Abstract:

Hospitalization is a frequent event in early childhood. In Brazil, the National Household Sample Survey of 2008 showed a 9% hospitalization rate among children in the first four years of life. The study aimed to describe the characteristics of hospitalization in the first six years of life and analyze the early factors associated with hospitalization in a birth cohort in southern Brazil. A zero-inflated Poisson model was used to simultaneously examine the effects of co-variables for the occurrence of a given event and to count events. The frequency of at least one episode of hospitalization during the study period was 33.4% (95%CI: 31.8-34.9), and was highest in the first year (19.1%; 95%CI: 17.9-20.4), remaining stable at approximately 10% between the first and fourth years, decreasing to 8.4% (95%CI: 7.6-9.4) between the fourth and sixth years. diseases of the respiratory system were among the leading causes of hospitalization, followed by infectious and parasitic diseases. History of prior hospitalization was one of the most important predictors of odds of hospitalization and risk of multiple hospitalizations. In early childhood, birth weight, gestational age, Apgar score, sex, and type of pregnancy were inversely associated with hospitalization, and environmental characteristics such as maternal smoking in pregnancy, mother’s skin color, and low family income were associated statistically with number of hospitalizations. The results point to the importance of focusing efforts on reducing hospitalizations from diseases of the respiratory system, especially in children under one year.

Keywords:
Hospitalization; Infant; Cohort Studies

Introduction

Hospitalization is defined as the admission and stay in a hospital establishment for more than 24 hours in the attempt to recover the individual’s health ¹1. Departamento de Sistemas e Redes Assistenciais, Secretaria de Assistência à Saúde, Ministério da Saúde. Padronização da nomenclatura do censo hospitalar. 2ª Ed. Brasília: Ministério da Saúde; 2002. (Série A, Normas e Manuais Técnicos).. Worldwide, hospitalizations account for nearly half of public spending in health ²2. Russell RB, Green NS, Steiner CA, Meikle S, Howse JL, Poschman K, et al. Cost of hospitalization for preterm and low birth weight infants in the United States. Pediatrics 2007; 120:e1-9.. Costs are related to the characteristics of the diseases that caused the hospitalization and the patient’s age ²2. Russell RB, Green NS, Steiner CA, Meikle S, Howse JL, Poschman K, et al. Cost of hospitalization for preterm and low birth weight infants in the United States. Pediatrics 2007; 120:e1-9.^,33. Yorita KL, Holman RC, Sejvar JJ, Steiner CA, Schonberger LB. Infectious disease hospitalizations among infants in the United States. Pediatrics 2008; 121:244-52.^,44. Berenstein CK. O perfil etário dos custos de internação na saúde pública no Brasil: uma análise para as capitais das regiões metropolitanas do Brasil em 2000 [Dissertação de Mestrado]. Belo Horizonte: Universidade Federal de Minas Gerais; 2005.^,55. Ferrer AP, Sucupira AC, Grisi SJ. Causes of hospitalization among children ages zero to nine years old in the city of São Paulo, Brazil. Clinics 2010; 65:35-44.. Diseases with high incidence can lead to the hospitalization of a high number of individuals, which entails higher costs ²2. Russell RB, Green NS, Steiner CA, Meikle S, Howse JL, Poschman K, et al. Cost of hospitalization for preterm and low birth weight infants in the United States. Pediatrics 2007; 120:e1-9.^,33. Yorita KL, Holman RC, Sejvar JJ, Steiner CA, Schonberger LB. Infectious disease hospitalizations among infants in the United States. Pediatrics 2008; 121:244-52.^,44. Berenstein CK. O perfil etário dos custos de internação na saúde pública no Brasil: uma análise para as capitais das regiões metropolitanas do Brasil em 2000 [Dissertação de Mestrado]. Belo Horizonte: Universidade Federal de Minas Gerais; 2005.. In addition, costs tend to be higher in extreme age groups, in rare diseases, and in diseases whose treatment requires highly specialized technology ²2. Russell RB, Green NS, Steiner CA, Meikle S, Howse JL, Poschman K, et al. Cost of hospitalization for preterm and low birth weight infants in the United States. Pediatrics 2007; 120:e1-9.^,33. Yorita KL, Holman RC, Sejvar JJ, Steiner CA, Schonberger LB. Infectious disease hospitalizations among infants in the United States. Pediatrics 2008; 121:244-52.^,44. Berenstein CK. O perfil etário dos custos de internação na saúde pública no Brasil: uma análise para as capitais das regiões metropolitanas do Brasil em 2000 [Dissertação de Mestrado]. Belo Horizonte: Universidade Federal de Minas Gerais; 2005..

Hospitalization is a frequent event in early childhood and preschool age ⁵5. Ferrer AP, Sucupira AC, Grisi SJ. Causes of hospitalization among children ages zero to nine years old in the city of São Paulo, Brazil. Clinics 2010; 65:35-44.. In Brazil, the National Household Sample Survey (PNAD 2008) showed an 8.9% hospitalization rate among children from zero to four years of age. Among hospitalized children, 54.7% were males and 80.7% had been hospitalized only once ⁶6. Instituto Brasileiro de Geografia e Estatística. Pesquisa Nacional por Amostra de Domicílios - PNAD 2008. Um panorama da saúde no Brasil. Acesso e utilização dos serviços, condições de saúde e fatores de risco e proteção à saúde 2008. Rio de Janeiro: Instituto Brasileiro de Geografia e Estatística; 2010..

