Heitor Pons Leite

Malnutrition as an independent predictor of clinical outcome in critically ill children

OBJECTIVE To determine the nutritional status of a cohort of children admitted to a pediatric intensive care unit (ICU) and to assess the effect of malnutrition as an independent risk factor affecting outcome in this patient group. METHODS In a prospective cohort study, 385 children admitted to the ICU of a teaching hospital over a 2-y period were assessed for nutritional status at admission and clinical outcome. The outcome variables were 30-d mortality, length of ICU stay, and length of mechanical ventilation. Potential exposure variables were gender, age, diagnosis (clinical versus surgical), septic shock, malnutrition, and scores on the Pediatric Index of Mortality and Pediatric Logistic Organ Dysfunction. Nutritional status was determined using z scores of weight for age, height for age, and body mass index, based on the World Health Organization child growth standards. Patients with z score < -2 of anthropometric indexes were considered malnourished. RESULTS 175 patients (45.5%) were malnourished on admission. Sixteen patients of the malnourished group (9.14%) and 25 patients (11.9%) of the non-malnourished group died. Malnutrition was associated with greater length of mechanical ventilation and length of ICU stay, but not with mortality on univariate analysis. Malnutrition was associated with greater length of ventilation on the multiple logistic regression model (OR 1.76, 95%; CI 1.08-2.88; P = 0.024). CONCLUSION Malnutrition is common among children admitted to an ICU. This factor was not a predictor of mortality but showed independent association with length of mechanical ventilation.

Outcomes in a pediatric intensive care unit before and after the implementation of a nutrition support team.

BACKGROUND We evaluated the effect of parenteral nutrition (PN) and enteral nutrition (EN) on in-pediatric intensive care unit (PICU) mortality before and after a continuous education program in nutrition support that leads to implementation of a nutrition support team (NST). METHODS We used a historical cohort study of infants hospitalized for >72 hours at the PICU from 1992 to 2003. Five periods were selected (P1 to P5), considering the modifications incorporated into the program: P1, without intervention; P2, basic themes and original articles discussion; P3, clinical and nursing staff participation; P4, clinical visits; P5, NST. The samples were compared in terms of sex, age, admitting service (ie, medical vs surgical), prognostic index of mortality, length of stay (LOS), duration of mechanical ventilation, in-PICU mortality rate, and percentage of time receiving EN and PN for each patient. Bi- and multivariate analyses were performed. Statistical significance was set at 0.05 level. RESULTS Progressive increase was observed in EN use ( p = .0001), median values for which were 25% in P1 and rose to 67% by P5 in medical patients; there was no significant difference in surgical patients. A reduction was observed in PN use; in P1 medians were 73% and 69% for medical and surgical patients respectively, and decreased to 0% in P5 for both groups ( p = .0001). There was significant reduction in-PICU mortality rate during P4 and P5 among medical patients ( p < .001). The risk of death was 83% lower in patients that received EN for >50% of LOS (odds ratio, 0.17; confidence interval, 0.066-0.412; p = .000). CONCLUSIONS The program motivated an increase in EN and a decrease in PN use, mainly after implementation of NST and reduced in-PICU mortality rate.

Low blood thiamine concentrations in children upon admission to the intensive care unit: risk factors and prognostic significance

BACKGROUND Thiamine deficiency has been associated with poorer clinical outcomes. Early recognition of thiamine deficiency is difficult in critically ill patients because clinical signs are nonspecific. OBJECTIVE We determined the prevalence of and identified risk factors associated with low blood thiamine concentrations upon admission of children to a pediatric intensive care unit and evaluated this condition as a predictor of clinical outcomes. DESIGN A prospective cohort study was conducted in 202 children who had whole-blood thiamin concentrations assessed by HPLC upon admission to the intensive care unit. The following independent variables for thiamine deficiency were analyzed: age, sex, nutritional status, clinical severity scores upon admission (ie, the revised Pediatric Index of Mortality and Pediatric Logistic Organ Dysfunction score), systemic inflammatory response measured by C-reactive protein serum concentrations, severe sepsis or septic shock, heart failure, and cardiac surgery. The dependent variables in the outcome analyses were mortality, length of stay, and time on mechanical ventilation. RESULTS Low blood thiamine concentrations upon admission were detected in 57 patients (28.2%) and were shown to be independently associated with C-reactive protein concentrations >20 mg/dL (odds ratio: 2.17; 95% CI: 1.13, 4.17; P = 0.02) but not with malnutrition. No significant association was shown between low blood thiamine concentrations upon admission and outcome variables. CONCLUSIONS The incidence of low blood thiamine concentrations upon admission was high. Of the risk factors examined, only the magnitude of the systemic inflammatory response showed an independent association with this event. The association between thiamine deficiency upon admission and prognosis requires further investigation.

Enteral nutrition in critically ill children: are prescription and delivery according to their energy requirements?

