The Low Sensitivity and Specificity of a Nutrition Screening Tool in Real Circumstances in a Tertiary Care Hospital Setting

Abstract

Nutrition screening is an essential process to detect children at risk of malnutrition during hospitalization and provide appropriate nutrition management. STRONGkids is a nutrition screening tool which has been implemented in a tertiary-care hospital service in Bangkok, Thailand. This study aimed to evaluate the performance of STRONGkids in the real-situation setting. Electronic Medical Records (EMR) of hospitalized pediatric patients aged 1 month to 18 years from January to December 2019 were reviewed. Those with incomplete medical records and re-admission within 30 days were excluded. Nutrition risk scores and clinical data were collected. Anthropometric data were calculated to Z-score based on the WHO growth standard. The sensitivity (SEN) and specificity (SPE) of STRONGkids were determined against malnutrition status and clinical outcomes. In total, 3914 EMRs (2130 boys, mean age 6.22 ± 4.72 years) were reviewed. The prevalence of acute malnutrition (BMI-for-age Z-score < −2) and stunting (height-for-age Z-score < −2) were 12.9 and 20.5%. SEN and SPE of STRONGkids against acute malnutrition were 63.2 and 55.6%, stunting values were 60.6 and 56.7%, and overall malnutrition values were 59.8 and 58.6%. STRONGkids had low SEN and SPE to detect nutrition risks among hospitalized children in a tertiary-care setting. Further actions are required to improve the quality of nutrition screening in hospital services.

Keywords

• nutrition screening tool
• hospitalized children
• STRONGkids
• malnutrition
• validation