The integration of nutritional assessment into the complex routine of child hospitalization is essential, regardless of the severity and risks of morbidity and mortality associated with it.1 This is due to the possibility of identifying the presence of nutritional disorders, such as malnutrition and obesity, which may be associated with diseases and/or other unfavorable social conditions.2,3

Child malnutrition is considered the second most common cause of death among children under 5 years of age in developing countries. At this stage of life, malnutrition results in a loss of physical growth and neuropsychomotor development, and causes greater vulnerability to several comorbidities4 and, consequently, a greater risk of hospital admission.1

Hospitalization, in turn, is a risk factor for the development and/or worsening of malnutrition, as it often causes a loss in the nutritional status of the pediatric patient. This nutritional impairment can significantly affect morbidity and mortality rates, as well as length of hospital stay and hospitalization costs. Thus, there will be a lower hospital bed turnover rate, which consequently hinders the service to the population.1 Therefore, it is essential to perform the early nutritional assessment of all these patients at admission, to diagnose not only malnutrition but also the risk of nutritional status impairment of this group.5

According to Resolution No. 304/2003 of the Federal Council of Nutritionists, the implementation of nutritional assessment and diagnosis through anthropometric, biochemical, clinical, and dietary data favors the creation of an adequate nutritional care plan.6 Ideally, the nutritional assessment should be quick, practical, easy to apply, noninvasive, and capable of being performed at the bedside. Additionally, it is important that it can provide a result as fast as possible.7

Anthropometrics is a widely used science to assess, classify, and monitor nutritional status8 in several age groups, such as in childhood and adolescence, and it allows the perception of growth variations.1 In pediatric patients, the most often used anthropometric data for the analysis of nutritional status are weight, height, weight-to-height ratio, and body mass index (BMI),8 which are assessed according to the curves and reference values created by the World Health Organization (WHO) according to gender and age group; and the use of weight/age, height/age, weight/height, and BMI/age indexes.9,10 In Brazil, the Food and Nutrition Surveillance System (Sistema de Vigilância Alimentar e Nutricional [SISVAN]) suggests that the classification of nutritional status must be based on these curves.11

In addition to anthropometrics, screening methods have been widely used aiming at identifying malnutrition or the risk of developing it.8 In adults, the use of these methods is well established. However, in the case of pediatric patients, there is no consensus on the best technique for nutritional risk screening at the time of hospital admission and during hospitalization.8

There are some tools in the literature used for nutritional screening in pediatric patients. Overall, these tools correlate anthropometric data and patient history information, such as weight history, presence of disease, gastrointestinal symptoms, and food intake data. The best-known pediatric screening tools include the STAMP tool (Screening Tool for the Assessment of Malnutrition in Pediatrics),12 Pediatric Nutritional Risk Score,13 Pediatric Yorkhill Malnutrition Score,14 Strong Kids,15 and Subjective Global Nutritional Assessment (SGNA) for children.16 However, only the last two have been translated into Portuguese and validated for the Brazilian population.8

In 2007, the SGNA questionnaire was adapted for the pediatric age group by Secker and Jeejeebhoy.16 It was validated for this same group in the Brazilian population by Carniel et al., in 2015.17 In the field of child and adolescent health, for the purposes of nutritional assessment, the use of the WHO growth curves9,10 and the nutritional status classification proposed by them are already well known, and have been recognized regarding their importance in the nutritional diagnosis of this population. However, this methodology can identify malnutrition only after it is installed, i.e., it does not allow the detection of milder degrees of malnutrition and/or the risk of developing it.18 Thus, in the individual assessment, anthropometric data are more significant for follow-up, since they allow verifying whether the growth rate considered normal is maintained.19

For this reason, new nutritional assessment tools for this target population are being developed with the objective of improving existing resources. However, it is crucial that these new tools be tested to evaluate their applicability and agreement with other methods already used in the pediatric age group.18 Additionally, few studies have used the SGNA in pediatric patients, or compared their results with methods currently more widely used in the evaluation of children and adolescents, such as the WHO curves.

Thus, the objective of this study was to evaluate the agreement between the results of the SGNA questionnaire, adapted for children and adolescents of the Brazilian population, and the nutritional status assessment method through the WHO growth curves and classification in a pediatric hospital service.

The characterization of the studied sample is described in Table 1, which consisted of 61 hospitalized children, aged between 1 and 155 months. There was a predominance of males (52.4%) and children younger than 2 years (36%) in the sample, with 47.5% females. The median hospital length of stay was ten days.

Most patients who participated in the study were classified as well nourished by the SGNA (70.5%) and as normal weight according to the WHO (67%), while the lowest percentages were found for severely malnourished and very thin children, according to the SGNA (5%) and WHO (1.5%), respectively.

The frequencies of each classification (severely malnourished, moderately malnourished, and well-nourished) were analyzed for each of the assessed methods (Table 2), noting that there was a difference in the frequencies between them (p=0.001).

When assessing the agreement between the two tools, it was verified that all growth curves showed agreement with the SGNA, except for height/age. However, the degree of this agreement is classified as low (0.20–0.40)25 (Table 3).

