Trend of infant mortality in cities of metropolitan region of Porto Alegre, Brazil from 1996 to 2021

Abstract

The aim is to verify the trend in infant mortality (IM) and preventable infant mortality (PIM) in municipalities in the metropolitan region of Porto Alegre from 1996 to 2021 and their associations with per capita public spending on health and population coverage of the Family Health Strategy (FHS). Ecological study with data from information systems: on mortality, live births, public budgets and primary health care. The deaths described in the Ministry of Health's list were considered preventable. Statistical analysis used Prais-Winsten regression. A decrease in MIE was observed in the municipalities studied with the exception of Esteio. There was an association between per capita public spending on health and the study outcomes in Novo Hamburgo, Canoas and Porto Alegre. An association was found between the FHS coverage and IM in Novo Hamburgo, Sapucaia do Sul, Canoas and Porto Alegre. PIM was associated with FHS coverage in Novo Hamburgo, Canoas and Porto Alegre. Despite the decrease in the IM and PIM coefficients, the association of this indicator was more impacted by per capita public spending on health than by the FHS population coverage.

Keywords:
Infant Mortality; Causes of death; Healthcare Financing; Public Expenditures on Health; Family Health Strategy

Resumo

O objetivo é verificar a tendência da mortalidade infantil (MI) e da mortalidade infantil evitável (MIE) em municípios da região metropolitana de Porto Alegre de 1996 a 2021 e suas associações com o gasto público per capita em saúde e a cobertura populacional de Estratégia Saúde da Família (ESF). Estudo ecológico com dados dos sistemas de informações: sobre mortalidade, nascidos vivos, de orçamentos públicos e de atenção básica. Foram consideradas evitáveis as mortes descritas na lista do Ministério da Saúde. A análise estatística utilizou a regressão de Prais-Winsten. Foi observada diminuição da MIE nos municípios estudados com exceção de Esteio. Houve associação entre o gasto público per capita em saúde e os desfechos do estudo em Novo Hamburgo, Canoas e Porto Alegre. Encontrou-se associação entre a cobertura de ESF e a MI em Novo Hamburgo, Sapucaia do Sul, Canoas e Porto Alegre. A MIE esteve associada à cobertura de ESF em Novo Hamburgo, Canoas e Porto Alegre. Apesar da diminuição dos coeficientes de MI e de MIE, a associação deste indicador foi mais impactada pelo gasto público per capita em saúde do que pela cobertura populacional de ESF.

Palavras-chave:
Mortalidade Infantil; Causas de Morte; Financiamento da Assistência à Saúde; Gastos Públicos com Saúde; Estratégia Saúde da Família

Resumen

El objetivo es verificar la tendencia de la mortalidad infantil (MI) y la mortalidad infantil prevenible (MIP) en municipios de la región metropolitana de Porto Alegre de 1996 a 2021 y sus asociaciones con el gasto público per cápita en salud y la cobertura poblacional de la Estrategia Brasileña de Salud de la Familia (ESF). Estudio ecológico con datos de sistemas de información: sobre mortalidad, nacidos vivos, presupuestos públicos y cuidados básicos. Las defunciones descritas en la lista del Ministerio de Salud se consideraron evitables. El análisis estadístico utilizó la regresión de Prais-Winsten. Se observó una disminución del MIP en los municipios estudiados a excepción de Esteio. Hubo una asociación entre el gasto público per cápita en salud y los resultados del estudio en Novo Hamburgo, Canoas y Porto Alegre. Se encontró una asociación entre la cobertura de la ESF y la MI en Novo Hamburgo, Sapucaia do Sul, Canoas y Porto Alegre. La MIP estuvo asociado a la cobertura de la ESF en Novo Hamburgo, Canoas y Porto Alegre. A pesar de la disminución de los coeficientes MI y MIP, la asociación de este indicador se vio más impactada por el gasto público per cápita en salud que por la cobertura poblacional de la ESF.

Palavras clave:
Mortalidad Infantil; Causas de Muerte; Financiamiento de la Atención Médica; Gasto en Salud Pública; Estrategia de Salud Familiar

Introduction

Despite the reduction in infant mortality (IM) observed in Brazil. national levels are still higher than those identified in high-income countries^1-2. It is well-known that IM can be influenced by numerous factors. such as socioeconomic conditions^3-4. characteristics of the newborn and the mother. and availability of health services^3-5. including childbirth care⁶. but evidence has been consolidated regarding the determining role of public policy financing^7-8-9.

There is a clear insufficiency of public spending on health in Brazil when compared to developed countries. Some international evidence has shown that countries with universal systems have presented public spending on health that reaches at least 70% of total spending. In Brazil. public spending on health was 47% of the total. lower than the previous 53% that private spending on health had reached. Furthermore. This percentage has typically been much lower when compared with other countries. such as Canada. 71.1%; the United Kingdom. 83.2%; and Norway. 85.5%^10-11-12.

As a way of dealing with the economic crisis. The government opted to contain public spending. Constitutional amendment No. 95 of 2016 froze the federal budget. including health spending. for 20 years.¹³ According to Donieke and collaborators¹⁴. “In 2017, the government had already failed to meet the minimum investment in the health budget guaranteed by the Constitution. around R$692 million.” These restrictions created expectations of direct consequences for health indicators. causing an increase in IM¹⁵. One study analyzing the economic slowdown and IM showed that in middle- and low-income countries. Even after controlling demographic and infrastructure variables. A 1% reduction in the Gross Domestic Product (GDP) per capita resulted in an increase in IM¹⁶. In 2008, the Health Surveillance Secretariat of the Ministry of Health released a list of preventable deaths through interventions by the Unified Health System. with comprehensive objectives that include reducing IM in Brazil. The list of preventable deaths was updated in 2010 and were considered to be those that would not occur with the provision of effective care at the appropriate time and place¹⁷. highlighting the importance of primary health care (PHC) as the gateway and organizer of the healthcare network.

Thus, the Family Health Strategy (FHS) became the predominant model of PHC in Brazil. Important findings from Macinko et al.¹⁸ have already revealed the impact of the FHS in reducing IM. After, several studies also highlighted the impact of the FHS on IM and preventable infant mortality (PIM)^19-22.