Respiratory diseases are the most frequent causes of morbidity and mortality in children under five years worldwide, especially in high-income countries ⁷7. Rudan I, Boschi-Pinto C, Biloglav Z, Mulholland K, Campbell H. Epidemiology and etiology of childhood pneumonia. Bull World Health Organ 2008; 86:408-16.^,88. Bryce J, Boschi-Pinto C, Shibuya K, Black RE; WHO Child Health Epidemiology Reference Group. WHO estimates of the causes of death in children. Lancet 2005; 365:1147-52.. In medium and low-income countries, infectious diseases are still important causes of hospitalization, despite the decline in hospitalizations from acute diarrhea since the advent of oral rehydration therapy and such interventions as maternal breastfeeding and immunizations ⁷7. Rudan I, Boschi-Pinto C, Biloglav Z, Mulholland K, Campbell H. Epidemiology and etiology of childhood pneumonia. Bull World Health Organ 2008; 86:408-16.^,88. Bryce J, Boschi-Pinto C, Shibuya K, Black RE; WHO Child Health Epidemiology Reference Group. WHO estimates of the causes of death in children. Lancet 2005; 365:1147-52.^,99. Wiegering V, Kaiser J, Tappe D, Weissbrich B, Morbach H, Girschick HJ. Gastroenteritis in childhood: a retrospective study of 650 hospitalized pediatric patients. Int J Infect Dis 2011; 15:e401-7..

In Brazil, respiratory diseases are the leading cause of hospitalization in children under five years, followed by infectious intestinal diseases ⁵5. Ferrer AP, Sucupira AC, Grisi SJ. Causes of hospitalization among children ages zero to nine years old in the city of São Paulo, Brazil. Clinics 2010; 65:35-44.^,1010. Oliveira BRG, Viera CS, Collet N, Lima RAG. Causas de hospitalização no SUS de crianças de zero a quatro anos no Brasil. Rev Bras Epidemiol 2010; 13:268-77.. Hospitalizations from these causes are distributed unequally across the regions of Brazil. In the South, Southeast, and Central, hospitalizations from respiratory diseases are more common, while admissions for infectious and parasitic diseases predominate in the North and Northeast ⁶6. Instituto Brasileiro de Geografia e Estatística. Pesquisa Nacional por Amostra de Domicílios - PNAD 2008. Um panorama da saúde no Brasil. Acesso e utilização dos serviços, condições de saúde e fatores de risco e proteção à saúde 2008. Rio de Janeiro: Instituto Brasileiro de Geografia e Estatística; 2010.^,1010. Oliveira BRG, Viera CS, Collet N, Lima RAG. Causas de hospitalização no SUS de crianças de zero a quatro anos no Brasil. Rev Bras Epidemiol 2010; 13:268-77.^,1111. Ministério da Saúde. Pesquisa Nacional de Demografia e Saúde da Criança e da Mulher - PNDS 2006: dimensões do processo reprodutivo e da saúde da criança. Brasília: Ministério da Saúde; 2009. (Série G. Estatística e Informação em Saúde)..

In the South of Brazil, Matijasevich et al. ¹²12. Matijasevich A, Cesar JA, Santos IS, Barros AJ, Dode MA, Barros FC, et al. Hospitalizations during infancy in three population-based studies in Southern Brazil: trends and differentials. Cad Saúde Pública 2008; 24 Suppl 3:S437-43. evaluated hospitalizations during childhood in three longitudinal population-based studies. All-cause hospitalization at least once in the first year of life remained constant from 1982 to 2004 (19.6% in 1982; 18.1% in 1993; and 19.2% in 2004; p = 0.7). However, there was a sharp reduction in hospitalizations from diarrhea during the same period (6.3% in 1982; 3.2% in 1993; and 1.0% in 2004; p < 0.001).

The current study aims to describe the characteristics of hospitalization in the first six years of life and analyze early factors associated with odds of hospitalization and risk of multiple hospitalizations in a birth cohort in southern Brazil.

Methods

Study site

Pelotas is a medium-sized municipality, predominantly urban, located in Rio Grande do Sul State, Brazil. The population in 2004 was estimated at 340,000. At the time, Pelotas was an impoverished municipality, with a per capita Gross Domestic Product (GDP) below the mean for Brazil and Rio Grande do Sul (Instituto Brasileiro de Geografia e Estatística. Censo Demográfico 2000. http://www.ibge.gov.br/home/estatistica/populacao/censo2000/universo.php?tipo=31o/tabela13_1.shtm&paginaatual=1&uf=43&letra=P, accessed on 28/Jul/2016).

Study population and design

From January 1 to December 31, 2004, all the hospitals in the city were visited daily to identify births. Postpartum women were interviewed on their socioeconomic, demographic, reproductive, and healthcare characteristics during pregnancy and childbirth. The newborns were examined by trained field researchers. The original cohort consisted of 4,231 live newborns, corresponding to 99.2% of the births in Pelotas in the year 2004.

Children participating in the study were followed at one, two, four, and six years of life. At each follow-up, the mothers were interviewed and the children were measured and weighed by trained field researchers. The follow-up rate at all visits exceeded 90% ¹³13. Santos IS, Barros AJ, Matijasevich A, Zanini R, Chrestani Cesar MA, Camargo-Figuera FA, et al. Cohort profile update: 2004 Pelotas (Brazil) Birth Cohort Study. Body composition, mental health and genetic assessment at the 6 years follow-up. Int J Epidemiol 2014; 43:1437-1437a-f.. All the follow-up waves of the 2004 Birth Cohort were approved by the Institutional Review Board of the School of Medicine, Pelotas Federal University.

Variables and statistical analyses

The study focused on the rate, number, and causes of hospitalization in early childhood, considering hospitalizations that occurred after hospital discharge following birth.