BACKGROUND The purpose of this study was to compare the differences between prescribed and delivered energy among critically ill children and to identify the factors that impede the optimal delivery of enteral nutrition in the first 5 days of nutrition support. METHODS In a prospective cohort study, we evaluated 55 critically ill children aged 8.2 ± 11.4 months (0-162.3 months), who were fed for ≥2 days through a gastric or postpyloric tube. The patients were followed from admission until day 10 of enteral nutrition. Prescribed and delivered energy were recorded daily and compared with each other and with the estimated basal metabolic rate (BMR). The Paediatric Index of Mortality 2 (PIM 2) was used to estimate illness severity. RESULTS The ratio of delivered:required energy was <90% in 55.7% of the enteral nutrition days. Low prescription was the main reason for not achieving the energy goal in the first 5 days of enteral nutrition. Discrepancies between prescribed and delivered: energy were attributable to interruptions in feeding caused by clinical instability, airway management, radiologic and surgical procedures, and accidental feeding tube removal. The other factors associated with the delivery of less than required energy were PIM 2 ≥15%, gastrointestinal complications, dialysis, and use of α-adrenergic vasoactive drugs. The latter was the only variable in multivariate analysis that was associated with not ultimately achieving energy goal. CONCLUSIONS The prescription and delivery of energy were not adequate in >50% of enteral nutrition days. The gap between the effective administration and energy requirements can be explained by both underprescription and underdelivery. Administration of vasoactive drugs was the only variable independently associated with a low energy supply.

Hypophosphatemia in critically ill children: prevalence and associated risk factors.

Background: Hypophosphatemia is a disorder with potential complications and is often unrecognized in critically ill patients. Aims: To identify the prevalence of hypophosphatemia and risk factors associated to this disorder in critically ill children. Methods: In a prospective cohort study, 82 children admitted consecutively to a pediatric intensive care unit (ICU) were monitored regarding phosphorus serum levels during the first 10 days of admission. The following variables were analyzed as independent for hypophosphatemia: age, gender, diagnosis at admission, malnutrition, phosphorus intake, clinical severity score at admission (pediatric index of mortality 2) and daily scores (Pediatric Logistic Organ Dysfunction), sepsis, use of dopamine, furosemide and steroids, starvation period, and refeeding. Children with a z score of less than −2 of expected weight for age or body mass index (National Center for Health Statistics, 2000) were considered malnourished. Variables significantly associated with hypophosphatemia by bivariate analysis (p < 0.1) were included in a multiple logistic regression model. Results: The rate of hypophosphatemia was 61% during the first 10 days of pediatric ICU stay, and 12 patients developed hypophosphatemia during the study period. Malnutrition was present in 39.1% of patients, and the sera phosphorus concentration was significantly lower in malnourished than in well-nourished children (2.6 ± 0.7 mg/dL vs. 3.5 ± 0.8 mg/dL, p = 0.01). The multiple logistic regression model indicated the diagnosis of acute respiratory disease (odds ratio: 3.22; confidence interval: 1.03–10.1; p = 0.04), use of dopamine (odds ratio: 8.65; confidence interval: 1.58–47.3; p = 0.01), and malnutrition (odds ratio: 3.96; confidence interval: 1.19–13.3; p = 0.02) as independent risk factors for hypophosphatemia. None of the other potential risk factors discriminated for hypophosphatemia. Conclusions: Hypophosphatemia was common in the first 10 days of ICU hospitalization and was associated with the diagnosis of respiratory disease, use of dopamine, and malnutrition. These factors should be taken into account during clinical follow up of critically ill children, especially when these conditions are found together.

Hypophosphatemia in critically ill children

The purpose of this paper is to review clinical studies on hypophosphatemia in pediatric intensive care unit patients with a view to verifying prevalence and risk factors associated with this disorder. We searched the computerized bibliographic databases Medline, Embase, Cochrane Library, and LILACS to identify eligible studies. Search terms included critically ill, pediatric intensive care, trauma, sepsis, infectious diseases, malnutrition, inflammatory response, surgery, starvation, respiratory failure, diuretic, steroid, antiacid therapy, mechanical ventilation. The search period covered those clinical trials published from January 1990 to January 2004. Studies concerning endocrinological disorders, genetic syndromes, rickets, renal diseases, anorexia nervosa, alcohol abuse, and prematurity were not included in this review. Out of 27 studies retrieved, only 8 involved pediatric patients, and most of these were case reports. One clinical trial and one retrospective study were identified. The prevalence of hypophosphatemia exceeded 50%. The commonly associated factors in most patients with hypophosphatemia were refeeding syndrome, malnutrition, sepsis, trauma, and diuretic and steroid therapy. Given the high prevalence, clinical manifestations, and multiple risk factors, the early identification of this disorder in critically ill children is crucial for adequate replacement therapy and also to avoid complications.

