Diagnosis of urinary tract infection in infants under 3 months with fever without a source: reliability of urinalysis and urine culture.

BACKGROUND: Urinary tract infection (UTI) is infants' most common serious bacterial infection. This study aimed to investigate the reliability of urianalysis (UA) to predict UTI, to specify the colony forming units (CFU)/ml threshold for diagnosis, and to identify variables that help suspect bacteremia in infants under 3 months with UTI. METHODS: We reviewed clinical records of children under 3 months hospitalized for a fever without source and recorded age, sex, days of fever pre-consultation, temperature and severity at admission, discharge diagnoses, laboratory tests, and treatments. According to the discharge diagnosis, we divided them into UTIs (-) and (+) with or without bacteremia. RESULTS: A total of 467 infants were admitted: 334 with UTI and 133 without UTI. In UTIs (+), the pyuria had a sensitivity of 95.8% and bacteria (+) 88.3%; specificity was high, especially for nitrites (96.2%) and bacteria (+) (92.5%). Positive predictive value (PPV) for nitrites was 95.9%, for bacteria 96.7%, and oyuria 92.5%. Escherichia coli was present in 83.8% of urine and 87% of blood cultures. UTIs with bacteremia had inflammatory urinalysis, urine culture > 100,000 CFU/ml, and higher percentage of C reactive protein (CRP) > 50 mg (p= 0.002); 94.6% of the urine culture had > 50,000 CFU. CONCLUSIONS: The pyuria and bacteria (+) in urine obtained by catheterization predict UTI. The cut-off point for diagnosis was ≥ 50,000 CFU/ml. No variables to suspect bacteremia were identified in this study.

INTRODUCCIÓN: La infección del tracto urinario (ITU) es una infección bacteriana grave frecuente en lactantes. El objetivo de este trabajo fue investigar la fiabilidad del análisis de orina (AO) para predecirla, precisar el umbral de unidades formadoras de colonias (UFC)/ml para el diagnóstico y buscar variables que ayuden a sospechar de bacteriemia en lactantes menores de 3 meses con ITU. MÉTODOS: Se revisaron fichas clínicas de lactantes menores de 3 meses hospitalizados por fiebre sin foco evidente, registrando edad, sexo, días de fiebre preconsulta, temperatura y gravedad al ingreso, diagnósticos de egreso, exámenes de laboratorio y tratamientos. Según diagnóstico de egreso, se separaron en ITU (-) y (+), con o sin bacteriemia. RESULTADOS: Ingresaron 467 lactantes: 334 con ITU y 133 sin ITU. En ITU (+), la sensibilidad de la piuria fue de 95.8% y bacterias (+) 88.3%; la especificidad fue alta para nitritos (96.2%) y bacterias (+) (92.5%). El valor predictivo positivo (VPP) fue de 95.9% para nitritos, 96.7% para bacterias y 92.5% para piuria. Escherichia coli se encontró en el 83.8% de los urocultivos (UC) (+) y en el 87% de los hemocultivos (+). Las ITU con bacteriemia presentaron elementos inflamatorios, UC con ≥ 100,000 UFC/ml y mayor porcentaje de proteína C reactiva (PCR) > 50 mg/l (p= 0.002); el 94.6% de los UC (+) tuvo ≥ 50,000 UFC/ml. CONCLUSIONES: La piuria y bacterias (+) en el AO son excelentes para pronosticar ITU en orina obtenida con sonda vesical y el punto de corte para el diagnóstico debe ser ≥ 50,000 UFC/ml. No encontramos señales que ayudaran a sospechar ITU con bacteriemia.

Diagnosis of urinary tract infection in infants under 3 months with fever without a source: reliability of urinalysis and urine culture / Diagnóstico de infección del tracto urinario en lactantes menores de 3 meses con fiebre sin foco identificado: fiabilidad del análisis de orina y urocultivo