Information on hospitalization in the first year of life was collected at 12 months using the following question: “Has your child been admitted to hospital since birth and until now?”. Hospitalization between 1 and 2 years, 2 and 4, and 4 and 6 years was collected using the previous question but referring to the respective period at the 2, 4, and 6-year follow-up visits, respectively. We calculated the rate of at least one hospitalization in the period from one to six years (sometime in life) and in the following periods: in the first year, between the first and second year, between the second and fourth year, and between the fourth and sixth year of life. To calculate the lifetime hospitalization rate, we only included children with information on hospitalization at all the follow-up visits. The hospitalization rate in each period was collected as discrete, and for the analysis it was categorized: none, one, two, three, four, and five or more times. Causes of hospitalization were collected with an open question: “Why was <the child> hospitalized?” or “Why was he/she admitted to hospital?” Answers were coded according to the International Classification of Diseases, 10^th edition (ICD-10). Later, the ten leading causes of hospitalization were listed for each follow-up wave according to the ICD-10 groups.

The early factors were collected in the perinatal study. Birth weight was classified in four groups (< 2,500g, 2,500g-2,999g, 3,000g-3,499g, and ≥ 3,500g). Gestational age was categorized as ≥ 37 weeks (term) and < 37 weeks (preterm). Family income was divided into quintiles, with the first quintile including the poorest 20% and the fifth quintile the wealthiest 20%. The monthly income quintile cutoffs were: BRL 0-≤ 240 (Q1); 241-≤ 400 (Q2); 401-≤ 600 (Q3); 601-≤ 1,000 (Q4); and 1,001-≤ 22,000 (Q5).

Parity was categorized as < 2 or ≥ 2 children, and 5-minute Apgar score as ≥ 7 or < 7 points. The mother’s and child’s skin color were observed by the interviewer and categorized as white, black, or brown.

The other variables, such as sex (male/female), type of delivery (vaginal/cesarean), and maternal smoking (yes/no) were analyzed the same way as collected. Alcohol consumption during the pregnancy was assessed through a question addressed to the mother and analyzed dichotomously (yes/no).

Early factors associated with hospitalization were evaluated with the Zero-Inflated Poisson model (ZIP) ¹⁴14. Lambert D. Zero-inflated Poisson regression, with an application to defects in manufacturing. Technometrics 1992; 34:1-14., which allows simultaneously examining the effects of co-variables on the binomial and counting components. Thus, the early factors that affect the odds of hospitalization were estimated simultaneously with the factors that affect the number of hospitalizations in the children. The estimated measures of association were odds ratio of hospitalization and relative risk for the number of hospitalizations. Wald’s chi-square test was used to assess whether the models’ estimates were statistically different from zero.

The crude and adjusted analyses were limited to children with information for the outcome and predictors. In the adjusted model, for each study period, we included the mother’s and child’s characteristics with significance < 0.20 in the crude analysis. The variables were included only once in the equation and were then removed one by one until the best predictors were identified (backward elimination strategy). The variables included in the final model varied according to the child’s age.

Interactions with the sex variable were tested, and no significant interactions were observed. Colinearity was seen between the variables. The correlation matrix did not show highly correlated variables.

The ZIP model was compared to the traditional Poisson model using the Voung test ¹⁵15. Vuong QH. Likelihood ratio tests for model selection and non-nested hypotheses. Econometrica 1989; 57:307-33., and in all cases ZIP was more adequate for the data analysis. The statistical analyses used Stata 13.1 (StataCorp LP, College Station, USA).

Results

Of the 4,231 newborns that entered the 2004 Pelotas cohort, 3,907, 3,869, 3,799, and 3,722 children were followed at two, four, and six years of age, respectively (with rates greater than 90% at each follow-up). At one, two, four, and six years of age, there was information on hospitalization for 3,907 (100%), 3,868 (99%), 3,799 (100%), and 3,646 (98%) children, respectively.

Among the children studied, 51.9% were males, 28.9% had black or brown skin color, 45.2% were delivered by cesarean, 10.9% with birth weight < 2500g, 11.6% with gestational age < 37 weeks, and 2.1% with 5-minute Apgar < 7. As for maternal characteristics, 72.8% of the mothers were white, 60.7% had two or more children, 21.1% smoked, and 3.3% had consumed alcohol during the pregnancy.

The rate of at least one lifetime hospitalization from birth to six years of age was 33.4% (95%CI: 31.8-34.9). Analyzing hospitalization according to cohort follow-up waves, the hospitalization rate in the first year of life was 19.1% (95%CI: 17.9-20.4), remained stable at 10% from the first to fourth years, and decreased to 8.4% (95%CI: 7.6-9.4) from the fourth to the sixth years of life (Table 1).

At all the follow-up waves, the majority of the children had been hospitalized once. In the first year of life, 15.0% (95%CI: 13.9-16.1) had one hospitalization, decreasing in the second year to 8.2% (95%CI: 7.4-9.1), and then remaining stable throughout the period (7.6% - 2 to 4 years; 6.8% - 4 to 6 years). The proportion of children with two or more hospitalizations varied from 4.2% in the first year of life to 1.7% from the fourth to the sixth year (Table 1).

Table 2 shows the causes of hospitalization according to ICD-10 chapters. Diseases of the Respiratory System (Chapter X) were among the leading causes of hospitalization in all the years, with 9.6% in the first year of life, around 5% from one to four years, decreasing to 2.9% from the fourth to the sixth year of life. The second leading cause of hospitalization was Infectious and Parasitic Diseases (Chapter I), with proportions around 1%.

As for ranking of the ten leading causes of hospitalization by ICD-10 groups (Table 3), “Influenza and pneumonia” and “Chronic diseases of the lower airways” were among the three leading causes of hospitalization in all the follow-up visits. “Infectious intestinal diseases” such as diarrhea, viral gastroenteritis, and other diarrheas not specified were classified from third to fifth, depending on the age at follow-up. from two to six years of age, surgical causes ranked from second to third.