Thus, the present study verified the trends of IM and PIM in municipalities in the metropolitan region of Porto Alegre from 1996 to 2021 and their associations with per capita public spending on health and population coverage of FHS in order to subsidize the health departments and their respective municipal health councils. with the aim of detecting inequities in the determinants of the health-disease process.

Methods

An ecological study was carried out to verify trends of the IM and PIM coefficients in the main municipalities of the metropolitan region, which are part of the Trensurb line (Novo Hamburgo, São Leopoldo, Sapucaia do Sul, Esteio, Canoas and Porto Alegre), in the period between 1996 and 2021.

The number of deaths was available in the Mortality Information System (Sistema de Informações sobre Mortalidade - SIM) on the website of the Department of Information Technology of the Unified Health System (Departamento de Informática do Sistema Único de Saúde - DATASUS). The population data on live births for each municipality in all the years of this study were obtained from the Live Birth Information System (Sistema de Informações sobre Nascidos Vivos - SINASC). also available on DATASUS. Information on public health spending was collected from the Public Health Budget Information System (Sistema de Informações sobre Orçamentos Públicos em Saúde - SIOPS) and the population coverage of the Family Health Strategy was obtained from the website of the Department of Primary Care (Departamento de Atenção Básica - DAB). The number of deaths in children under one year of age was collected in each municipality from 1996 (first year of implementation of the 10th Revision of the International Classification of Diseases) to 2021 (last year available in SIM).

Preventable deaths in children under one year of age were classified as: reducible by immunoprevention actions; by adequate attention to women’s health during pregnancy. childbirth. the fetus. and the newborn; by resolute actions to promote health or linked to these; and by ill-defined causes of death¹⁷. Infant mortality rates were calculated using the formula: [(number of deaths in children under one year of age with mothers residing in the municipality in the year/number of live births with mothers residing in the municipality in the year) x 1000].

PIM rates were calculated using the formula: [(number of preventable deaths in children under one year of age with mothers residing in the municipality in the year/number of live births with mothers residing in the municipality in the year) x 1000].

Proportional mortality rates in the neonatal period (from birth to 27 days) and preventable neonatal mortality (PNM) were calculated among infant deaths for each year of the historical series in the municipalities using the following equations: (number of neonatal deaths x 100 / number of infant deaths) and (number of preventable neonatal deaths x 100 / number of preventable infant deaths).

Preventable deaths in children under one year of age were classified as: reducible by immunoprevention actions; by adequate attention to women’s health during pregnancy. childbirth. the fetus. and the newborn; by resolute actions to promote health or linked to these; and by ill-defined causes of death¹⁷. Infant mortality rates were calculated using the formula: [(number of deaths in children under one year of age with mothers residing in the municipality in the year/number of live births with mothers residing in the municipality in the year) x 1000].

PIM rates were calculated using the formula: [(number of preventable deaths in children under one year of age with mothers residing in the municipality in the year/number of live births with mothers residing in the municipality in the year) x 1000].

Proportional mortality rates in the neonatal period (from birth to 27 days) and preventable neonatal mortality (PNM) were calculated among infant deaths for each year of the historical series in the municipalities using the following equations: (number of neonatal deaths x 100 / number of infant deaths) and (number of preventable neonatal deaths x 100 / number of preventable infant deaths).

Public spending on health was determined through the Summary Reports of Budget Execution of each municipality available in SIOPS, from 2000 to 2020. The total amount of resources applied by the municipalities were determined based on the expenditure settled in each year of the direct administration subfunction in health in SIOPS. Per capita public spending on health was prepared as the ratio of the total spending made in each municipality in the year and the population of the municipalities according to the year.

To reduce price variations due to inflation on the values presented. the deflator resource was used. The values of all years of the historical series of per capita public spending on health were brought to the present day. through correction by the National Broad Consumer Price Index (IPCA) of 2020, the last year of the series^23-24.

The population coverage of FHS between 1998 and 2020 was obtained from the website of the DAB and presented as a percentage²⁵. The annual average was calculated for each municipality.

The trends in the IM and PIM coefficients. per capita health expenditure and population coverage of FHS in the municipalities were verified. The associations between mortality coefficients, per capita public health expenditure. and population coverage of FHS in each municipality were analyzed.

The data were tabulated using Microsoft Excel 2010® spreadsheets. in which information was displayed by municipality and year to construct the indicators. The Stata 11 program was used for statistical analysis.

The temporal trend was analyzed based on the Prais-Winsten Generalized Regression models. indicated for a time series of more than seven years, as it corrects for first-order serial autocorrelation. a common occurrence in a time series. which overestimates adjustment measures²⁶.

The results with a p-value<0.05 were statistically significant. with the trend considered to be increasing when the regression coefficient was positive and decreasing when the regression coefficient was negative. The trend with a p-value > 0.05 was considered stationary or insignificant. In the association analysis, the IM coefficients and the PIM coefficients were the dependent variables and per capita public spending on health and population coverage of FHS were the independent variables.

This study was carried out with secondary data available in public information systems. protecting the identification of individuals. This study was exempt from the requirement for approval by a research ethics board in accordance with Resolution 466/2012 of the National Health Council.

Results

Between 1996 and 2021, the IM ranged from 21.9 in Canoas (1996) to 1.7 deaths per thousand live births in Esteio (2011). With the exception of the municipality of Esteio, all other municipalities had average coefficients between 11 and 13 deaths per thousand live births. During the period studied, neonatal mortality predominated in all municipalities, with average percentages above 60% (Table 1).

Table 1
Description of infant mortality rates and neonatal mortality rates in municipalities in the metropolitan region of Porto Alegre from 1996 to 2021.