Abstract Background: Urinary tract infection (UTI) is infants´ most common serious bacterial infection. This study aimed to investigate the reliability of urianalysis (UA) to predict UTI, to specify the colony forming units (CFU)/ml threshold for diagnosis, and to identify variables that help suspect bacteremia in infants under 3 months with UTI. Methods: We reviewed clinical records of children under 3 months hospitalized for a fever without source and recorded age, sex, days of fever pre-consultation, temperature and severity at admission, discharge diagnoses, laboratory tests, and treatments. According to the discharge diagnosis, we divided them into UTIs (-) and (+) with or without bacteremia. Results: A total of 467 infants were admitted: 334 with UTI and 133 without UTI. In UTIs (+), the pyuria had a sensitivity of 95.8% and bacteria (+) 88.3%; specificity was high, especially for nitrites (96.2%) and bacteria (+) (92.5%). Positive predictive value (PPV) for nitrites was 95.9%, for bacteria 96.7%, and oyuria 92.5%. Escherichia coli was present in 83.8% of urine and 87% of blood cultures. UTIs with bacteremia had inflammatory urinalysis, urine culture > 100,000 CFU/ml, and higher percentage of C reactive protein (CRP) > 50 mg (p= 0.002); 94.6% of the urine culture had > 50,000 CFU. Conclusions: The pyuria and bacteria (+) in urine obtained by catheterization predict UTI. The cut-off point for diagnosis was ≥ 50,000 CFU/ml. No variables to suspect bacteremia were identified in this study.

Resumen Introducción: La infección del tracto urinario (ITU) es una infección bacteriana grave frecuente en lactantes. El objetivo de este trabajo fue investigar la fiabilidad del análisis de orina (AO) para predecirla, precisar el umbral de unidades formadoras de colonias (UFC)/ml para el diagnóstico y buscar variables que ayuden a sospechar de bacteriemia en lactantes menores de 3 meses con ITU. Métodos: Se revisaron fichas clínicas de lactantes menores de 3 meses hospitalizados por fiebre sin foco evidente, registrando edad, sexo, días de fiebre preconsulta, temperatura y gravedad al ingreso, diagnósticos de egreso, exámenes de laboratorio y tratamientos. Según diagnóstico de egreso, se separaron en ITU (-) y (+), con o sin bacteriemia. Resultados: Ingresaron 467 lactantes: 334 con ITU y 133 sin ITU. En ITU (+), la sensibilidad de la piuria fue de 95.8% y bacterias (+) 88.3%; la especificidad fue alta para nitritos (96.2%) y bacterias (+) (92.5%). El valor predictivo positivo (VPP) fue de 95.9% para nitritos, 96.7% para bacterias y 92.5% para piuria. Escherichia coli se encontró en el 83.8% de los urocultivos (UC) (+) y en el 87% de los hemocultivos (+). Las ITU con bacteriemia presentaron elementos inflamatorios, UC con ≥ 100,000 UFC/ml y mayor porcentaje de proteína C reactiva (PCR) > 50 mg/l (p= 0.002); el 94.6% de los UC (+) tuvo ≥ 50,000 UFC/ml. Conclusiones: La piuria y bacterias (+) en el AO son excelentes para pronosticar ITU en orina obtenida con sonda vesical y el punto de corte para el diagnóstico debe ser ≥ 50,000 UFC/ml. No encontramos señales que ayudaran a sospechar ITU con bacteriemia.

[Febrile syndrome in children younger than 29 days]. / Síndrome febril en niños menores de 29 días.

INTRODUCTION: Acute fever of unknown origin (FUO) in children under 29 days is a worrying situation because of the risk of serious bacterial infection (SBI). OBJECTIVE: to study the main clinical and laboratory characteristics of a group of hospitalized children under 29 days with diagnosis of FUO. PATIENTS AND METHOD: Retrospective study of children under 29 days hospitalized due to FUO. The clinical records of the patients were reviewed, recording age, sex, history of fever before consultation, temperature at admission, estimated severity at admission and discharge, discharge diagnoses, laboratory tests, and indicated treatments. Patients were classified according to the severity of the discharge diagnosis, as severe (S) and non-severe (NS). The inclusion criteria were term newborn, age less than 29 days, fe ver > 38°C registered at home or admission, and history of < 4 days. RESULTS: 468 children with FUO were admitted. Concordance between severity at admission and discharge was low (Kappa = 0.125; p = 0.0007). 26.1% of children were S and 73.9% NS. In the S group, urinary tract infection domínate (70.5%) and in the NS, FUO (67.6%). The cut-off levels for leukocytes/mm3, C-reactive protein, and neutrophils/mm3 showed negative predictive values to rule out severe bacterial infection. Conclu sions: Most of the newborns presented mild severity at admission, but 24% of them had SBI, thus hospitalization and close clinical observation are always necessary. Laboratory tests, such as CRP, white blood cell and neutrophils count are not good predictors of SBI. Early treatment with antibio tics for patients who meet the low-risk criteria is debatable.