Get access

The Malnutrition Screening Tool versus objective measures to detect malnutrition in hip fracture

Authors

  • J. J. Bell,

    Corresponding author
    1. Department of Nutrition and Dietetics, The Prince Charles Hospital, Queensland Health, Brisbane, Queensland, Australia
    2. Centre for Dietetics Research, School of Human Movement Studies, University of Queensland, Brisbane, Queensland, Australia
    • Correspondence

      J. Bell, The Prince Charles Hospital, Department of Nutrition and Dietetics, Rode Road, Chermside 4035, Brisbane, Queensland, Australia.

      Tel.: +61 7 3139 5589

      Fax: +61 7 3139 5589

      E-mail: jack_bell@health.qld.gov.au

    Search for more papers by this author
  • J. D. Bauer,

    1. Centre for Dietetics Research, School of Human Movement Studies, University of Queensland, Brisbane, Queensland, Australia
    Search for more papers by this author
  • S. Capra

    1. Centre for Dietetics Research, School of Human Movement Studies, University of Queensland, Brisbane, Queensland, Australia
    Search for more papers by this author

Abstract

Background

The Malnutrition Screening Tool (MST) is the most commonly used screening tool in Australia. Poor screening tool sensitivity may lead to an under-diagnosis of malnutrition, with potential patient and economic ramifications. The present study aimed to determine whether the MST or anthropometric parameters adequately detect malnutrition in patients who were admitted to a hip fracture unit.

Methods

Data were analysed for a prospective convenience sample (n = 100). MST screening was independently undertaken by nursing staff and a nutrition assistant. Mid upper arm circumference (MUAC) was measured by a trained nutrition assistant. Nutritional risk [MST score ≥ 2, body mass index (BMI) < 22 kg m–2, or MUAC < 25 cm] was compared with malnutrition diagnosed by accredited practicing dietitians using International Classification of Diseases version 10-Australian Modification (ICD10-AM) coding criteria.

Results

Malnutrition prevalence was 37.5% using ICD10-AM criteria. Delirium, dementia or preadmission cognitive impairment was present in 65% of patients. The BMI as a nutrition risk screen was the most valid predictor of malnutrition (sensitivity 75%; specificity 93%; positive predictive value 73%; negative predictive value 84%). Nursing MST screening was the least valid (sensitivity 73%; specificity 55%; positive predictive value 50%; negative predictive value 77%). There was only fair agreement between nursing and nutrition assistant screening using the MST (κ = 0.28).

Conclusions

In this population with a high prevalence of delirium and dementia, further investigation is warranted into the performance of nutrition screening tools and anthropometric parameters such as BMI. All tools failed to predict a considerable number of patients with malnutrition. This may result in the under-diagnosis and treatment of malnutrition, leading to case-mix funding losses.

Get access to the full text of this article

Ancillary