Year	Novo Hamburgo		São Leopoldo		Sapucaia do Sul		Esteio		Canoas		Porto Alegre
	IM	NM (%)	IM	NM (%)	IM	NM (%)	IM	NM (%)	IM	NM (%)	IM	NM (%)
1996	19.6	67.5	18.3	60.8	16.9	60.5	9.9	72.2	21.9	55.5	18.5	52.6
1997	15.9	65.2	11.8	59.6	13.5	78.8	15.0	68.2	16.8	63.3	15.7	59.1
1998	13.4	53.7	13.9	60.4	14.2	57.1	11.0	58.8	14.9	43.9	16.4	51.6
1999	14.4	64.5	13.7	64.9	15.3	70.0	10.9	64.7	14.0	69.1	12.2	58.2
2000	13.5	65.5	15.0	79.7	13.8	55.9	12.4	66.7	15.4	61.9	14.8	56.5
2001	13.8	57.1	16.2	57.1	8.3	38.9	14.7	60.9	13.6	66.7	14.2	60.3
2002	12.8	63.6	13.3	59.6	14.2	65.5	13.3	55.0	14.5	56.9	14.0	51.4
2003	15.0	56.9	13.1	54.8	12.2	50.0	8.6	66.7	15.4	57.9	13.3	55.9
2004	17.2	70.7	10.2	68.6	9.5	73.7	16.0	45.8	12.7	55.2	12.2	61.1
2005	12.2	78.1	8.5	53.6	12.1	73.9	10.7	68.8	16.0	62.0	12.9	56.2
2006	10.4	41.2	13.6	66.7	15.8	61.5	7.7	50.0	13.3	62.1	12.1	58.7
2007	13.5	57.1	11.4	50.0	12.4	73.7	7.2	63.6	11.8	63.2	11.9	53.3
2008	17.1	64.0	11.9	70.3	14.4	52.4	5.4	75.0	11.5	71.4	11.6	65.3
2009	8.9	76.9	14.9	67.4	11.7	68.4	9.6	75.0	12.7	58.3	9.8	61.2
2010	14.2	73.8	11.2	66.7	11.2	66.7	6.2	42.9	8.3	80.0	10.5	64.6
2011	10.6	76.5	10.3	77.4	7.4	92.3	1.7	100.0	13.0	59.4	9.3	63.4
2012	11.9	71.1	15.6	77.8	7.4	78.6	13.9	58.8	7.2	64.7	9.2	62.2
2013	10.7	58.8	10.0	77.4	11.5	72.7	9.8	81.8	9.6	70.6	9.3	62.3
2014	9.4	76.7	12.4	63.2	8.6	76.5	14.0	80.0	10.3	65.5	9.8	67.4
2015	11.7	65.9	14.8	80.4	9.2	52.6	7.3	88.9	11.0	66.7	9.2	69.2
2016	9.2	82.8	11.5	75.8	11.0	76.2	13.8	73.3	10.2	52.8	9.0	64.3
2017	12.8	78.1	11.1	66.7	10.2	78.9	11.8	92.3	8.6	53.5	9.0	68.1
2018	8.3	69.2	12.8	75.7	9.7	52.9	6.2	85.7	9.8	68.8	9.0	66.7
2019	8.8	81.5	11.4	67.7	11.6	88.9	8.5	77.8	9.9	70.8	8.7	64.6
2020	9.2	74.1	6.9	94.7	8.60	92.9	2.0	100.0	8.1	83.3	7.8	69.7
2021	8.1	77.3	13.4	83.3	10.9	62.5	7.8	85.7	7.0	73.3	9.2	67.7
Average	12.4	68.0	12.6	68.5	11.6	68.2	9.8	71.5	12.2	63.8	11.5	61.2

• IM: Infant mortality rate per 1.000 inhabitants; NM: neonatal mortality rate.

In relation to PIM, in 1996, Canoas had the highest coefficient of the period. with 16.3 deaths per thousand live births. while in 2020. Esteio had the lowest value. with only one death per thousand live births (Table 2). The descriptive analysis showed that in all municipalities the average coefficients in the period were below nine deaths per thousand live births. except in São Leopoldo. In the historical series. most of the PIM was concentrated in the neonatal component. with average percentages above 70% in most municipalities, except Porto Alegre and Canoas (Table 2).

Table 2
Description of avoidable infant mortality rates and neonatal mortality rates in municipalities in the metropolitan region of Porto Alegre from 1996 to 2021.

Year	Novo Hamburgo		São Leopoldo		Sapucaia do Sul		Esteio		Canoas		Porto Alegre
	PIM	PNM (%)	PIM	PNM (%)	PIM	PNM (%)	PIM	PNM (%)	PIM	PNM (%)	PIM	PNM (%)
1996	15.1	70.3	14.8	63.3	13.8	61.3	8.2	80.0	16.3	61.1	14.0	52.8
1997	7.5	72.6	8.5	67.7	10.2	80.0	10.9	68.6	13.4	67.9	12.4	61.4
1998	8.7	54.3	11.3	65.1	10.6	61.5	9.1	57.1	10.3	47.4	11.6	50.9
1999	9.0	69.2	8.7	75.0	11.4	73.3	7.0	72.7	10.0	71.7	9.1	62.4
2000	9.3	68.4	9.2	80.6	8.9	68.2	4.8	71.4	9.9	67.8	10.2	58.9
2001	7.9	67.9	11.0	63.2	6.0	38.5	10.2	75.0	9.2	68.6	10.1	62.6
2002	8.7	76.7	9.3	63.6	7.8	81.3	9.3	57.1	8.5	71.7	9.3	54.3
2003	10.0	61.8	7.2	69.6	9.2	44.4	7.8	72.7	11.0	64.8	8.4	59.9
2004	12.2	78.1	5.2	55.6	6.0	83.3	11.4	47.1	8.9	61.7	8.6	60.5
2005	7.2	91.7	6.1	65.0	6.8	76.9	8.0	75.0	10.7	64.8	8.8	60.8
2006	6.4	47.6	8.4	60.7	10.9	50.0	2.6	100.0	7.7	63.2	8.3	64.5
2007	9.0	60.7	9.8	44.8	8.5	69.2	3.3	60.0	8.5	70.7	7.8	56.1
2008	11.6	70.6	9.0	75.0	13.0	57.9	2.0	100.0	7.8	78.9	7.8	70.3
2009	4.1	91.7	11.0	61.8	8.0	84.6	7.2	77.8	8.2	69.2	7.0	67.2
2010	9.8	82.8	8.1	75.0	8.7	78.6	3.5	75.0	5.0	83.3	6.3	73.9
2011	6.8	90.9	7.6	82.6	5.1	100.0	1.7	100.0	9.8	66.7	5.8	69.7
2012	8.1	80.8	12.5	83.3	5.8	81.8	9.0	54.5	4.8	64.0	6.4	65.3
2013	6.0	73.7	5.2	81.3	8.4	75.0	8.0	88.9	6.2	87.9	6.3	68.0
2014	5.9	78.9	5.9	66.7	7.5	73.3	9.3	100.0	7.1	65.8	6.4	79.5
2015	6.9	66.7	9.7	83.3	5.8	66.7	6.5	87.5	7.3	70.0	5.8	79.1
2016	6.3	90.0	9.4	81.5	6.3	91.7	12.0	76.9	6.7	65.7	5.5	69.9
2017	9.3	83.3	7.7	69.6	6.6	75.0	9.0	100.0	4.4	63.6	6.2	74.6
2018	4.1	92.3	8.0	78.3	4.6	25.0	3.5	75.0	7.3	75.0	4.9	77.9
2019	5.5	76.5	8.5	78.3	10.3	93.8	5.7	83.3	6.0	75.9	5.9	70.1
2020	6.2	83.3	5.4	100.0	6.1	100.0	1.0	100.0	4.7	81.0	4.8	73.3
2021	4.0	72.7	10.0	85.2	6.1	77.8	4.5	100.0	4.9	76.2	6.0	78.8
Média	8.3	74.7	9.0	72.2	8.4	71.9	6.9	78.9	8.4	69.3	7.9	66.3

