Malnutrition Universal Screening Tool

1. Abstract

The Malnutrition Universal Screening Tool (MUST) is a validated, five-step clinical screening instrument developed by the Malnutrition Advisory Group (MAG), a standing committee of the British Association for Parenteral and Enteral Nutrition (BAPEN), spearheaded by Professor Marinos Elia in 2003. Engineered to identify adult patients across varied healthcare settings who are malnourished, at risk of malnutrition (undernutrition), or obese, MUST functions as a predictive, action-oriented triaging scale. The instrument measures nutritional vulnerability through three core physiological parameters across distinct operational steps: current nutritional depletion measured via Body Mass Index (BMI; Step 1), chronological trajectory of nutritional loss via percentage of unintentional weight loss over the preceding 3 to 6 months (Step 2), and acute metabolic disruption coupled with dietary arrest induced by severe illness for more than 5 days (Step 3). Each step produces a categorical score ranging from 0 to 2, yielding an aggregate continuous sum from 0 to 6. This cumulative index categorizes patients into three pragmatic risk strata: low risk (Score 0), medium risk (Score 1), and high risk (Score ≥ 2), each directly linked to evidence-based management protocols (Step 4 and Step 5). Psychometric evaluations across acute inpatient, geriatric, primary care, and community cohorts demonstrate exceptional inter-rater reliability (Cohen’s kappa (κ) typically ranging between 0.80 and 1.00), robust convergent validity with detailed comprehensive nutritional assessments, and powerful predictive validity for length of hospital stay, discharge destination, mortality, and healthcare costs. By bridging objective anthropometry with acute clinical pathophysiology, MUST remains one of the most widely implemented and psychometrically sound clinical risk screening protocols globally.

2. Keywords

Malnutrition Universal Screening Tool, MUST, nutritional screening, undernutrition, Body Mass Index, unintentional weight loss, acute disease effect, psychometrics, clinical assessment, BAPEN

3. Authors

The Malnutrition Universal Screening Tool was devised and validated under the leadership of Professor Marinos Elia, MD, FRCP, Professor of Clinical Nutrition and Metabolism at the University of Southampton, United Kingdom, acting on behalf of the Malnutrition Advisory Group (MAG) of the British Association for Parenteral and Enteral Nutrition (BAPEN). Key collaborators and contributing committee members include clinicians, dietitians, and researchers associated with BAPEN, such as Rebecca J. Stratton, Christine Russell, and Alastair Forbes. The Dutch clinical translation and institutional cross-cultural adaptation were executed under the auspices of the Stuurgroep Ondervoeding (Dutch Malnutrition Steering Group), ensuring standardized clinical implementation across primary care, residential homes, and acute hospitals in the Netherlands and Flanders.

4. Purpose

The primary clinical objective of the Malnutrition Universal Screening Tool (MUST) is the rapid, reliable, and standardized detection of protein-energy undernutrition and the identification of individuals at high risk of developing severe nutritional depletion across acute, ambulatory, and institutional care settings. Undernutrition frequently manifests as an insidious clinical syndrome characterized by progressive loss of lean tissue mass, blunted immune competence, impaired wound healing, delayed post-operative recovery, elevated readmission rates, and heightened morbidity and mortality. Despite its profound clinical consequences, malnutrition remains notoriously underdiagnosed in both inpatient and outpatient settings due to overlapping disease symptomatology and inconsistent clinical documentation. MUST was formulated to overcome these systematic oversights by providing a universal, standardized, rapid checklist that requires fewer than five minutes to complete per patient.

From an applied clinical perspective, MUST serves not merely as a diagnostic classification tool, but as an operationalized clinical decision algorithm. It bridges screening with definitive intervention: Step 4 integrates the diagnostic score into a tripartite risk matrix (low, medium, or high risk), and Step 5 outlines customized clinical care plans. These range from routine rescreening schedules in low-risk individuals to intensive dietary counseling, oral nutritional supplementation, or artificial enteral/parenteral nutrition overseen by specialized dietetic teams in high-risk individuals. Furthermore, MUST incorporates practical surrogate measures (such as ulna length for height estimation, knee height, or mid-upper arm circumference [MUAC]) to ensure high feasibility in bedbound, unconscious, or critically compromised patients who cannot undergo standard standing anthropometry.

In epidemiological and clinical research paradigms, MUST provides a standardized, operationalized metric of nutritional vulnerability. It allows researchers to quantify the prevalence of malnutrition across diverse disease cohorts, evaluate the independent prognostic contribution of nutritional status to clinical outcomes (such as critical care survivorship, oncological treatment tolerance, and surgical site infections), and control for baseline physiological reserve in prospective clinical trials. By delineating acute nutritional failure from chronic, insidious physical wasting, MUST provides researchers with a robust, reproducible index of metabolic vulnerability.