In the first year of life, according to the crude analysis, birth weight, gestational age, and type of pregnancy were inversely associated with odds of hospitalization (Table 4). In the adjusted analysis, children with gestational age < 37 weeks and those with low birth weight (< 2,500g) showed odds of hospitalization three times higher than term infants and those with birth weight from 3,000g to 3,499g, respectively. Infants with low birth weight (LBW - RR = 1.7; 95%CI: 1.3-2.2) and those from the poorest families showed higher risk of multiple hospitalizations when compared to the reference category. Maternal smoking during pregnancy was associated with increased risk of readmissions (RR = 1.2; 95%CI: 1.0-1.4). Girls showed approximately 20% lower risk of readmission when compared to boys.

From the first to the second year of life (Table 4), in the adjusted analysis, children with LBW and 5-minute Apgar < 7 showed odds of hospitalization twice as high (OR = 2.2; 95%CI: 1.0-4.8) and six times higher (OR 6.1; 95%CI: 1.8-21.1), respectively, than the reference category. Children hospitalized in the first year of life showed nearly three times greater risk of multiple hospitalizations than those not hospitalized in the previous year (RR = 2.9; 95%CI: 2.3-3.8). Children of mothers that had smoked during pregnancy and from families in the poorest quintile showed 40% (RR = 1.4; 95%CI: 1.1-1.8) and 80% (RR = 1.8; 95%CI: 1.3-2.6) higher risk of readmission when compared to children of mothers that had not smoked during pregnancy and those from families in the highest income quintile, respectively.

From two to four years, in the crude analysis, the only factors associated with odds of hospitalization were gestational age, hospitalization from to one to two years of life, and maternal smoking during pregnancy (Table 5). After adjustment, odds of hospitalization were higher in children that had been hospitalized during the previous period (OR = 3.4; 95%CI: 2.0-5.9) and in children of mothers who had smoked during pregnancy (OR = 1.7; 95%CI: 1.0-2.8). Children hospitalized from the first to the second years of life and children of mothers that had smoked during pregnancy showed approximately four and three times greater risk of multiple hospitalizations, respectively, compared to children in the reference categories.

From four to six years, girls showed 30% (OR = 0.7; 95%CI: 0.5-0.9) lower odds of hospitalization than boys. Children of multiple pregnancies showed six times higher odds of hospitalization (OR = 6.3; 95%CI: 1.9-20.9) than those of singleton pregnancies. Children hospitalized from two to four years of life showed five times greater odds of hospitalization (OR = 5.3; 95%CI: 2.9-9.8) than those not hospitalized in the previous period. Children hospitalized from two to four years and those born to multiple pregnancies showed twice the odds (RR = 2.2; 95%CI: 1.3-3.6) and three times the odds (RR = 2.8; 95%CI: 1.6-4.9), respectively, of those in the reference categories (Table 5).

Discussion

In the current study, one out of five children required hospitalization in the first year of life. Diseases of the respiratory system were among the leading causes of hospitalization, followed by infectious and parasitic diseases. Early factors that predicted hospitalization and number of hospitalizations were not the same. In the first two years of life, the child’s and mother’s biological factors such as birth weight, gestational age, Apgar score, sex, and multiple pregnancy were inversely associated with hospitalization, while environmental characteristics such as history of maternal smoking during pregnancy, mother’s skin color, and low family income were associated with number of hospitalizations. For each follow-up period analyzed, history of previous hospitalization was one of the most important predictors of hospitalization and number of hospitalizations.

In keeping with previous studies in Pelotas ¹²12. Matijasevich A, Cesar JA, Santos IS, Barros AJ, Dode MA, Barros FC, et al. Hospitalizations during infancy in three population-based studies in Southern Brazil: trends and differentials. Cad Saúde Pública 2008; 24 Suppl 3:S437-43.^,1616. Menezes AMB, Noal RB, Cesar JA, Hallal PC, Araujo CL, Dumith SC, et al. Hospital admissions from birth to early adolescence and early-life risk factors: the 11-year follow-up of the 1993 Pelotas (Brazil) birth cohort study. Cad Saúde Pública 2010; 26:1980-9. with a similar methodology to that used in this study, hospitalization in childhood, especially in the first year of life, has remained high and stable for at least three decades, at around 20%, thus corroborating the current study’s findings.

Other Brazilian studies ⁵5. Ferrer AP, Sucupira AC, Grisi SJ. Causes of hospitalization among children ages zero to nine years old in the city of São Paulo, Brazil. Clinics 2010; 65:35-44.^,66. Instituto Brasileiro de Geografia e Estatística. Pesquisa Nacional por Amostra de Domicílios - PNAD 2008. Um panorama da saúde no Brasil. Acesso e utilização dos serviços, condições de saúde e fatores de risco e proteção à saúde 2008. Rio de Janeiro: Instituto Brasileiro de Geografia e Estatística; 2010.^,1111. Ministério da Saúde. Pesquisa Nacional de Demografia e Saúde da Criança e da Mulher - PNDS 2006: dimensões do processo reprodutivo e da saúde da criança. Brasília: Ministério da Saúde; 2009. (Série G. Estatística e Informação em Saúde).^,1717. Castro MSM, Travassos C, Carvalho MS. Fatores associados às internações hospitalares no Brasil. Ciênc Saúde Coletiva 2002; 7:795-811. on hospitalization of children found lower hospitalization rates than in the current study. Castro et al. ¹⁷17. Castro MSM, Travassos C, Carvalho MS. Fatores associados às internações hospitalares no Brasil. Ciênc Saúde Coletiva 2002; 7:795-811., using data from the PNAD 1998, showed an 11.1% hospitalization rate in children under one year and 6.9% from one to four years. In PNAD 2008 ⁶6. Instituto Brasileiro de Geografia e Estatística. Pesquisa Nacional por Amostra de Domicílios - PNAD 2008. Um panorama da saúde no Brasil. Acesso e utilização dos serviços, condições de saúde e fatores de risco e proteção à saúde 2008. Rio de Janeiro: Instituto Brasileiro de Geografia e Estatística; 2010., the hospitalization rate in children under four years of age was 8.4%. Ferrer et al. ⁵5. Ferrer AP, Sucupira AC, Grisi SJ. Causes of hospitalization among children ages zero to nine years old in the city of São Paulo, Brazil. Clinics 2010; 65:35-44. studied hospitalization of children from birth to nine years of age, from 2002 to 2006, in Brazil and in the city of São Paulo, and used data from the Brazilian Hospital Information System of the Brazilian Unified National Health System (SIH-SUS), showing a hospitalization rate of 17% in Brazil and 16% in São Paulo ⁵5. Ferrer AP, Sucupira AC, Grisi SJ. Causes of hospitalization among children ages zero to nine years old in the city of São Paulo, Brazil. Clinics 2010; 65:35-44.. The National Demographic and Health Survey (PNDS 2006) ¹¹11. Ministério da Saúde. Pesquisa Nacional de Demografia e Saúde da Criança e da Mulher - PNDS 2006: dimensões do processo reprodutivo e da saúde da criança. Brasília: Ministério da Saúde; 2009. (Série G. Estatística e Informação em Saúde). assessed the morbidity and mortality profile in children less than five years of age in Brazil. In the 12-23-month and 24-35-month age brackets, PNDS 2006 showed higher hospitalization rates than in the current study. Regional inequalities could explain these differences. In PNDS 2006, in the North and Northeast, the hospitalization rates in children under five years were 14.1% and 13.6%, respectively, compared to 10.9% in the South. The hospitalization rates from diarrhea and respiratory diseases were also higher in these age brackets in PNDS 2006 when compared to the current study.