• PIM: Infant Mortality Rate due to preventable causes per 1.000 inhabitants; PNM: percentage of preventable neonatal mortality.

It is important to note that in 2021, the most significant year of the pandemic in Brazil. São Leopoldo was the only municipality that presented IM and PIM coefficients of higher than the average for the period.

Regarding the trend analysis. between 1996 and 2021, there was a statistically significant decrease in IM in all municipalities (Table 3). Regarding the PIM trend, a significant drop in the coefficients was observed in all municipalities. except in Esteio. which showed stability (Table 3).

Table 3
Trends in infant mortality and preventable infant mortality rates in municipalities in the metropolitan region of Porto Alegre from 1996 to 2021.

	Infant Mortality				Preventable Infant Mortality
	b	95% CI		p-valor	b	95% CI		p-valor
Novo Hamburgo	-0.291	-0.376	-0.207	<0.001	-0.253	-0.342	-0.163	<0.001
São Leopoldo	-0.133	-0237	-0.029	0.014	-0.116	-0.225	-0.006	0.040
Sapucaia do Sul	-0.209	-0.327	-0.091	0.001	-0.186	-0.305	-0.067	0.004
Esteio	-0.204	-0.393	-0.015	0.035	-0.153	-0.363	0.057	0.147
Canoas	-0.380	-0.468	-0.292	<0.001	-0.309	-0.384	-0.234	<0.001
Porto Alegre	-0.338	-0.400	-0.276	<0.001	-0.293	-0.363	-0.224	<0.001

• b: regression coefficient; 95% CI: 95% confidence interval; p-value for Prais-Winsten regression.

All municipalities showed a significant growth trend in relation to per capita public spending on health. except Sapucaia do Sul. whose values in the area were stable (Table 4). Between 1998 and 2020, there was a significant increase in the population coverage of FHS in all of the municipalities analyzed in this study (Table 4).

Table 4
Trends in per capita public spending on health and population coverage of the Family Health Strategy in municipalities in the metropolitan region of Porto Alegre.

	Per capita public spending on health*				Population coverage of FHS**
	b	95% CI		p-valor	b	95% CI		p-valor
Novo Hamburgo	48.357	28.940	67.775	<0.001	2.781	1.095	4.467	0.003
São Leopoldo	37.803	17.422	58.183	0.001	0.908	0.582	1.234	<0.001
Sapucaia do Sul	21.491	-8.438	51.419	0.149	2.690	1.979	3.401	<0.001
Esteio	70.634	35.446	105.823	<0.001	2.690	1.431	3.948	<0.001
Canoas	74.388	55.463	93.312	<0.001	2.695	1.350	4.040	<0.001
Porto Alegre	20.657	3.654	37.660	0.020	2.193	1.670	2.716	<0.001

• * values in Reais from 2000 to 2020 corrected by the IPCA; ** average of annual percentages from 1998 to 2020. b: regression coefficient; 95% CI: 95% confidence interval; p-value for Prais-Winsten regression.

The highest and lowest values of per capita public spending on health in the period were found in Esteio, with R$1.638.48 in 2020 and R$87.49 in 1996, respectively. Regarding the average amounts invested in health in the municipalities, only Porto Alegre reached levels above R$1.000 per inhabitant (Supplementary Table 1).

In 1998, only Esteio and Porto Alegre had some form of FHS coverage. both with percentages below 5% of the population. Throughout the period, the highest FHS coverage was found in 2018, in the municipality of Canoas, with 63.4%. In 2020, the last year of the historical series. two municipalities had FHS coverage between 60% and 65% and São Leopoldo had a coverage percentage below 20% (Supplementary Table 2). The median percentages of population coverage by FHS were analyzed, with emphasis on Porto Alegre and Sapucaia do Sul. which obtained percentages above 20% in 2010 and 2011.

A significant and inverse association was found between both IM and PIM and per capita public spending on health in the municipalities of Novo Hamburgo. Canoas and Porto Alegre, that is, an increase of one real per capita per year in the aforementioned municipalities led to a decrease from four to seven infant deaths per thousand live births and from three to four preventable infant deaths per thousand live births (Table 5).

Table 5
Association between infant mortality rates and preventable infant mortality with per capita public spending on health and the Family Health Strategy in municipalities in the metropolitan region of Porto Alegre from 2000 to 2020.

