Development and validation of the ‘MUST’
1. In which healthcare settings was ‘MUST’ piloted?
The reliability of ‘MUST’ was established by assessing extent to which the malnutrition risk obtained independently by different healthcare workers on the same group of patients agreed with each other (inter-rater agreement). A series of studies were undertaken in medical and surgical wards, outpatient clinics, nursing / residential homes and a GP surgery. Agreement was >95% in all studies. Most other screening tools have not been tested in this way and where they have, the level of agreement has generally been lower than with the ‘MUST’.
2. Has ‘MUST’ been validated for use in subjects with learning or physical disabilities? Are the BMI cut-off values relevant for this client group?
BMI is used as a general indicator of protein energy status and recommended for this purpose by national and international organisations. Whilst some clients with learning or physical disabilities were included in the pilot phase of ‘MUST’, the tool was not specifically validated in this client group. Research is required to assess if BMI reflects the same or different body composition in those subjects with learning or physical disabilities.
3. Is ‘MUST’ suitable for use in patients with renal or liver disease?
‘MUST’ has been designed for use in all care settings by all healthcare workers. There is no reason why it cannot be used in patients with renal or liver disease, but care should be taken when interpreting BMI and weight loss if fluid balance is disturbed (see details in the ‘MUST’ Report and ‘MUST’ Explanatory Booklet).
4. Why haven’t the cut-off points for BMI used in ‘MUST’ been adjusted for use in the elderly?
The lower BMI cut-off points for malnutrition risk used in ‘MUST’ are consistent with those given in a WHO report on loss of physiological function in relation to BMI. A wide range of BMI cut-off points (<17 kg/m2 to <24kg/m2) have been used to indicate malnutrition in older subjects. Use of these results in enormous differences in the prevalence of malnutrition in this population group and in the strategies and resources required to manage it.
Confusion appears to have arisen from the inappropriate extrapolation of BMI cut-off values obtained from public health initiatives to clinical practice. BMI cut-off values for public health are primarily intended for groups of subjects without overt disease whilst in clinical practice they are intended for those subjects with disease.
In public health, BMI is typically used to aid the prediction and prevention of mortality, often over many years, mainly from cardiovascular disease. In clinical practice, BMI is typically used to aid the prediction of current nutritional status and body function and the likely response to treatment, usually over a much shorter time frame.
Public health and clinical approaches may not yield the same BMI values or require the same nutritional interventions.
In the UK, public health surveys have not used the elevated lower cut-off values but have consistently used BMI of < 20 kg/m2 to indicate underweight in adults, including those individuals over 65 years of age. In USA, the 1990 Dietary Guidelines for Americans which suggested age specific BMI ranges were withdrawn in 1995 as it was recognised that other variables (e.g. smoking, lifestyle) affected mortality not just BMI.
In clinical practice, many professional organisations, agencies and healthcare workers recommend a BMI cut-off value of 18.5 – 20 kg/m2 to identify nutritional risk in a range of ages including older subjects who account for ~50% hospital population and 90% of those in nursing homes.
The cut-off values used in ‘MUST’ are based on physiological and clinical observations on loss of body function as BMI decreases, the apparently normal body function in many older subjects with a BMI >20 kg/m2 and randomised controlled trials showing the benefits of nutritional support in subjects, particularly those in the community with a BMI <20 kg/m2.
5. How was the score of 2 for the effect of acute disease derived?
There are 2 aspects to take into account when considering the effect of acute disease on malnutrition risk:
(i) No or virtually no food intake for more than 5 days i.e. starvation
(ii) The catabolic effect of acute disease.
During starvation, taller and heavier individuals tend to lose weight faster than those who are shorter and lighter. Hence men tend to lose weight faster than women. However, malnutrition risk is influenced more by the percentage weight loss than by actual weight loss. The same absolute weight loss results in a greater percentage weight loss in leaner individuals than in heavier subjects.
After 5 days of total starvation, subjects with an initial BMI of 17.5 – 18.0 kg/m2 lose close to 10% body weight; those between 20 – 25 kg/m2 lose about 6-8% body weight; and those with a BMI of ~ 35 kg/m2 lose about 5% body weight.
In the presence of acute disease, basal metabolic rate is increased but physical activity is decreased and an increase in energy expenditure may not occur. However, there is an increase in protein oxidation and nitrogen loss, therefore in the absence of fluid retention, the rate of weight loss and loss of lean body mass tends to be more rapid in the presence of acute disease. Typically the percentage weight loss in acute disease together with no or virtually no food intake for more than 5 days is comparable to > 10% over 3 –6 months. Hence a score of 2 has been assigned.
6. What is the evidence for the equal weightings given to the 3 criteria used in ‘MUST’?
Each of the 3 components of ‘MUST’ can occur independently or together. Each can have detrimental effects on physiological function and clinical outcome, therefore it is reasonable to assign an important weighting to all of them. The components vary in importance according to patient group, outcome variable and healthcare setting.
The predictive value of the individual components was independently assessed in a variety of healthcare specialities and settings. ‘MUST’ categorisation was found to be significantly related to mortality, length of stay in hospital and number of GP visits by patients in the community. The relative importance of each factor was found to vary with the care setting and type of patient (i.e. medical speciality), therefore, the MAG Committee decided to assign equal weightings to BMI, recent weight loss and acute disease effect in order to simplify use of the tool whilst retaining overall predictive validity.
7. Where can I obtain further information on the development and validation of the ‘MUST’?
“The ‘MUST’ Report” examines the need to screen, contains the evidence base for the criteria used in ‘MUST’ and describes the development and validation of the ‘MUST’. Copies can be purchased from the BAPEN office:
Tel: 01527 457850
8. Has ‘MUST’ been validated for use in children?
No. ‘MUST’ is only validated for use in adults.