As shown in other studies in different settings in Brazil ⁵5. Ferrer AP, Sucupira AC, Grisi SJ. Causes of hospitalization among children ages zero to nine years old in the city of São Paulo, Brazil. Clinics 2010; 65:35-44.^,66. Instituto Brasileiro de Geografia e Estatística. Pesquisa Nacional por Amostra de Domicílios - PNAD 2008. Um panorama da saúde no Brasil. Acesso e utilização dos serviços, condições de saúde e fatores de risco e proteção à saúde 2008. Rio de Janeiro: Instituto Brasileiro de Geografia e Estatística; 2010.^,1010. Oliveira BRG, Viera CS, Collet N, Lima RAG. Causas de hospitalização no SUS de crianças de zero a quatro anos no Brasil. Rev Bras Epidemiol 2010; 13:268-77.^,1616. Menezes AMB, Noal RB, Cesar JA, Hallal PC, Araujo CL, Dumith SC, et al. Hospital admissions from birth to early adolescence and early-life risk factors: the 11-year follow-up of the 1993 Pelotas (Brazil) birth cohort study. Cad Saúde Pública 2010; 26:1980-9.^,1818. Caetano JJR, Bordin IA, Puccini RF, Peres CCA. Factors associated to hospitalization of children under five years of age, São Paulo, Brazil. Rev Saúde Pública 2002; 36:285-91., Diseases of the Respiratory System (ICD-10 - chapter X) were the leading causes of hospitalization from birth to six years of age and ranked from first to third among the ten leading causes of hospitalization.

Most respiratory diseases in childhood are caused by viruses and/or bacteria that develop more readily in humid, cold, closed, and crowded environments. Gonzales et al. ¹⁹19. Gonzalez DA, Victora CG, Gonçalves H. Efeitos das condições climáticas no trimestre de nascimento sobre asma e pneumonia na infância e na vida adulta em uma coorte no Sul do Brasil. Cad Saúde Pública 2008; 24:1089-102., studying the effects of climate on hospitalizations from asthma and pneumonia in preschoolers in Pelotas, found that children born from April to June (autumn) showed higher risk of hospitalization from pneumonia and asthma/bronchitis in the first three years of life than those born from January to March (summer).

Infectious and Parasitic Diseases (ICD-10 - chapter I) were the second leading cause of hospitalization, accounting for 1% of admissions over the study period. Diarrheic diseases, although presenting lower and more stable percentages, are still an important public health problem. Studies have shown that cases of acute, serious diarrhea persist and are associated with one-third of childhood hospitalizations and an estimated 600,000 deaths per year in children worldwide ³3. Yorita KL, Holman RC, Sejvar JJ, Steiner CA, Schonberger LB. Infectious disease hospitalizations among infants in the United States. Pediatrics 2008; 121:244-52.^,88. Bryce J, Boschi-Pinto C, Shibuya K, Black RE; WHO Child Health Epidemiology Reference Group. WHO estimates of the causes of death in children. Lancet 2005; 365:1147-52.^,99. Wiegering V, Kaiser J, Tappe D, Weissbrich B, Morbach H, Girschick HJ. Gastroenteritis in childhood: a retrospective study of 650 hospitalized pediatric patients. Int J Infect Dis 2011; 15:e401-7..