Malnutrition May Worsen the Prognosis of Critically Ill Children With Hyperglycemia and Hypoglycemia

OBJECTIVES To determine whether hyperglycemia and hypoglycemia are associated with higher mortality, longer length of intensive care unit (ICU) stay, and fewer ventilator-free days in critically ill children while taking into account the clinical severity and nutrition status. PATIENTS AND METHODS A prospective observational cohort study was conducted on 221 children admitted to the ICU. Blood glucose levels were analyzed in the first 72 hours. Potential exposure variables for adverse prognosis included hyperglycemia (blood glucose >150 mg/dL), hypoglycemia (blood glucose ≤60 mg/dL), age <1 year, sex, nutrition status, the revised Pediatric Index of Mortality (PIM 2), and the Pediatric Logistic Organ Dysfunction (PELOD). RESULTS Of the patients, 47.1% were malnourished. Controlling for nutrition status, both hyperglycemia and hypoglycemia increased the risk of mortality in the malnourished patients compared with the well-nourished ones. Adjusting for clinical severity, the odds ratio of mortality was higher in malnourished patients with hyperglycemia (odds ratio [OR], 3.98; 95% confidence interval [CI], 1.14-13.94; P = .03), whereas no significant associations were detected in the well-nourished patients. After controlling for nutrition status, hypoglycemia was associated with longer length of ICU stay (OR, 6.5; 95% CI, 1.30-32.57; P < .01) and fewer ventilator-free days (OR, 4.11; 95% CI, 1.26-13.40; P < .01) only in the malnourished group of patients. CONCLUSIONS Compared with the well nourished, malnourished patients with hyperglycemia are at a greater risk of mortality, independent of clinical severity. Hypoglycemia was shown to be associated with mortality, longer length of ICU stay, and fewer ventilator-free days only in malnourished patients.

Nutritional assessment and surgical risk makers in children submitted to cardiac surgery

In order to assess the nutritional status of children with heart diseases and to evaluate nutritional parameters for predicting postoperative complications, 50 children undergoing to cardiac surgery and classified in high and low surgical risk prospectively evaluated. Assessment parameters included anthropometry and plasma proteins albumin, transferrin and prealbumin. The nutritional classification according to Waterlows modified criteria showed a high prevalence of malnutrition in the population studied (78%). The measures of arm circumference when located below the 5th percentile showed a significant association with general postoperative complications in the high risk group (arm circumference, p = 0.0019; arm muscle circumference, p = 0.0419). The percentage of weight per height, serum albumin and transferrin has not played a prognostic role concerning postoperative morbidity. The mean value of prealbumin was significantly lower in high risk group patients developing postoperative infections (p < 0.01) compared to those who did not. The sensitivity-specificity analysis of prealbumin as risk indicator for postoperative infection was 87.5% and 59% respectively. The nutritional risk classification seems to be a good way to identify the subgroups of children with additional postoperative surgical risk. However, more specific and sensitive tests are desirable to provide an individual identification of these children.

Hypophosphatemia in children hospitalized within an intensive care unit.

The aims of this study were to estimate the occurrence of hypophosphatemia and to identify potential risk factors and outcome measures associated with this disturbance in children admitted to a pediatric intensive care unit. Data concerning 42 children admitted consecutively to 1 pediatric intensive care unit over a 1-year period were examined. Serum phosphorus levels were measured on the third day of admission, where levels below 3.8 mg/dL were considered indicative of hypophosphatemia. Hypophosphatemia was found in 32 children (76%), and there was a significant association between this disturbance and malnutrition ( P = .04). Of the potential risk factors such as sepsis, diuretic/steroid therapy, starvation (over 3 days), and Pediatric Index of Mortality, none discriminated for hypophosphatemia. There were no associations between hypophosphatemia and mortality, length of stay in the pediatric intensive care unit, or time on mechanical lung ventilation. Hypophosphatemia was a common finding in critically ill children and was associated with malnutrition.

Interdisciplinary approach improves nutritional status of children with heart diseases.

OBJECTIVES We evaluated the role of an interdisciplinary team in improving the nutrient intake and nutritional status of children with heart diseases. METHODS Thirty-five children 10.0 +/- 7.5 mo old were studied over a 6-mo period with regard to anthropometric measurements and dietary intake. RESULTS On admission, malnutrition prevalence was 57% and vitamin and mineral intakes were below estimated average requirement levels (dietary reference intakes) in the majority of patients. Stunting and wasting were more frequent among patients with pulmonary hypertension and/or uncompensated congestive heart failure. Follow-up analyses showed increased Z scores of weight for height (-1.17 +/- 1.03 versus -0.32 +/- 1.08, P < 0.01) and height for age (-1.09 +/- 0.96 versus -0.51 +/- 1.36, P < 0.01). In patients with pulmonary hypertension or uncompensated congestive heart failure, the height-for-age index remained unchanged. Energy intake did not differ (112.0 +/- 20.4 and 119.0 +/- 18.0 kcal.kg(-1).d(-1)) and significant increases were seen in intakes of micronutrients and minerals over the study period. CONCLUSION Intervention by an interdisciplinary team improved the nutrient intake and nutritional status of patients overall but was insufficient to improve growth in the subgroup with pulmonary hypertension or uncompensated congestive heart failure. Nutritional support should be made routine in the treatment of children with heart disease.