5. Psychological Construct

The Malnutrition Universal Screening Tool measures the multidimensional physiological and functional construct of nutritional vulnerability and protein-energy undernutrition. Rather than treating nutritional status as a static, unidimensional index, MUST conceptualizes malnutrition as a dynamic interplay between baseline somatic reserves, chronicity of nutritional decline, and acute catabolic demand. The construct comprises three distinct operational domains:

Current Somatic Nutritional Depletion (Step 1: Body Mass Index)

Body Mass Index represents the individual’s current static nutritional reserve adjusted for height. In the construct of undernutrition, BMI reflects the long-term equilibrium between sustained energy intake and expenditure. A severely suppressed BMI (<18.5 kg/m²) indicates clinically significant depletion of adipose stores and lean muscle mass. However, MUST conceptualizes this dimension continuously and transitionally, recognizing an intermediate margin (18.5–20.0 kg/m²) where individuals possess diminished physiological buffer capacity and are predisposed to rapid clinical decompensation when exposed to secondary physiological stressors.

Kinetic/Chronological Trajectory of Nutritional Loss (Step 2: Unplanned Weight Loss)

Static BMI measurements can obscure acute and severe wasting states, particularly in individuals with pre-existing overweight or obesity (sarcopenic obesity) who lose substantial lean tissue while retaining a BMI within normative ranges. Therefore, the second construct dimension captures the kinetic trajectory of energy balance via unintentional weight loss over the preceding 3 to 6 months. Weight loss exceeding 5% indicates early negative energy balance, whereas weight loss exceeding 10% indicates profound, clinically pathological tissue catabolism associated with altered cellular physiology, muscular atrophy, and heightened mortality, regardless of the patient’s baseline BMI.

Acute Metabolic Disruption and Intake Arrest (Step 3: Acute Disease Effect)

The third dimension accounts for the acute physiological, neuroendocrine, and metabolic shifts that occur during severe illness. Under conditions of critical systemic illness (e.g., severe sepsis, major trauma, extensive burns, acute pancreatitis, or major abdominal surgery), the body enters a hypermetabolic state characterized by profound cytokine release, accelerated proteolysis, and insulin resistance. Concurrently, physiological or mechanical factors may completely suppress dietary intake. MUST defines this dimension through the concurrent presence of acute illness and the absence (or anticipated absence) of nutritional intake for more than 5 days. This step reflects immediate metabolic vulnerability, recognizing that acute intake cessation in a hypercatabolic state leads to rapid somatic wasting and compromised vital organ function.

6. Theoretical Framework

The theoretical framework underpinning the Malnutrition Universal Screening Tool is rooted in the pathophysiology of metabolic starvation and stress-induced catabolism. Classically elucidated by clinical nutritionists and metabolic researchers, the human adaptation to non-complicated starvation involves a downregulation of resting energy expenditure, a reduction in protein breakdown, and an adaptive shift toward utilizing ketone bodies derived from fatty acid oxidation. However, when acute disease or trauma is superimposed upon nutritional deprivation, this adaptive conservation mechanism fails. Neuroendocrine activation of the hypothalamic-pituitary-adrenal axis and systemic release of pro-inflammatory cytokines (such as Tumor Necrosis Factor-alpha, Interleukin-1, and Interleukin-6) induce a state of unrelenting hypercatabolism, muscle proteolysis, and gluconeogenesis, which cannot be reversed by simple rest.

From an applied measurement perspective, MUST is built upon the conceptual model of tri-axial risk stratification. This model posits that clinical nutritional vulnerability ($V$) is a function of static tissue depletion ($D$), dynamic rate of loss ($L$), and acute systemic stress ($S$):

V = f(D, L, S)

By formalizing each theoretical component into an objective, standardized step, MUST aligns with the World Health Organization’s conceptual definitions of adult undernutrition while addressing the limitations of single-marker assessments. Traditional diagnostic methods relied heavily on biochemical indices such as serum albumin or prealbumin. However, contemporary nutritional science recognizes these hepatic transport proteins as negative acute-phase reactants indicative of systemic inflammation and vascular permeability rather than true markers of nutritional status. MUST therefore intentionally excludes visceral protein biomarkers, grounding its framework entirely in reliable anthropometric indices and clear, objective clinical indicators of intake arrest.

7. Validity

The Malnutrition Universal Screening Tool has undergone extensive psychometric and clinical validation across diverse global healthcare environments, demonstrating robust construct, convergent, criterion, and predictive validity.

