1. The Nutrition Screening Survey
1.1 This report provides a summary of the largest nutrition screening survey undertaken in the UK. Reporters from 175 hospitals, 173 care homes and 22 mental health units in the UK completed a general questionnaire and an anonymous patient questionnaire as part of a national audit on nutritional screening. Unlike previous studies that used different criteria to identify malnutrition in various care settings, this survey used the same criteria based on the ‘Malnutrition Universal Screening Tool’ (‘MUST’) in all care settings. Data were collected on patients during the first three days of admission to hospitals and acute mental health units, and on residents admitted to care homes and long stay / rehabilitation mental health units in the previous six months.
2.1 Of 9336 patients who were screened on admission to hospital, 28% were found to be at risk of malnutrition, (high risk, 22%; medium risk, 6%). The combination of these two risk categories is henceforth referred to as ‘malnutrition’ for simplicity.
2.2 Since the presence of ‘malnutrition’ at or shortly after admission to hospitals suggests that it largely originated in the community, strategies to prevent and treat malnutrition in the community setting should be considered.
2.3 ‘Malnutrition’ varied significantly according to source of admission (25% from home, 31% from another hospital, 32% from another ward, and 43% from a care home), type of admission (32% for emergency admission, 20% for elective admission), and type of ward (e.g. 43% in oncology wards and 15% in orthopaedic/trauma wards). It was also greater in hospitals that had a screening policy than those that did not (28% v 24%), and considerably greater in large hospitals with ≥1000 beds than in those with <1000 beds (38% v 26%).
2.4 ‘Malnutrition’ was common in all age groups and diagnostic categories, but it was significantly more common in women, who were older than men (29% v 26%), in subjects aged over 65 years than under 65 years (30% v 24%), and in certain diagnostic categories than others (e.g. gastrointestinal disease (43%) and neurological disease (33%) versus cardiovascular (21%) and musculoskeletal conditions (18%)). A low body mass index (BMI <20 kg/m2) contributed to a ‘MUST’ score in 4 out of 10 ‘malnourished’ patients.
2.5 Most hospitals reported that they had a screening policy (89%), but weighing on all wards was carried out in less than half (less than a quarter of patients involved in the survey were in hospitals where weighing was carried out on all wards). This suggests that much malnutrition continues to be under-recognised and under-treated.
3. Care homes
3.1 Of 1610 residents screened 30% were ‘malnourished’ (20% high risk, 10% medium risk).
3.2 The prevalence of ‘malnutrition’ was greater in residents admitted from hospitals (35%) and other care homes (30%) than in those admitted from their own homes (24%). The prevalence was also greater in nursing homes (35%) than residential homes (22%).
3.3 ‘Malnutrition’ was more prevalent in care homes that had a screening policy (31%) (these care homes accounted for 82% of the residents in the survey) than in those that did not (23%).
3.4 A low BMI (<20 kg/m2) contributed to the ‘MUST’ score in about 9 out of 10 ‘malnourished’ residents. Underweight was 2- to 3-fold more common than obesity. The subjects in care homes were older than those in hospitals and mental health units, half of them being 85 years and over. The prevalence of ‘malnutrition’ tended to increase with age (26% in those <70 years, 29% in those 70-84 years and 32% in those ≥ 85 years) and duration of stay (up to 6 months), but neither of these were statistically significant.
3.5 Women were older and had a greater prevalence of ‘malnutrition’ than men (32% v 27%).
3.6 Half the residents had neurological conditions with an associated ‘malnutrition’ prevalence that was intermediate (33%) between those found in respiratory (43%) and musculoskeletal conditions (21%), which accounted for only 5% and 9% of all those screened respectively.
4. Mental Health Units
4.1 Of 332 adults screened, 19% were ‘malnourished’ (12% high risk, 7% medium risk).
4.2 The prevalence of ‘malnutrition‘ on admission to combined Acute and Long-stay/Rehabilitation units, which provided data on most of the population, was only 17%. This was lower that that found in only Acute (31%) or only Long-stay rehabilitation units (29%).
4.3 Less than half the units had a nutrition screening policy, but more than two thirds of patients were reported from such units. The prevalence of ‘malnutrition’ on admission (31%) was higher in units that had a screening policy than those that did not (17%), and in those that had access to a nutrition support team (39%) than those that did not (17%).
4.4. A low BMI (<20kg/m2) was present in 14% of patients (8% with a BMI < 18.5 kg/m2). Subjects aged 65 years and over (24%) had a greater prevalence of ‘malnutrition’ (24%) than those less than 65 years (14%).
5. A comparison across care settings
5.1 The prevalence of ‘malnutrition’ on admission to hospitals (28%) was similar to that found in care homes (30%), but higher than in mental health units (19%), although a much smaller number of subjects were reported from mental health units.
5.2 In all care settings most of the ‘malnutrition’ was high risk ‘malnutrition’.
5.3 The prevalence of ‘malnutrition’ amongst patients admitted to hospitals or care homes varied significantly according to source of admission, being lower in those that came from their own homes than from institutions (other wards, hospitals and care homes). The differences in mental health units were not significant.
5.4 In each setting women outnumbered men (ratio 1.1:1 in care homes and mental health units and 2.1:1 in care homes). Women were also older and had a slightly greater prevalence of ‘malnutrition’ than men.
5.5 BMI contributed to ‘MUST’ scores in about 4 out of 10 subjects in acute hospitals, 7 out of 10 in community hospitals and mental health units, and 9 out of 10 in care homes. Underweight (BMI <20kg/m2) was most common in care homes, affecting more than a quarter of residents. The mean BMI in care homes (23.4 kg/m2) was significantly lower (p <0.001) than in hospitals (26.2 kg/m2) and mental health units (25.7 kg/m2). In care homes underweight was more common than obesity (BMI >30kg/m2), whilst in hospitals and mental health units the reverse was true.
5.6 In each care setting ‘malnutrition’ was more common where a nutrition screening policy was in place than where it was not. Paradoxically, screening policies, which were most common in hospitals (89%), were associated with the least amount of weighing on admission (a little under half ). In contrast, in mental health units, where screening policies were least common (45%), weighing was apparently undertaken routinely on all patients. In care homes a nutrition screening policy was in place in at least 82% of them, and almost all residents were weighed on admission and regularly throughout their stay. The extent to which results of screening and weighing were linked into appropriate clinical care requires investigation.
6.1 Patients or residents admitted to all institutional care settings should be screened, and repeat measurements made at intervals according to care setting, using accurate and reliable instruments (see ‘MUST’ report).
6.2 The results of nutritional screening should be linked to care plans, which may vary according to local resources and policies.
6.3 Access to nutrition advice and nutrition support teams should be available.
6.4 Consistent strategies to detect, prevent, and treat malnutrition should be in place in all care settings, including the community, where most malnutrition originates.