Usage
  • 39 views
  • 44 downloads

Can screening on admission identify children who are malnourished?

  • Author / Creator
    Carter, Laura
  • Background: Children are at high risk for malnutrition during hospital admission. Over half of children admitted to hospital will exhibit signs of nutrition deterioration such as weight loss. Screening for malnutrition is a critical step in the nutrition care process, however there is currently a lack of validated screening tools specifically for children in Alberta. Multiple screening tools have been proposed for pediatric use globally, but none are validated in a Canadian population. There is insufficient information available to select one screening tool over the others for widespread use. A rigorous process of selection and validation is required to determine which of the available pediatric nutrition screening tools is appropriate for use in a specific population.Objectives: The aim of this paper is to first, compare available pediatric nutrition screening tools and select two appropriate for use in an Alberta institute. The second aim is to compare the two selected tools in a validation study at the Stollery Children’s Hospital in Edmonton, Alberta, and propose one appropriate for implementation into clinical care.Methods: A literature review identified five nutrition screening tools created and validated for use in children admitted to hospital. Of those five tools, two were believed to be appropriate for use in Alberta; the Pediatric Nutrition Screening Tool (PNST), and the Screening Tool for Risk on Nutritional Status and Growth (STRONGkids). These two tools were evaluated to determine which was best able to identify malnutrition risk on admission with acceptable sensitivity, specificity, and agreement with the Subjective Global Nutritional Assessment (SGNA). Patients admitted to surgery and medicine units at an Alberta pediatric hospital were approached to participate (n=165). Both screening tools were completed on each patient by a nurse and a nutrition risk score was calculated basediiion recommended cut-offs. The SGNA was then completed by a trained dietitian, blinded to the results of the screen. Statistics: Sensitivity and specificity were calculated for both screening tool against the SGNA. A Receiver Operator Characteristic (ROC) curve was used to assess alternate cut-offs for each tool. Results: Based on the SGNA, 29% of patients were malnourished on admission. Using the recommended cut-offs STRONGkids identified 56% and 16% as at moderate and severe nutrition risk respectively with a sensitivity of 89%, specificity of 35%, and Cohen’s K of 0.483. PNST identified 26% as at nutrition risk with a sensitivity of 58%, specificity of 88%, and Cohen’s K of 0.601. Using adjusted cut-offs based on ROC curve analysis, the PNST improved to a sensitivity of 87%, specificity of 71%, and Cohen’s K of 0.681, and STRONGkids improved to a sensitivity of 80%, specificity of 61%, and Cohen’s K of 0.5. Those who were malnourished based on the SGNA stayed in hospital 2.9 days longer than those well-nourished (p < 0.05). Children identified as at nutrition risk by both tools using original and adjusted cut-offs had significantly longer lengths of hospital stay.Conclusion: This study showed neither tool was able to identify children at nutrition risk with acceptable concurrent validity in this population. When the nutrition risk cut-offs were adjusted to better fit the study population, both tools had better agreement with the SGNA. The PNST with adjusted cut-offs had the strongest concurrent validity and appears to be the tool best suited for use in Alberta pediatric hospitals. Selection of a nutrition screening tool is the first step in creation of pediatric nutrition care algorithm to guide clinicians and positively impact the nutrition care of children while admitted to hospital.

  • Subjects / Keywords
  • Graduation date
    Spring 2019
  • Type of Item
    Thesis
  • Degree
    Master of Science
  • DOI
    https://doi.org/10.7939/r3-hxjr-jm18
  • License
    Permission is hereby granted to the University of Alberta Libraries to reproduce single copies of this thesis and to lend or sell such copies for private, scholarly or scientific research purposes only. Where the thesis is converted to, or otherwise made available in digital form, the University of Alberta will advise potential users of the thesis of these terms. The author reserves all other publication and other rights in association with the copyright in the thesis and, except as herein before provided, neither the thesis nor any substantial portion thereof may be printed or otherwise reproduced in any material form whatsoever without the author's prior written permission.