Construct and Convergent Validity

Construct validity has been repeatedly demonstrated by evaluating MUST against comprehensive, multidimensional nutritional assessment batteries, including full dietetic clinical evaluations, the Subjective Global Assessment (SGA), and the Mini Nutritional Assessment (MNA). In foundational validation studies conducted by Elia and colleagues (2003) across acute inpatient wards, MUST displayed high concordance with full clinical assessments, yielding sensitivity values between 78% and 95% and specificity values between 76% and 93%. When compared directly against SGA in acute hospitalized adult cohorts, MUST consistently achieves significant diagnostic agreement, with kappa values commonly ranging from 0.65 to 0.88, confirming that the tool accurately captures the theoretical construct of undernutrition.

Predictive Validity

Predictive validity represents one of the strongest psychometric attributes of MUST. Extensive prospective cohort investigations have demonstrated that MUST risk categories are strong independent predictors of adverse clinical endpoints, even after controlling for primary medical diagnosis, age, and comorbid disease burden:

• Length of Hospital Stay: Patients categorized as high risk on MUST consistently demonstrate significantly longer hospital stays compared to low-risk counterparts (often 2 to 4 times longer, p < .001).
• Mortality: In prospective evaluations across general medical, surgical, and geriatric populations, patients categorized as high risk exhibit substantially elevated in-hospital mortality rates (odds ratios frequently ranging from 2.5 to 5.0) and post-discharge 6-month mortality rates.
• Healthcare Utilization: In primary care and community settings, individuals identified as medium or high risk on MUST have higher rates of general practitioner consultations, more frequent antibiotic prescriptions, and greater rates of unplanned hospital admissions.

Discriminant Validity

MUST demonstrates clear discriminant validity by distinguishing uncomplicated chronic medical conditions without nutritional impact from conditions characterized by active, uncompensated catabolic wasting. Its step-based design prevents individuals who are chronically thin but metabolically stable (e.g., constitutional low BMI without weight loss or acute illness) from being inappropriately classified as high risk, preserving clinical resources for those with genuine, active tissue wasting.

8. Reliability

Given that MUST is an observational screening checklist combining anthropometric measurements with clinical history, its reliability is primarily evaluated via inter-rater agreement and intra-rater test-retest reproducibility rather than internal consistency metrics such as Cronbach’s alpha (which are conceptually unsuited for additive, index-based clinical risk scales).

Inter-Rater Reliability

Extensive multicenter investigations have examined inter-rater reliability across distinct clinical disciplines, comparing assessments performed by registered dietitians (gold standard), registered nurses, healthcare assistants, and medical doctors. In the seminal validation trials by Elia (2003) and Stratton et al. (2004), inter-rater reliability for overall MUST risk classification across acute hospital wards yielded Cohen’s kappa values of κ = 0.80 to 1.00, denoting near-perfect agreement. Individual step analysis typically demonstrates:

• Step 1 (BMI category): κ = 0.88 to 0.96
• Step 2 (Weight loss category): κ = 0.78 to 0.90
• Step 3 (Acute disease score): κ = 0.65 to 0.82

The slightly lower, yet still substantial, agreement observed for Step 3 reflects subjective clinical variations in anticipating whether an acutely ill patient will maintain intake over the subsequent 5-day horizon.

Test-Retest Reliability and Feasibility

Test-retest assessments conducted within 24 to 48 hours of initial hospital admission demonstrate excellent temporal stability (κ > 0.85), provided no rapid fluid shifts or acute resuscitations occur. Furthermore, usability studies report that healthcare staff successfully complete MUST in more than 95% of admitted patients within an average completion time of 3 to 5 minutes, underscoring its operational feasibility in fast-paced clinical environments.

9. Factor Analysis

The Malnutrition Universal Screening Tool is conceptually and mathematically structured as a formative, cumulative risk index rather than a reflective latent factor model. In classical psychometric theory, reflective models (such as those analyzed via standard Confirmatory Factor Analysis [CFA]) assume that an underlying, unobserved latent construct causally produces the observable responses across parallel items. In contrast, formative indices integrate separate, clinically meaningful physical and pathological parameters to construct a cumulative vulnerability score; a change in one parameter (e.g., developing an acute infection) does not necessarily require a concurrent change in another (e.g., past 3-to-6-month weight loss).

Consequently, while standard Exploratory Factor Analysis (EFA) or CFA indices (such as Root Mean Square Error of Approximation [RMSEA] or Comparative Fit Index [CFI]) are rarely applied to MUST in clinical validation studies, structural investigations into the dimensionality of malnutrition screening tools yield important psychometric findings:

• Item Independence and Communality: When subjected to principal component analysis, the three steps load onto two distinct physiological dimensions accounting for over 70% of the total variance: a Chronic Nutritional Reserve Dimension (heavily loaded by Step 1 [BMI] and Step 2 [Weight loss]) and an Acute Stress/Inflammatory Dimension (governed predominantly by Step 3 [Acute disease effect]).
• Multi-collinearity Absence: Variance Inflation Factors (VIF) among the three parameters consistently remain well below 2.0, demonstrating that the steps provide non-redundant, independent diagnostic information to the cumulative clinical score.
• Additivity and Formative Integrity: Clinical validation demonstrates that all three components contribute significant, independent prognostic weight in multivariable logistic regression models predicting clinical deterioration, justifying their additive, unweighted combination into the composite clinical index.