	Per capita public spending on Health
	Infant Mortality				Preventable Infant Mortality
	b	95% CI		p-value	b	95% CI		p-value
Novo Hamburgo	-4.60	-7.45	-1.75	0.004	-3.04	-5.76	-0.32	0.031
São Leopoldo	1.02	-2.65	4.69	0.560	1.39	-3.52	6.31	0.553
Sapucaia do Sul	0.20	-7.76	8.16	0.957	-0.34	-8.66	1.88	0.189
Esteio	-0.60	-9.90	8.70	0.893	1.49	-7.12	10.10	0.717
Canoas	-4.59	-6.61	-2.58	<0.001	-3.13	-4.46	-1.79	<0.001
Porto Alegre	-7.61	-12.07	-3.14	0.003	-4.61	-8.12	-1.10	0.014
	Family Health Strategy
	Infant Mortality				Preventable Infant Mortality
	b	95% CI		p-value	b	95% CI		p-value
Novo Hamburgo	-0.063	-0.091	-0.035	<0.001	-0.047	-0.073	-0.020	0.001
São Leopoldo	-0.061	-0.210	0.088	0.406	-0.041	-0.179	0.097	0.541
Sapucaia do Sul	-0.059	-0.108	-0.010	0.021	-0.041	-0.088	0.006	0.086
Esteio	-0.049	-0.136	0.038	0.253	-0.029	-0.115	0.057	0.488
Canoas	-0.074	-0.118	-0.030	0.002	-0.060	-0.089	-0.031	<0.001
Porto Alegre	-0.129	-0.174	-0.084	<0.001	-0.109	-0.144	-0.073	<0.001

• b: regression coefficient; 95% CI: 95% confidence interval; p-value for Prais-Winsten regression.

Similarly, an association between outcomes and FHS population coverage was observed. The IM coefficients showed a significant association with the average percentage of FHS population coverage in the municipalities of Novo Hamburgo, Sapucaia do Sul, Canoas and Porto Alegre. The PIM coefficients were associated with the average percentage of FHS coverage in Novo Hamburgo, Canoas and Porto Alegre only. For each 1% increase in FHS population coverage, the decrease in outcomes was less than one death per thousand live births (Table 5).

Discussion

This study showed a decrease in IM and PIM in all analyzed municipalities. It was observed that the persistence of IM was proportionally more concentrated in the neonatal period. The analysis showed that per capita public spending on health was more relevant in reducing mortality than in FHS coverage.

The reduction in mortality rates followed the trends found in other studies conducted in Brazil. A publication by the Ministry of Health showed a drop in IM rates between 1990 and 2019. The average for the final three-year period (2017-19) in Brazil was 13.3 deaths for every thousand live births, while in the South Region it reached 10.1 deaths for every thousand live births².

Other studies have shown a decreasing trend. In Rio Branco. from 1999 to 2015, there was a decrease in IM coefficients, with values ​​falling from 27.0 to 14.5 deaths per thousand live births²⁷.

In Aracajú. between 2001 and 2010. IM coefficients decreased from 29.5 to 17.7 deaths per thousand live births²⁸. Consequently, it can be seen that the coefficients observed in our analysis were lower than those cited, but they were still above those verified in Florianópolis, 5.34 per thousand live births in 2016, considered to be the Brazilian capital with the lowest IM²⁹.

In 2020 and 2021, no major changes were observed in mortality rates, except in São Leopoldo. An ecological study covering all of Brazil revealed a small impact of the pandemic on IM³⁰.

As for PIM, Esteio was the only municipality in which the trend did not decrease significantly. However, this municipality had the lowest IM rates in the period, a fact that may have contributed to the lack of statistical association in this outcome. Studies conducted in other locations throughout Brazil also found a decrease in PIM. In Minas Gerais, a reduction in PIM was found, although it was more severe in highly vulnerable areas of the Jequitinhonha Valley³¹. A time series study in eastern Minas Gerais also showed a decrease in IM, showing that 68% of all deaths among children under one year of age were preventable by actions linked to the SUS³². A time series study in Mato Grosso showed a decline in PIM between 2007 and 2020 in children under one year of age³³.

The results also showed the predominance of neonatal mortality, both in IM and PIM. This result points to the need to improve prenatal care, both in primary care and in support for high-risk situations, at the time of delivery, which specifically requires the provision and qualification of hospital services^29.-34-35-36.

This finding is in line with other studies carried out in the country. The highest IM coefficients were also concentrated in the early neonatal period, reaching an average proportion of 52%²⁷.

Regarding neonatal mortality in Brazil, an average rate of 9.46 per thousand live births was found from 2007 to 2017, with a reduction of 2.2% per year, with a greater decline in early neonatal mortality when compared to late mortality³⁷. PNM in Brazil, between 2000 and 2018, decreased from 10.98 per thousand live births in 2000 to 6.76 in 2018³⁸.

An increase in per capita public spending on health was observed in all municipalities. The World Health Organization (WHO) has recommended that universal health coverage require a minimum expenditure of US$112 per capita and 7.5% of GDP applied to health³⁹, a figure exceeded by all municipalities in the last year of the analysis. Nevertheless, studies on health expenditures are still limited in Brazil. One study conducted in São Paulo, including seven municipalities of Rota dos Bandeirantes, between 2009 and 2012, showed per capita public spending on health ranging from R$291 to R$1.354, with an average of R$624, that is, values ​​at the time similar to those achieved in the present study⁴⁰.

The increase observed in the analyzed period, as compared to more than 20 years ago, can be explained by the numerous changes in the financing of SUS, such as the expansion of management levels, the increase in the complexity of health actions under municipal responsibility and, primarily. the constitutional definition of sectoral spending by federated entities. A study carried out between 2000 and 2010 had already shown an 81.4% growth in per capita public spending on health in the Southern region of Brazil, showing, at the time, a restriction in federal spending but an increase in contributions made by states (from 16% to 22%) and municipalities (from 25% to 36%)⁴¹. Although it should be noted that per capita spending on public health actions and services by the federal government in 2019 was R$560 (13.8% of the federal government’s net current revenue), this amount was the lowest invested since 2014, when it was R$600, corresponding to 14.3% of net current revenue⁴². What is clear is the lack of definition of criteria and restrictions on federal spending on health throughout the period.