According to Ferrer et al. ⁵5. Ferrer AP, Sucupira AC, Grisi SJ. Causes of hospitalization among children ages zero to nine years old in the city of São Paulo, Brazil. Clinics 2010; 65:35-44., Certain Conditions Originating in the Perinatal Period (ICD-10 - chapter XVI) were the second cause of hospitalization in São Paulo and the third in Brazil, accounting for 15.9% and 10.1% of admissions, respectively. In the current study, the proportion of hospitalizations from these causes was low (< 1%). Such differences can be attributed to methodological issues between studies and improvements in some maternal-child indicators in the municipality in recent decades, for example improved perinatal care, longer breastfeeding, and an increase in maternal schooling ⁵5. Ferrer AP, Sucupira AC, Grisi SJ. Causes of hospitalization among children ages zero to nine years old in the city of São Paulo, Brazil. Clinics 2010; 65:35-44.^,2020. Barros FC, Victora CG. Maternal-child health in Pelotas, Rio Grande do Sul State, Brazil: major conclusions from comparisons of the 1982, 1993, and 2004 birth cohorts. Cad Saúde Pública 2008; 24 Suppl 3:S461-7.. In the study by Ferrer et al. ⁵5. Ferrer AP, Sucupira AC, Grisi SJ. Causes of hospitalization among children ages zero to nine years old in the city of São Paulo, Brazil. Clinics 2010; 65:35-44., calculation of the hospitalization rate included all hospitalizations of children under one year, without excluding readmissions, which may have overestimated the hospitalization rate from these causes. In addition, the authors used the Hospital Information System of the Unified Health System (SIH-SUS), fed by the Authorizations for Hospital Admissions (AIHs), which have limitations in the quality of diagnosis at hospital admission ⁴4. Berenstein CK. O perfil etário dos custos de internação na saúde pública no Brasil: uma análise para as capitais das regiões metropolitanas do Brasil em 2000 [Dissertação de Mestrado]. Belo Horizonte: Universidade Federal de Minas Gerais; 2005.^,2121. Veras CMT, Martins MS. Reliability of data from Authorization Forms for Hospital Admittance, Rio de Janeiro, Brazil. Cad Saúde Pública 1994; 10:339-55..

Age at first hospitalization, number of previous hospitalizations, disease severity, and socio-environmental conditions are considered risk factors for re-hospitalization of children ²²22. Lasmar L, Goulart E, Sakurai E, Camargos P. Risk factors for hospital admissions among asthmatic children and adolescents. Rev Saúde Pública 2002; 36:409-19.^,2323. Bloomberg RG, Trinkaus KM, Fisher Jr. EB, Musick JR, Strunk RC. Hospital readmissions for childhood asthma: a 10-year metropolitan study. Am J Resp Crit Care Med 2003; 167:1068-76.. Lasmar et al. ²²22. Lasmar L, Goulart E, Sakurai E, Camargos P. Risk factors for hospital admissions among asthmatic children and adolescents. Rev Saúde Pública 2002; 36:409-19. studied asthma patients under 15 years of age and found that odds of re-hospitalization in children with a history of first hospitalization before 12 months of age were 2.5 times higher when compared to children whose first hospitalization occurred after 24 months of age. Bloomberg et al. ²³23. Bloomberg RG, Trinkaus KM, Fisher Jr. EB, Musick JR, Strunk RC. Hospital readmissions for childhood asthma: a 10-year metropolitan study. Am J Resp Crit Care Med 2003; 167:1068-76. studied hospitalizations of children over the course of 10 year and found that after the first hospitalization from asthma, the probability of a second, third, and fourth hospitalization was 30%, 46%, and 59%, respectively. Neuman et al. ²⁴24. Neuman MI, Hall M, Gay JC, Blaschke AJ, Willians DJ, Parikh K, et al. Readmissions among children previously hospitalized with pneumonia. Pediatrics 2014; 134:100-9. investigated factors associated with re-hospitalization after a previous admission for pneumonia. Higher number of admissions due to pneumonia (including more prolonged hospitalizations) was associated with re-hospitalization. The authors also reported that children with complicated pneumonia or those with chronic comorbidities such as neurological, digestive, and circulatory diseases were more predisposed to experiencing re-hospitalization from all causes, when compared to children without chronic diseases.

Meanwhile, the high rate of preterm births in the 2004 birth cohort (15% preterm ¹³13. Santos IS, Barros AJ, Matijasevich A, Zanini R, Chrestani Cesar MA, Camargo-Figuera FA, et al. Cohort profile update: 2004 Pelotas (Brazil) Birth Cohort Study. Body composition, mental health and genetic assessment at the 6 years follow-up. Int J Epidemiol 2014; 43:1437-1437a-f. and 10.8% late preterm ²⁵25. Santos IS, Matijasevich A, Domingues MR, Barros AJ, Victora CG, Barros FC. Late preterm birth is a risk factor for growth faltering in early childhood: a cohort study. BMC Pediatrics 2009; 9:71.) could also explain the high odds of re-hospitalization in the study sample. Preterm or late preterm children have higher odds of hospitalization and re-hospitalization than term children ²⁶26. Boyle EM, Poulsen G, Field DJ, Kurinczuk JJ, Wolke D, Alfirevic Z, et al. Effects of gestational age at birth on health outcomes at 3 and 5 years of age: population based cohort study. BMJ 2012; 344:e896.. In a retrospective birth cohort in Australia that followed children from birth to 18 years, gestational age was inversely associated with greater risk of re-hospitalization. Nearly 62% of the cohort was re-hospitalized at least once during childhood, and more than 90% of children born with gestational age ≤ 33 weeks were re-hospitalized in the first 18 years of life, compared to 59% in children born with gestational age ≥ 39 weeks ²⁷27. Srinivasjois R, Slimings C, Einarsdóttir K, Burgner D, Leonard H. Association of gestational age at birth with reasons for subsequent hospitalisation: 18 years of follow-up in a Western Australian Population Study. PLoS One 2015; 10:e0130535..

Finally, repeated hospitalization may lead to social and clinical problems for the family and child, disorganization of family functioning, and consequences for the child’s physical and emotional development, in addition to increasing the potential for developing hospital infections, especially in children with chronic comorbidities ²⁸28. Nakamura MM, Toomey SL, Zaslavsky AM, Berry JG, Lorch SA, Jha AK, et al. Measuring pediatric hospital readmission rates to drive quality improvement. Acad Pediatr 2014; 14(5 Suppl):S39-46..