10. Instrument / Measurement Tool

• Tool Name: Malnutrition Universal Screening Tool (MUST)
• Instrument Type: Clinical screening tool / Observational checklist
• Target Population: Adult patients (including geriatric populations) across acute hospital wards, outpatient clinics, residential and nursing homes, and primary care settings
• Administration Time: Approximately 3 to 5 minutes per patient
• Total Steps: 5 procedural steps (Steps 1 to 3 comprise the diagnostic assessment; Steps 4 and 5 comprise overall risk stratification and clinical management planning)
• Response / Scoring Scale: Step-specific categorical scoring: Step 1 (BMI: >20.0 = 0, 18.5-20.0 = 1, <18.5 = 2); Step 2 (Unplanned weight loss in past 3-6 months: <5% = 0, 5-10% = 1, >10% = 2); Step 3 (Acute disease effect: Acutely ill with no nutritional intake for >5 days = 2, otherwise = 0)
• Scoring Interpretation: Overall risk of malnutrition is calculated by summing the scores of Steps 1, 2, and 3: Score 0 = Low risk (routine clinical care); Score 1 = Medium risk (observe); Score 2 or more = High risk (treat).
• Clinical Management Protocol (Step 5):
  • Low Risk (Score 0): Routine clinical care. Repeat screening weekly in hospitals, monthly in care homes, and annually in the community for vulnerable groups.
  • Medium Risk (Score 1): Observe and monitor. Document dietary intake for 3 days; monitor body weight; rescreen weekly in hospitals and at least monthly in care homes.
  • High Risk (Score ≥ 2): Treat. Refer to registered dietitian, nutritional support team, or implement local nutritional management protocols; improve overall nutritional intake; monitor and review clinical care plan regularly.

11. Permissions & Fee and Test Year

The Malnutrition Universal Screening Tool was formally established and published in 2003 by the Malnutrition Advisory Group (MAG) of the British Association for Parenteral and Enteral Nutrition (BAPEN). The copyright of MUST and its instructional materials is retained by BAPEN. However, BAPEN actively encourages widespread, non-commercial clinical, educational, and academic research use of the instrument without charging licensing fees, provided that the tool is reproduced without modification and proper attribution is preserved. Any electronic integration into proprietary electronic medical record (EMR) software or commercial healthcare platforms requires formal permission and licensing clearance from BAPEN. Details regarding reproduction permissions, clinical implementation toolkits, and educational resources are available through the official BAPEN website.

12. References

Elia, M. (2003). The ‘MUST’ report: Nutritional screening of adults: A multidisciplinary responsibility. Development and implementation of the ‘Malnutrition Universal Screening Tool’ (‘MUST’) for adults. Malnutrition Advisory Group (MAG), British Association for Parenteral and Enteral Nutrition (BAPEN). https://www.bapen.org.uk/screening-and-must/must/must-report/

Stratton, R. J., Hackston, A., Longmore, D., Dixon, R., Price, S., Stroud, M., King, C., & Elia, M. (2004). Malnutrition in hospital outpatients and inpatients: Prevalence, concurrent validity and ease of use of the ‘Malnutrition Universal Screening Tool’ (‘MUST’). British Journal of Nutrition, 92(5), 799–808. https://doi.org/10.1079/BJN20041258

Stratton, R. J., King, C. L., Stroud, M. A., Jackson, A. A., & Elia, M. (2006). ‘Malnutrition Universal Screening Tool’ predicts mortality and length of hospital stay in acutely ill elderly patients. British Journal of Nutrition, 95(2), 325–330. https://doi.org/10.1079/BJN20051622

Todorovic, V., Russell, C., & Elia, M. (2011). The ‘MUST’ explanatory booklet: A guide to the ‘Malnutrition Universal Screening Tool’ (‘MUST’) for adults. British Association for Parenteral and Enteral Nutrition (BAPEN).

Kondrup, J., Allison, S. P., Elia, M., Vellas, B., & Plauth, M. (2003). ESPEN guidelines for nutrition screening 2002. Clinical Nutrition, 22(4), 415–421. https://doi.org/10.1016/S0261-5614(03)00098-0

Stuurgroep Ondervoeding. (2019). Screening en behandeling van ondervoeding bij volwassenen. Stuurgroep Ondervoeding Nederland. https://www.stuurgroepondervoeding.nl

13. Items of the Scale