The present study pointed out the relationship between IM and PIM and the increase in per capita public spending on health in at least three municipalities. This relationship has been shown by other studies^15-43-44-45.

The FHS coverage found in this study was lower than necessary. The emblematic study conducted by Aquino et al.⁴⁶ considered that effective FHS coverage should reach at least 70% of the population within four years of implementation. No municipality included in the study achieved this coverage. Although FHS coverage is associated with a reduction in IM, the analysis demonstrated a low impact on this drop in indicators. The analysis of the association showed that the reduction in mortality was more impacted by public spending on health than by FHS population coverage. Ecological studies have a weak design to establish causal inferences, and it is worth noting that the data collection was carried out in different time periods. The present study did not analyze aspects that are known to determine IM, such as the structure. Organization, and quality of health services; socioeconomic factors; and individual characteristics of the children. However, the objective of the study was to monitor mortality trends in relation to public spending and FHS population coverage. The analytical treatment given to per capita public spending on health followed the rigor of economic studies, which was updated through the appropriate index, and FHS coverage considered a long observation period.

Changes in IN and PIM were observed in relation to per capita public spending on health in the years investigated in this study, even under a scenario of cost containment since 2016. It is important to note that, despite the positive results observed in the municipalities, the increase in health costs and the aging of the population require that financial contributions be adjusted, accompanying these changes for more promising results.

In this context, it is important to highlight the change in the PHC model with the strengthening of the FHS through the expansion and consolidation of coverage that should be priorities in management planning. Adequate prenatal care for pregnant women. qualified care at the time of birth, monitoring of newborns, and care during the postpartum period are essential to reduce mortality among children under one year of age.

Acknowledgements

To the Coordination for the Improvement of Higher Education Personnel (CAPES). for the doctoral scholarship awarded to Douglas Nunes Stahnke.