Various perinatal health conditions are associated with hospitalization in early childhood ¹²12. Matijasevich A, Cesar JA, Santos IS, Barros AJ, Dode MA, Barros FC, et al. Hospitalizations during infancy in three population-based studies in Southern Brazil: trends and differentials. Cad Saúde Pública 2008; 24 Suppl 3:S437-43.^,1616. Menezes AMB, Noal RB, Cesar JA, Hallal PC, Araujo CL, Dumith SC, et al. Hospital admissions from birth to early adolescence and early-life risk factors: the 11-year follow-up of the 1993 Pelotas (Brazil) birth cohort study. Cad Saúde Pública 2010; 26:1980-9.^,1818. Caetano JJR, Bordin IA, Puccini RF, Peres CCA. Factors associated to hospitalization of children under five years of age, São Paulo, Brazil. Rev Saúde Pública 2002; 36:285-91.. In this study, low birth weight, gestational age < 37 weeks, and 5-minute Apgar score < 7 were associated with hospitalization and number of hospitalizations in the first and/or second years of life. These findings are consistent with those of previous studies ¹²12. Matijasevich A, Cesar JA, Santos IS, Barros AJ, Dode MA, Barros FC, et al. Hospitalizations during infancy in three population-based studies in Southern Brazil: trends and differentials. Cad Saúde Pública 2008; 24 Suppl 3:S437-43.^,1616. Menezes AMB, Noal RB, Cesar JA, Hallal PC, Araujo CL, Dumith SC, et al. Hospital admissions from birth to early adolescence and early-life risk factors: the 11-year follow-up of the 1993 Pelotas (Brazil) birth cohort study. Cad Saúde Pública 2010; 26:1980-9.^,1818. Caetano JJR, Bordin IA, Puccini RF, Peres CCA. Factors associated to hospitalization of children under five years of age, São Paulo, Brazil. Rev Saúde Pública 2002; 36:285-91.. Children with low birth weight are more vulnerable to environmental and social factors and show greater risk of morbidity and mortality in early childhood ¹²12. Matijasevich A, Cesar JA, Santos IS, Barros AJ, Dode MA, Barros FC, et al. Hospitalizations during infancy in three population-based studies in Southern Brazil: trends and differentials. Cad Saúde Pública 2008; 24 Suppl 3:S437-43.^,1818. Caetano JJR, Bordin IA, Puccini RF, Peres CCA. Factors associated to hospitalization of children under five years of age, São Paulo, Brazil. Rev Saúde Pública 2002; 36:285-91.. Selling et al. ²⁹29. Selling KE, Carstensen J, Finnstrom O, Josefsson A, Sydsjo G. Hospitalizations in adolescence and early adulthood among Swedish men and women born preterm or small for gestational age. Epidemiology 2008; 19:63-70. assessed the long-term effects of preterm birth and intrauterine growth restriction and found greater risk of hospitalization in adolescence and adulthood among those born small for gestational age, preterm, and those born both preterm and small for gestational age ²⁹29. Selling KE, Carstensen J, Finnstrom O, Josefsson A, Sydsjo G. Hospitalizations in adolescence and early adulthood among Swedish men and women born preterm or small for gestational age. Epidemiology 2008; 19:63-70.. Matijasevich et al. ¹²12. Matijasevich A, Cesar JA, Santos IS, Barros AJ, Dode MA, Barros FC, et al. Hospitalizations during infancy in three population-based studies in Southern Brazil: trends and differentials. Cad Saúde Pública 2008; 24 Suppl 3:S437-43., in a study on hospitalization over the course of 22 years in three birth cohorts, found that birth weight was one of the factors most closely associated with higher hospitalization rates. Apgar score, the method used to assess the newborn’s clinical status at birth, is considered a predictor of neonatal complications and mortality ³⁰30. American Academy of Pediatrics Committee on Fetus and Newborn; American College of Obstetricians and Gynecologists Committee on Obstetric Practice. The Apgar score. Pediatrics 2015; 136:819-22.. Low 5-minute Apgar score is associated with increased risk of brain damage ³¹ and greater risk of hospitalization due to epilepsy in the first 12 years of life ³¹31. Ehrenstein V, Sørensen HT, Pedersen L, Larsen H, Holsteen V, Rothman KJ. Apgar score and hospitalization for epilepsy in childhood: a registry-based cohort study. BMC Public Health 2006; 6:1-7..

In this study, male gender was associated with greater odds of hospitalization and more hospitalizations in children under one year and from four to six years. Other studies have shown the predominance of male gender in morbidity and hospitalization from diseases of the respiratory tract in childhood, with an inversion in puberty, when girls show higher incidence of hospitalization ¹⁶16. Menezes AMB, Noal RB, Cesar JA, Hallal PC, Araujo CL, Dumith SC, et al. Hospital admissions from birth to early adolescence and early-life risk factors: the 11-year follow-up of the 1993 Pelotas (Brazil) birth cohort study. Cad Saúde Pública 2010; 26:1980-9.^,1717. Castro MSM, Travassos C, Carvalho MS. Fatores associados às internações hospitalares no Brasil. Ciênc Saúde Coletiva 2002; 7:795-811.^,1818. Caetano JJR, Bordin IA, Puccini RF, Peres CCA. Factors associated to hospitalization of children under five years of age, São Paulo, Brazil. Rev Saúde Pública 2002; 36:285-91.. A possible explanation for this finding relates to the lower caliber of airways and immunological differences in boys during childhood and hormonal changes in girls in puberty ³²32. Almqvist C, Worm M, Leynaert B; Working Group of GALEN WP 2.5 Gender. Impact of gender on asthma in childhood and adolescence: a GA2LEN review. Allergy 2008; 63:47-57..

Children born to multiple pregnancies are at greater risk of preterm birth and/or low birth weight and are more likely to need additional health care ³³33. Henderson J, Hockley C, Petrou S, Goldacre M, Davidson L. Economic implications of multiple births: inpatient hospital costs in the first 5 years of life. Arch Dis Child Fetal Neonatal Ed 2004; 89:F542-5.. In Henderson et al. ³³33. Henderson J, Hockley C, Petrou S, Goldacre M, Davidson L. Economic implications of multiple births: inpatient hospital costs in the first 5 years of life. Arch Dis Child Fetal Neonatal Ed 2004; 89:F542-5., the duration of hospitalization of twin and triplet infants was twice as long and eight times longer, respectively, compared to singleton children.

