Using machine learning classifiers to assist healthcare-related decisions: classification of electronic patient records.

Surveillance Levels (SLs) are categories for medical patients (used in Brazil) that represent different types of medical recommendations. SLs are defined according to risk factors and the medical and developmental history of patients. Each SL is associated with specific educational and clinical measures. The objective of the present paper was to verify computer-aided, automatic assignment of SLs. The present paper proposes a computer-aided approach for automatic recommendation of SLs. The approach is based on the classification of information from patient electronic records. For this purpose, a software architecture composed of three layers was developed. The architecture is formed by a classification layer that includes a linguistic module and machine learning classification modules. The classification layer allows for the use of different classification methods, including the use of preprocessed, normalized language data drawn from the linguistic module. We report the verification and validation of the software architecture in a Brazilian pediatric healthcare institution. The results indicate that selection of attributes can have a great effect on the performance of the system. Nonetheless, our automatic recommendation of surveillance level can still benefit from improvements in processing procedures when the linguistic module is applied prior to classification. Results from our efforts can be applied to different types of medical systems. The results of systems supported by the framework presented in this paper may be used by healthcare and governmental institutions to improve healthcare services in terms of establishing preventive measures and alerting authorities about the possibility of an epidemic.

Can the insulin sensitivity index (ISI) in association with insulin-like growth factor binding protein-1 identify insulin resistance early in overweight children?

UNLABELLED: Association between insulin resistance (IR) and non-alcoholic fatty liver disease (NAFLD) has been reported. This prompted us to evaluate the power of the insulin sensitivity index (ISI) in association with IGFBP-1 to identify IR early in obese children/adolescents. OGTT was performed in 34 obese/overweight children/adolescents. Glucose, insulin and IGFBP-1 were measured in serum samples and ISI was calculated. Considering the presence of three or more risk factors for IR as a criterion for IR, ISI < 4.6 showed 87.5% sensitivity and 94.5% specificity in diagnosing IR. IGFBP-1 was lower in the group with ISI < 4.6 (p < 0.01). In this group, three patients had higher than expected IGFBP-1, suggesting hepatic IR, while three patients with ISI > 4.6 showed very low IGFBP-1 levels. CONCLUSION: ISI < 4.6 is a good indicator of early peripheral IR and, associated with IGFBP-1, can identify increased risk of hepatic IR. Low IGFBP-1 levels among non-IR children may indicate increased portal insulin levels.

Zinc serum levels and their association with vitamin A deficiency in preschool children.

OBJECTIVES: To identify the prevalence of zinc deficiency in a population with high prevalence of vitamin A deficiency; to verify whether zinc deficiency is associated with vitamin A deficiency in the population studied; to verify risk factors for zinc deficiency (sex, age, diarrhea and fever). METHOD: Cross-sectional study of 182 healthy children aged > or = 24 months and < 72 months. Peripheral blood samples were obtained from fasting children to determine zinc serum levels. Information about presence of diarrhea and/or fever during the 15 days preceding the study was also obtained. Vitamin A deficiency was identified by a serum 30-day dose-response test (+S30DR). RESULTS: Of the children studied, 0.5% (1/182) presented zinc serum levels < 65 microg/dL; however, 74.7% (136/182) of them had vitamin A deficiency. Zinc serum levels were not correlated with retinol serum levels. Zinc serum levels were not changed by previous diarrhea and/or fever. There was no difference in zinc levels between boys and girls. Children aged between > or = 48 and < 60 months tended to have lower zinc serum levels than children of other ages. CONCLUSION: Zinc deficiency prevalence was low and did not represent a risk factor for vitamin A deficiency. Children aged between > or = 48 and < 60 months tended to have lower zinc serum levels than children of other ages. Zinc serum levels were not changed by previous diarrhea and/or fever.

Nível sérico de zinco e sua associação com deficiência de vitamina A em crianças pré-escolares / Zinc serum levels and their association with vitamin A deficiency in preschool children

OBJETIVOS: Identificar a prevalência da deficiência de zinco em uma população com alta prevalência de deficiência de vitamina A; verificar se a deficiência de zinco apresenta associação com deficiência de vitamina A; verificar a influência de alguns fatores de risco (idade, sexo, diarréia e febre) na gênese da deficiência de zinco. MÉTODOS: Estudo transversal com 182 crianças saudáveis com idades > 24 meses e < 72 meses. Obtiveram-se amostras de sangue periférico em jejum para dosagem dos níveis séricos de zinco. Também foram obtidas informações sobre a presença de diarréia e/ou febre nos 15 dias precedentes à pesquisa. A identificação da deficiência de vitamina A foi realizada através do teste de dose-resposta após 30 dias a uma suplementação com vitamina A -+S30DR. RESULTADOS: Das crianças estudadas, 0,5 por cento (1/182) apresentou nível sérico de zinco < 65 µg/dL; entretanto, 74,7 por cento (136/182) apresentavam deficiência de vitamina A. Não houve correlação entre os níveis séricos de zinco e os de retinol. Episódios febris e/ou diarréicos não alteraram os níveis de zinco. Não houve também diferença entre os níveis de zinco entre os sexos. As crianças com idade entre > 48 e < 60 meses de idade tenderam a apresentar menores níveis de zinco do que as demais faixas etárias. CONCLUSÃO: A prevalência de deficiência de zinco foi baixa e não representou fator de risco para deficiência de vitamina A. As crianças com idades entre > 48 e < 60 meses tenderam a apresentar menores médias de nível sérico de zinco do que as demais faixas etárias. Febre e/ou diarréia prévios ao estudo não alteraram os níveis séricos de zinco.