References

• ¹ Szwarcwald CL, Almeida WS, Teixeira RA, França EB, Miranda MJ, Malta DC. Inequalities in infant mortality in Brazil at subnational levels in Brazil, 1990 to 2015. Popul Health Metrics 2020; 18(1):4. Doi: 10.1186/s12963-020-00208-1
  » https://doi.org/10.1186/s12963-020-00208-1
• ² Brasil. Ministério da Saúde (MS). Secretaria de Vigilância em Saúde. Mortalidade infantil no Brasil - Boletim Epidemiológico [Internet] 2021 [acessado 2022 Jan. 3]. Disponível em: https://www.gov.br/saude/pt-br/centrais-de-conteudo/publicacoes/boletins/epidemiologicos/edicoes/2021/boletim_epidemiologico_svs_37_v2.pdf
  » https://www.gov.br/saude/pt-br/centrais-de-conteudo/publicacoes/boletins/epidemiologicos/edicoes/2021/boletim_epidemiologico_svs_37_v2.pdf
• ³ Lima JC, Mingarelli AM, Neuber JS, Zavala AAZ, Takano AO. Estudo de base populacional sobre mortalidade infantil. Cien Saude Colet 2017; 22(3):931-939. Doi: 10.1590/1413-81232017223.12742016.
  » https://doi.org/10.1590/1413-81232017223.12742016
• ⁴ Tejada Cao, Triaca LM, Liermann NH, Ewerling F, Costa JC. Crises econômicas, mortalidade de crianças e o papel protetor do gasto público em saúde. Cien Saude Colet 2019; 24(12): 4395-4404. Doi: 10.1590/1413-812320182412.25082019
  » https://doi.org/10.1590/1413-812320182412.25082019
• ⁵ Martins JLA, Durans KCN, Brito JD, Freitas DS. Mortalidade infantil por causas evitáveis de crianças de 0-4 anos no Maranhão entre 2015 a 2019. Research, Society and Development 2022;11(7). Doi: 10.33448/rsd-v11i7.29952
  » https://doi.org/10.33448/rsd-v11i7.29952
• ⁶ Varela AR, Schneider BC, Bubach S, Silveira MF, Bertoldi AD, Duarte LSM, Menezes AMB, Domingues MR, Bassani DG. Fetal, neonatal, and post-neonatal mortality in the 2015 Pelotas (Brazil) birth cohort and associated factors. Cad Saude Publica 2019; 35(7). Doi: 10.1590/0102-311X00072918.
  » https://doi.org/10.1590/0102-311X00072918.
• ⁷ Ensor T, Cooper S, Davidson L, Fitzmaurice A, Graham WJ. The impact of economic recession on maternal and infant mortality: lessons from history. BMC Public Health 2010; 10(727):1-9. Doi: 10.1186/1471-2458-10-727.
  » https://doi.org/10.1186/1471-2458-10-727
• ⁸ Williams C, Gilbert BJ, Zeltner T, Watkins J, Atun R, Maruthappu M. Effects of economic crises on population health outcomes in Latin America, 1981-2010: an ecological study. BMJ Open 2016; 6 (1). Doi: 10.1136/bmjopen-2014-007546.
  » https://doi.org/10.1136/bmjopen-2014-007546
• ⁹ Maruthappu M, Watson RA, Watkins J, Zeltner T, Raine R, Atun, R. Effects of economic downturns on child mortality: a global economic analysis, 1981-2010. BMJ Global Health 2007; 2 (2). Doi: 10.1136/bmjgh-2016-000157
  » https://doi.org/10.1136/bmjgh-2016-000157
• ¹⁰ Macinko J, Harris MJ. Brazil's Family Health Strategy - Delivering Community-Based Primary Care in a Universal Health System. N Engl J Med 2015; 372(23):2177-2181. Doi: 10.1056/NEJMp1501140
  » https://doi.org/10.1056/NEJMp1501140
• ¹¹ Rocha R, furtado I, Spinola P. Financing needs, spending projection, and the future of health in Brazil. Health Econ 2021; 30(5):1082-1094.
• ¹² Brasil. Ministério da Saúde (MS). Contas de saúde na perspectiva da contabilidade internacional: conta SHA para o Brasil, 2015 a 2019 / Ministério da Saúde, Fundação Oswaldo Cruz, Instituto de Pesquisa Econômica Aplicada. - Brasília: IPEA, 2022.
• ¹³ Massuda A, Hone T, Lele FAG, Castro MC, Atun R. The Brazilian health system at crossroads: progress, crisis and resilience. BMJ global health 2018; 3 (4).
• ¹⁴ Doniec K, Dall´Alba R, King L. Brazil´s health catastrophe in the making. Lancet Brazil 2018; 18(1):30853-30855. Doi: 10.1016/S0140-6736(18)30853-308555.
  » https://doi.org/10.1016/S0140-6736(18)30853-308555
• ¹⁵ Rasella D, Basu S, Hone T, Paes-Souza R, Ocké-Reis CO, Millet C. Child morbidity and mortality associated with alternative policy responses to the economic crisis in Brazil: a nationwide microsimulation study. Plos Medicine 2018; 15(5):(1). Doi: 10.1371/journal.pmed.1002570.
  » https://doi.org/10.1371/journal.pmed.1002570.
• ¹⁶ Maruthappu M, NG KYB, Williams C, Atun R, Zeltner T. Government health care spending and child mortality. Pediatrics 2015; 135(4):887-894. Doi: 10.1542/peds.2014-1600.
  » https://doi.org/10.1542/peds.2014-1600
• ¹⁷ Malta DC, Sardinha LMV, Moura L, Lansky S, Leal MC, Szwarcwald CL, França E, Almeida MF, Duarte EC. Atualização da lista de causas de mortes evitáveis por intervenções do Sistema Único de Saúde do Brasil. Epidemiol Serv Saúde 2010; 19(2):173-176. Doi: 10.5123/S1679-49742010000200010.
  » https://doi.org/10.5123/S1679-49742010000200010
• ¹⁸ Macinko J, Souza MFM, Guanais FC, Simões CCS. Going to Scale with Community-Based Primary Care: an analysis of the family health program and infant mortality in Brazil, 1999-2004. Soc Sci Med 2007; 65(10):2070-2080. Doi: 10.1016/j.socscimed.2007.06.028.
  » https://doi.org/10.1016/j.socscimed.2007.06.028
• ¹⁹ Brentani A, Grisi SJFE, Taniguchi MT, Ferrer APS, De Moraes Bourroul ML, Fink G. Rollout of community-based family health strategy (Programa de Saúde de Família) is associated with large reductions in neonatal mortality in São Paulo, Brazil. SSM Popul Health 2016; 21(2):55-61. Doi: 10.1016/j.ssmph.2016.01.001
  » https://doi.org/10.1016/j.ssmph.2016.01.001
• ²⁰ Leal MDC, Szwarcwald CL, Almeida PVB., Aquino EML, Barreto ML, Barros F, Victora C. Saúde reprodutiva, materna, neonatal e infantil nos 30 anos do Sistema Único de Saúde (SUS). Ciên Saúde Colet 2018; 23(6):1915-1928.
• ²¹ Guerra AB, Guerra LM, Probst LF, Gondinho BVC, Ambrosano GMB, Melo EA, Brizon VSC, Bulgareli JV, Cortellazzi KL, Pereira AC. Can the primary health care model affect the determinants of neonatal, post-neonatal and maternal mortality? A studyfrom Brazil. BMC Health Services Research 2019; 19(1):1-11.
• ²² Fonseca JFA, Liberal MMC, Varela PS, Zucchi P. Evolução da estratégia Saúde da Família e sua influência nos indicadores da Atenção Básica em Saúde no Estado de Mato Grosso. Research, Society and Development 2021; 10(14). Doi: 10.33448/rsd-v10i14.21866.
  » https://doi.org/10.33448/rsd-v10i14.21866.
• ²³ Capucci P. Financiamento para atenção básica à saúde no Brasil: avanços e desafios. J Manag Prim Health Care 2014; 5(1):127-128. Doi: 10.14295/jmphc.v5i1.206.