In the analysis of the association between the nutritional assessment techniques, it was observed that the variables weight/height, BMI/age, and the final WHO classification were associated with SGNA (p<0.001) and also with the weight/age (p=0.002) and height/age (p=0.028) variables. The Kendall coefficient values showed associations ranging from low (0.20–0.39) to good (0.60–0.79),26 especially the weight/height variable, which showed a considerable (0.624) association with the SGNA (Table 4).

When concomitantly evaluating the patients according to the SGNA17 and the WHO,9,10 it was verified that most patients were classified as well-nourished and normal weight, respectively. Therefore, the greatest disagreements were found among the groups classified as moderately or severely malnourished, where the percentage of diagnosis of malnourished patients was higher in the SGNA than in the WHO curves. It was observed that of 11 patients classified as malnourished by the subjective assessment (one severely malnourished and ten moderately malnourished), all were considered well nourished by the WHO classification.

In 2006, Yamauti7 found a similar result when comparing nutritional diagnoses obtained by the SGNA and anthropometrics in adult patients with heart disease, in which there was a prevalence of malnutrition that was 9.4% higher by the SGNA than by anthropometrics. Similarly, in 1996, the study by Ek et al.,27 when comparing the same methods in hospitalized elderly individuals, also observed that the number of patients classified as malnourished by subjective methods exceeded that identified by objective methods.

These differences can be explained by the fact that the subjective methodology consists of qualitative questions, and thus, it is efficient in the early identification of nutritional risk and nutritional alterations.28

This can be explained by the fact that the SGNA questionnaire addresses some of the points that actually improve the study of the child's nutritional status, such as weight loss, anthropometric birth data for children younger than 2 years (weight and length), home eating habits, presence of gastrointestinal symptoms, physical examination, etc. For this reason, it is able to detect signs of malnutrition that might go undetected when considering only weight, height, and BMI in the nutritional assessment, as in the case of the classification according to the WHO curves.28

In the present study, it was observed that the nutritional statuses generated by the SGNA and the WHO have a low agreement. In 2014, Carniel18 similarly observed a low agreement (0.38, p=0.001) between the weight/height variable and the SGNA, whereas for the other variables, agreement levels were different from those found in this study.

Possibly, the combination of objective and subjective data proposed in the SGNA contributes to the fact that the final classifications of the methodologies disagree, which shows the importance of using both tools in clinical practice as complementary techniques for an adequate nutritional diagnosis.18

Based on the association between the nutritional assessment methods, it was verified that the weight/height variable showed a strong and positive correlation between the applied methods, unlike the others, which showed lower levels of correlation. This can be explained by the fact that weight/height index assesses whether the child's weight is adequate for her height, thus demonstrating the proportionality between these two measurements and eliminating the age element.18

These results demonstrate the importance of performing the nutritional assessment through objective and subjective methods at hospital admission, aiming to identify malnourished patients and/or those at nutritional risk. The early diagnosis of these conditions allows the establishment of an adequate diet therapy, which contributes to the treatment and improvement of the prognosis and hospital discharge of the individual. As demonstrated by Baccaro et al. in 2007,29 who evaluated the nutritional status of patients in a hospital in Argentina, malnourished patients were hospitalized for a period longer than seven days (on average) than patients with normal weight. Moreover, among the patients who died, 62.5% belonged to the group classified as severely malnourished.

Furthermore, the SGNA allows the concise and rapid collection of a large part of the patient's history, and also identifies those at greater nutritional risk who need a more careful investigation of their clinical history, indicating which individuals require a more intense follow-up, preventing the worsening of the nutritional picture. The same was observed by Rocha et al. in 2006,30 who evaluated the nutritional status of children younger than 5 years evaluated at admission and at hospital discharge. They found that 51.6% of the patients had weight loss during hospitalization, and 9.17% of the children with normal weight prior to hospital admission had their nutritional condition worsen during hospitalization.

However, the SGNA also has limitations, namely the need for training and experience of the observer, which is an important factor for the test accuracy, since the assessment accuracy depends on the ability to recognize nutritional variations. In addition, because it is a qualitative method, it has low sensitivity to identify small nutritional status alterations, and therefore it is not a good tool to follow the patient's evolution, with objective methods being more adequate in this case.28

At the same time, another negative point observed in the present study was the non-differentiation between children with normal weight from those with overweight and/or obesity, as proposed by the WHO classification.9,10 Therefore, all of these patients, although showing very different nutritional diagnoses and requiring diverse dietary therapies, are classified as well-nourished, which may induce errors. As well as the malnourished patients, those who are overweight also need special care, since this condition also characterizes a nutritional status imbalance and increases the risk of developing chronic diseases and, consequently, morbidity and mortality.

The fact that the assessments were made by only one observer is also a limitation of the present study, which could lead to a aferition bias. However, every precaution was taken to minimize possible errors. Furthermore, the authors emphasize the need for further studies that address this issue.

Low agreement was found among the assessed methods. Thus, in the absence of a gold standard method, it is concluded that the combination of objective and subjective methods, such as anthropometric assessment through the WHO curves and the SGNA, may be beneficial for pediatric patients, since it allows the early diagnosis of nutritional alterations and/or nutritional risk. Consequently, this indication can help in the creation of an adequate dietary therapy to minimize the risks of compromising the growth and development of the pediatric patient.

The authors declare no conflicts of interest.