

Good-quality care in reproductive planning, pre-natal care, and care during the birth will result in prevention of abortions, fetal deaths, maternal deaths and neonatal deaths¹¹11. Lawn JE, Blencowe H, Waiswa P, Amouzo A, Mathers C, Hogan D, Flenady V, Frøen JF, Qureshi ZU, Calder-wood C, Shiekh S, Jassi FB, You D, McClure M, Matha Mi, Cousens S, for The Lancet Ending Preventable Stillbirths Series study group* with The Lancet Stillbirth Epidemiology investigator group. Stillbirths: rates, risk factors, and acceleration towards 2030. Lancet 2016; 18..

References

Publication Dates

History