Various studies have shown racial inequalities in hospitalization rates. Hospitalization rates due to diarrhea were higher in black children under six months of age when compared to white children, and this difference was not modified following introduction of the rotavirus vaccine ³⁴34. Yen C, Steiner CA, Barrett M, Curns AT, Hunter K, Wilson E, et al. Racial disparities in diarrhea-associated hospitalizations among children in five US States, before and after introduction of rotavirus vaccine. Vaccine 2010; 28:7423-6.. African-American children showed greater risk of hospitalization from asthma/wheezing and greater risk of readmission than white children, and this inequality was attributed to differences in access to health care and associated socioeconomic variables ³⁴34. Yen C, Steiner CA, Barrett M, Curns AT, Hunter K, Wilson E, et al. Racial disparities in diarrhea-associated hospitalizations among children in five US States, before and after introduction of rotavirus vaccine. Vaccine 2010; 28:7423-6..

There is evidence in the literature of the association between maternal smoking during pregnancy and risk of hospitalization in childhood ³⁵35. Beck AF, Huang B, Auger KA, Ryan PH, Chen C, Kahn RS. Explaining racial disparities in child asthma readmission using a causal inference approach. JAMA Pediatr 2016; 170:695-703.^,3636. Kukla L, Hruba D, Tyrlik M. Influence of prenatal and postnatal exposure to passive smoking on infants&apos; health during the first six months of their life. Cent Eur J Public Health 2004; 12:157-60.. In this study, children of mothers that smoked showed higher hospitalization rates than children born to mothers that did not smoke during pregnancy.

Family income at birth was inversely associated with hospitalization in children under four years. Children of poor families are more prone to be exposed to pathogens and have greater risk of illness and less access to health services and medical treatments ³⁷37. DiFranza JR, Aligne CA, Weitzman M. Prenatal and postnatal environmental tobacco smoke exposure and children&apos;s health. Pediatrics 2004; 113(4 Suppl):1007-15.. Studies on hospitalization are heavily influenced not only by socioeconomic factors, but also by the health system’s characteristics and professional practices ²²22. Lasmar L, Goulart E, Sakurai E, Camargos P. Risk factors for hospital admissions among asthmatic children and adolescents. Rev Saúde Pública 2002; 36:409-19..

The study’s limitations include the possibility of Berkson’s bias, since children of low-income families can be hospitalized more often than children of higher-income families, increasing the hospitalization rate in this group, which already displays higher morbidity and mortality rates ³⁸38. Pearce N, Richiardi L. Commentary: three worlds collide: Berkson's bias, selection bias and collider bias. Int J Epidemiol 2014; 43:521-4..

Another potential limitation is the quality of information on cause of hospitalization, since it was reported retrospectively by the mother without confirmation on the medical record. To deal with this problem, all causes of hospitalization were classified in major groups with similar signs and symptoms, according to ICD-10 chapters.

Changes in Brazil’s economy in the last decade were marked by a combination of economic growth and improved income distribution. The improvement was due to access to a public conditional cash transfer policy and growth in the job supply, among other reasons, manifested as better quality of life for families, an improvement that was more pronounced in the poorest areas of the country ³⁹39. Victora CG, Aquino EM, Carmo Leal M, Monteiro CA, Barros FC, Szwarcwald CL. Maternal and child health in Brazil: progress and challenges. Lancet 2011; 377:1863-76.. There were changes in families’ income in the 2004 cohort over the course of the study period, especially in the poorest quintiles, with more stability among the wealthier. Some 62% of the children in the lowest income quintile at birth remained in the lower quintiles at 6 years. Meanwhile, 82% of the children in the highest income quintiles at birth remained in the upper quintiles at 6 years.

Skin color of the mother and child was assessed by the interviewer rather than by self-report. The effect of the chosen data collection procedure on racial classification is controversial, and various researchers have suggested that it has limited validity and reliability ⁴⁰40. Bastos JL, Peres MA, Peres KG, Dumith SC, Gigante DP. Socioeconomic differences between self- and interviewer-classification of color/race. Rev Saúde Pública 2008; 42:324-34.. If this type of classification error occurred, it was not differential in relation to the outcome, since the interviewers had no way of knowing who had been hospitalized previously. We would thus expect that any associations would be biased toward nullity (non-association).

Despite these limitations, this was a study with a prospective population-based longitudinal design, with a low percentage of losses to follow-up, in which the children were followed since birth, allowing to study the hospitalization rate and to follow its evolution over time. This deserves emphasis, since there is a relative scarcity of similar studies in middle-income countries like Brazil.

In summary, the study’s results confirm that the early childhood hospitalization rate remains quite high in Pelotas, Rio Grande do Sul State, especially in the first year of life, when compared to similar studies in other states of Brazil. Respiratory diseases are still the leading cause of hospitalization in childhood. Previous history of hospitalization was a strong predictor of hospitalization, and various perinatal biological and environmental factors were associated with hospitalization in early childhood. Increasing access to adequate treatments and improving childcare are necessary interventions to reduce childhood hospitalization. These findings are useful for planning health measures for children, thereby contributing to the improvement of health services.

Acknowledgments

The article analyzed data from the study Pelotas Birth Cohort, 2004, conducted by the Graduate Studies Program in Epidemiology at Pelotas Federal University, with the support of the Brazilian Association of Collective Health (ABRASCO). From 2009 to 2013, the 2004 birth cohort study was financed by the Wellcome Trust. Previous phases of the study were financed by the World Health Organization (WHO), the Program to Support Centers of Excellence (PRONEX), Brazilian National Research Council (CNPq), Brazilian Ministry of Health, and Children’s Pastoral.

References

Publication Dates

History