OBJECTIVES: To identify the prevalence of zinc deficiency in a population with high prevalence of vitamin A deficiency; to verify whether zinc deficiency is associated with vitamin A deficiency in the population studied; to verify risk factors for zinc deficiency (sex, age, diarrhea and fever). METHOD: Cross-sectional study of 182 healthy children aged > 24 months and < 72 months. Peripheral blood samples were obtained from fasting children to determine zinc serum levels. Information about presence of diarrhea and/or fever during the 15 days preceding the study was also obtained. Vitamin A deficiency was identified by a serum 30-day dose-response test. RESULTS:Of the children studied, 0.5 percent (1/182) presented zinc serum levels < 65 µg/dL; however, 74.7 percent (136/182) of them had vitamin A deficiency. Zinc serum levels were not correlated with retinol serum levels. Zinc serum levels were not changed by previous diarrhea and/or fever. There was no difference in zinc levels between boys and girls. Children aged between > 48 and < 60 months tended to have lower zinc serum levels than children of other ages. CONCLUSION: Zinc deficiency prevalence was low and did not represent a risk factor for vitamin A deficiency. Children aged between > 48 and < 60 months tended to have lower zinc serum levels than children of other ages. Zinc serum levels were not changed by previous diarrhea and/or fever.

[Prevalence of iron deficiency and its association with vitamin A deficiency in preschool children]. / Prevalência da carência de ferro e sua associação com a deficiência de vitamina A em pré-escolares.

OBJECTIVES: To identify the prevalence of iron deficiency in the population studied, as well as verifying if such deprivation is associated with vitamin A deficiency. METHODS: One hundred seventy-nine children, > or = 24 months and < 72 months of age, with no diarrhea and/or fever at collection were studied. Vitamin A deficiency identification was carried out through serum 30-day dose-response test. Samples of peripheral blood from fasting children was obtained for hemoglobin counts, serum iron, and unsaturated iron binding capacity assays. Information about the presence of diarrhea and/or fever during the 15 days preceding the study was also obtained. RESULTS: 35.8% (64/179) of the children presented iron deficiency and 75.4% (135/179), vitamin A deficiency. 29.1% (52/179) of the children presented both iron and vitamin A deficiencies. Iron deficiency was not associated with vitamin A deficiency. A separate analysis for each hematimetric index also demonstrated no significant difference between children with or without vitamin A deficiency. Children aged 24 to 36 months presented significantly higher prevalence rates of iron deficiency (p = 0.0005) as did children with diarrhea and/or fever during the 15 days preceding the study (p = 0.003). CONCLUSIONS: Although iron deficiency was not associated with vitamin A deficiency, high rates of both deficiencies were exhibited in a "healthy" population with low malnutrition indices. Such situations are known as "hidden hunger". Younger children presented a higher risk of iron deficiency as did children with diarrhea and/or fever during the 15 days preceding the study.

Prevalência da carência de ferro e sua associação com a deficiência de vitamina A em pré-escolares / Prevalence of iron deficiency and its association with vitamin A deficiency in preschool children

OBJETIVO: Estimar a carência de ferro na população estudada e verificar se isso se associa à falta de vitamina A. MÉTODOS: Foram estudadas 179 crianças com idade > 24 meses e < 72 meses, sem diarréia e/ou febre no momento da coleta. A identificação da deficiência de vitamina A foi realizada através do teste de resposta sérica de 30 dias. Foram obtidas amostras de sangue periférico em jejum para dosagem dos níveis de hemoglobina, ferro sérico e capacidade latente de fixação de ferro, além de informação sobre a presença de diarréia ou febre nos 15 dias precedentes à pesquisa. RESULTADOS: 35,8 por cento (64/179) das crianças apresentaram carência de ferro; 75,4 por cento (135/179), deficiência de vitamina A; e 29,1 por cento (52/179) apresentaram carência de ferro e deficiência de vitamina A, concomitantemente. A carência de ferro não apresentou associação com a deficiência de vitamina A, nem tampouco com cada índice hematimétrico analisado separadamente. As crianças entre 24 e 36 meses de idade apresentaram significativamente maior prevalência da carência de ferro (p = 0,0005), como também as crianças com episódios febris ou diarréicos nos 15 dias precedentes à entrada no estudo (p = 0,003). CONCLUSÕES: Apesar de a carência de ferro não apresentar associação à deficiência de vitamina A, ambas as carências apresentaram prevalências elevadas em uma população saudável e com baixo índice de desnutrição. Tal situação é conhecida como fome oculta. As crianças mais jovens apresentaram maior risco de portar carência de ferro, como também as crianças com episódios febris ou diarréicos nos 15 dias precedentes à entrada no estudo.