  » https://doi.org/10.14295/jmphc.v5i1.206
• ²⁴ Machado CV, Lima LD, Andrade CLT. Federal funding of health policy in Brazil: trends and challenges. Cad Saúde Pública 2014; 30(1):187-200. Doi: 10.1590/0102-311X00144012.
  » https://doi.org/10.1590/0102-311X00144012
• ²⁵ Brasil. Ministério da Saúde (MS). Secretaria de Atenção Primária à Saúde (SAPS). Histórico de Cobertura da Atenção Primária. [Internet] 2021 [acessado 2022 jan.3]. Disponível em: https://egestorab.saude.gov.br/paginas/acessoPublico/relatorios/relHistoricoCoberturaConsolidado.xhtml
  » https://egestorab.saude.gov.br/paginas/acessoPublico/relatorios/relHistoricoCoberturaConsolidado.xhtml
• ²⁶ Antunes JLF, Cardoso MRA. Uso da análise de séries temporais em estudos epidemiológicos. Epidemiol Serv Saúde 2015; 24(3):565-576. Doi: 10.5123/S1679-49742015000300024.
  » https://doi.org/10.5123/S1679-49742015000300024
• ²⁷ Ramalho AA, Andrade AM, Martins FA, Koifman RJ. Tendência da mortalidade infantil no município de Rio Branco, AC, 1999 a 2015. Rev Saúde Pública 2018; 52(1):33. Doi: 10.11606/S1518-8787.2018052000280.
  » https://doi.org/10.11606/S1518-8787.2018052000280
• ²⁸ Carvalho RAS, Santos VS, Melo CM, Gurgel RQ, Oliveira CCC. Desigualdades em saúde: condições de vida e mortalidade infantil em região do nordeste do Brasil. Rev Saude Publica 2015; 49(5):1-9. Doi:10.1590/S0034-8910.2015049004794.
  » https://doi.org/10.1590/S0034-8910.2015049004794
• ²⁹ Garcia LP, Fernandes CM, Traebert J. Risk factors for neonatal death in the capital city with the lowest infant mortality rate in Brazil. J Pediatr 2019; 95(2):194-200. Doi: 10.1016/j.jped.2017.12.007.
  » https://doi.org/10.1016/j.jped.2017.12.007
• ³⁰ Szwarcwald CL, Boccolin CS, Almeida WS, Soares Filho AM, Malta DC. Covid-19 mortality in Brazil, 2020-21: consequences of the pandemic inadequate management. Arch Public Health 2022; 80(255):1-9. Doi: 10.1186/s13690-022-01012-z.
  » https://doi.org/10.1186/s13690-022-01012-z
• ³¹ Barbosa TAGS, Gazzinelli A, Andrade GN. Mortalidade infantil evitável e vulnerabilidade social no Vale do Jequitinhonha, Minas Gerais, Brasil. REME - Rev Min Enferm 2019; 23(1):(1). Doi: 10.5935/1415-2762.20190094.
  » https://doi.org/10.5935/1415-2762.20190094.
• ³² Silva FR, Aguiar DN, Gonçalves MF, Gusmão RB, De Freitas ED, DE Oliveira Simões M, Alves WA. Análise da mortalidade infantil no leste de Minas Gerais, 2008-2019. HU Rev 2021; 47(1):1-11. Doi: 10.34019/1982-8047.2021.v47.34178.
  » https://doi.org/10.34019/1982-8047.2021.v47.34178
• ³³ Adamski K, Da Silva TG, Da Silva Pereira PP, Dos Santos Farias E, Cantarelli KJ, Mendes VA. Mortalidade infantil por causas evitáveis em macrorregião de saúde: série temporal 2007 a 2020. REAS, 2022;15(8):1-9.
• ³⁴ Maia LTS, Souza WV, Mendes ACG. Determinantes individuais e contextuais associados à mortalidade infantil nas capitais brasileiras: uma abordagem multinível. Cad Saude Publica 2020; 36(2):(0).
• ³⁵ Alipour A, Hantoushzadeh S, Hessami K, Saleh M, Shariat M, Yazdizadeh B, Babaniamansour S, Ghamari A, Aghajanian S, Moradi K, Abdolmaleki AS, Emami Z. A global study of the association of cesarean rate and the role of socioeconomic status in neonatal mortality rate in the current century. BMC Pregnancy Childbirth 2022; 22(821):1-9. Doi: 10.1186/s12884-022-05133-9.
  » https://doi.org/10.1186/s12884-022-05133-9
• ³⁶ Moura BLA, Alencar GP, Silva ZPD, Almeida MFD. Fatores associados à internação e à mortalidade neonatal em uma coorte de recém-nascidos do Sistema Único de Saúde, no município de São Paulo. Rev Bras Epidemiol 2020; 23:(2).
• ³⁷ Bernardino FBS, Gonçalves TM, Pereira TID, Xavier JS, Freitas BHBMD, Gaíva, MAM. Tendência da mortalidade neonatal no Brasil de 2007 a 2017. Ciên Saude Colet 2022; 27(2):567-578. Doi: 0000-0003-0339-9451.
  » https://doi.org/0000-0003-0339-9451
• ³⁸ Prezotto KH, Oliveira RR, Pelloso SM, Fernandes CAM. Tendência da mortalidade neonatal evitável nos Estados do Brasil. Rev Bras Saude Mater Infant 2021; 21(1):301-309. Doi: 10.1590/1806-93042021000100015.
  » https://doi.org/10.1590/1806-93042021000100015
• ³⁹ O´hare BA. International corporate tax avoidance and domestic government health expenditure. Bull World Health Organ 2019; 97(1):746-753. Doi: 10.2471/BLT.18.220897.
  » https://doi.org/10.2471/BLT.18.220897
• ⁴⁰ Santos Neto JAS, Mendes NA, Pereira AC, Paranhos LR. Análise do financiamento e gasto do Sistema Único de Saúde dos municípios da região de saúde Rota dos Bandeirantes do estado de São Paulo, Brasil. Ciên Saúde Colet 2017; 22(4):1260-1280. Doi: 10.1590/1413-81232017224.28452016.
  » https://doi.org/10.1590/1413-81232017224.28452016
• ⁴¹ Piola SF, França JRM, Nunes A. Os efeitos da Emenda Constitucional 29 na alocação regional dos gastos públicos no Sistema Único de Saúde no Brasil. Ciên Saúde Colet 2016; 21(2): 411-421. Doi: 10.1590/1413-81232015212.10402015.
  » https://doi.org/10.1590/1413-81232015212.10402015
• ⁴² Funcia Fr. Subfinanciamento e orçamento federal do SUS: referências preliminares para a alocação adicional de recursos. Ciên Saúde Colet 2019; 22(4):4405-4414. Doi: 10.1590/1413-812320182412.25892019.
  » https://doi.org/10.1590/1413-812320182412.25892019
• ⁴³ Araújo CEL, Gonçalves GQ, Machado JA. Os municípios brasileiros e os gastos próprios com saúde: algumas associações. Ciên Saúde Colet 2017; 22(3):953-963. Doi: 10.1590/1413-81232017223.15542016.
  » https://doi.org/10.1590/1413-81232017223.15542016
• ⁴⁴ Rintzel LT, Godoy MR, Spolavani R, Cavassola S. Redução da mortalidade infantil na Região Sul do Brasil nos anos de 2000-2010. Perspectiva Econômica 2018; 14(1):17-32. Doi: 10.4013/pe.2018.141.02.
  » https://doi.org/10.4013/pe.2018.141.02
• ⁴⁵ Jung RO, Agranonik M. Óbitos infantis evitáveis no Rio Grande do Sul: diferenças entre os períodos neonatal e pós-neonatal. Indic Econ FEE 2018; 45(3):51-66.
• ⁴⁶ Aquino R, Oliveira NF, Barreto ML. Impact of the Family Health Program on infant mortality in Brazilian municipalities. Am J Public Health 2009; 99(1): 87-93.

Edited by

• Chief Editors:
  Maria Cecília de Souza Minayo, Romeu Gomes, Antônio Augusto Moura da Silva

Publication Dates

• Publication in this collection
  02 June 2025
• Date of issue
  2025

History

• Received
  24 Oct 2023
• Accepted
  07 Mar 2024
• Published
  09 Mar 2024