Serum insulin-like growth factor (IGF)-I and IGF-binding protein-3 in girls with premature thelarche.

Premature thelarche (PT) is characterised by precocious breast development without any other sign of puberty, normal height velocity (HV) and normal bone maturation, while girls with central precocious puberty (CPP) show increased HV, bone maturation and increased serum IGF-I and IGFBP-3 levels. This prompted us to study serum IGF-I and IGFBP-3 concentrations in girls with PT. Thirty-nine girls with premature breast development were studied and classified as PT or CPP according to clinical and laboratory evaluation. Normal prepubertal and pubertal girls were studied as controls. Serum IGF-I and IGFBP-3 were determined in all girls by IRMA. IGF-I levels in PT (155 +/- 61 microg/l) were lower than in CPP (337 +/- 149 microg/l) or late-pubertal controls (355 +/- 84 microg/l) and similar to those found in prepubertal (113 +/- 72 microg/l) and early-pubertal (222 +/- 81 microg/l) girls. Considering the SDS of IGF-I for chronological age (CA), the values observed in PT were in an intermediate position between CPP and prepubertal controls and statistically similar to those observed in CPP and prepubertal girls. IGFBP-3 levels in PT (2.1 +/- 0.5 mg/l) were similar to those found in CPP (2.5 +/- 0.8 mg/l), but only the latter were higher than in prepubertal girls (1.9 +/- 0.9 mg/l). IGF-I/IGFBP-3 molar ratios in PT were in an intermediate position between CPP and prepubertal controls. In conclusion, IGF-I and IGF-I/IGFBP-3 values in PT are intermediate between those observed in prepubertal children and in CPP, suggesting that PT could be a very early stage of puberty with slight but real changes in the GH-IGF axis.

Growth of exclusively breast-fed infants from a poor urban population

The growth of two groups of infants were evaluated, one of them exclusively breast-fed (105 infants) and the other exclusively bottle-fed (61 infants), and compared with one another and with international standards (NCHS). All infants were evaluated by anthropometry at 1, 2, 3, 4, 5 and 6 months of age. A fourth order polynomial was adopted for each infant and for each anthropometric measurement in order to estimate individual growth, and the 5th, 50th and 95th percentiles for weight and length were obtained. For the age of 6 months, the weights of breast-fed boys and girls were always statistically equal to or higher than those of infants fed cow's milk or those of NCHS standards. Breast-fed boys presented significantly longer length than bottle-fed boys but shorter than NCHS standards, and breast-fed girls presented significantly shorter length than both bottle-fed girls and than NCHS standards. The greater weight of exclusively breast-fed infants when compared to NCHS standards at six months of age, which differentiates the present study from several other ones carried out in developing countries, was probably due to the association of the beneficial effects of breast-feeding with those of pediatric follow-up.

Prevalência de asma em escolares / Prevalence of asthma in schoolchildren

Objetivos: avaliar a prevalência dos sintomas e da doença de asma em crianças, em razão da alta frequência pediátricas com sintomas sugestivos da doença e da falta de dados de prevalência nesta população. Métodos0: num estudo transversal, foram avaliados 2.735 escolares na faixa etária de 6-7 anos (crianças) e 3.509 na de 13-14 anos (adolescentes), escolhidos por amostragem aleatória. Os dados foram coletados em 1998-99 utilizando-se o questionário escrito traduzido do Internacional Stady of Asthma and Allergies in Children (ISAAC), previamente testado e avaliado. O questionário dirigido às crianças foi respondido pelos seus responsáveis, enquanto os adolescentes responderam em sala de aula. Os dados foram transcritos e analisados pelo programa EPI-info. Resultados: taxa de devolução: 73,2 por cento (6-7 anos) e 94 por cento (13-14 anos). Prevalência dos sintomas nas crianças e adolescentes: sibilos algumas vez, 46,7 por cento e 44,3 por cento (p<0,05); sibilos nos últimos 12 meses, 22,7 por cento e 21,2 por cento; fala prejudicada por sibilos, 4,8 por cento e 4,0 por cento, sibilos aos exercícios, 6,3 por cento e 18,2 por cento (p<0,05); tosse seca noturna, 38,5 por cento e 45,8 por cento (p<0,05), asma (bronquite) alguma vez 28,2 por cento e 26,4 por cento (p=0,11). Conclusões: a prevalência dos sintomas e da doença de asma encontrados nos dois grupos está entre as maiores relatadas no Brasil, sendo compatíveis com as maiores taxas encontradas no mundo com o questionário do ISAAC. A alta prevalência dos sintomas sugere uma elevada prevalência da doença, o que a torna um importante problema de saúde pública nesta região