Title : Association of the malnutrition universal screening tool and ASA classification with 30-day and 1-year mortality after hip fracture: An observational cohort study
Abstract:
Background: Malnutrition is common in older patients admitted with a hip fracture and may contribute to poor outcomes. ASA grade is already used routinely when assessing these patients, while MUST is generally recorded as part of nutritional screening. We looked at whether mortality differed according to MUST score and whether this association remained after taking ASA and other patient characteristics into account.
Methods: We included consecutive patients aged 60 years and over who presented with a hip fracture between January 2022 and June 2025. Clinical information had been collected prospectively. Patients with a periprosthetic or distal femoral fracture were excluded. Death within 30 days was the primary outcome, and mortality at 1 year was examined as a secondary outcome. Deaths were identified through our institutional hip-fracture review process. We could not establish whether this captured every death occurring outside the Trust. For the main analysis, patients were divided into MUST 0, MUST 1 and MUST ≥2. We also repeated the analysis using MUST as a continuous score. Logistic regression was used to adjust for age, sex, year of presentation and ASA category. Models with and without MUST were compared using likelihood-ratio tests. AUCs were calculated from the same dataset and therefore represent in-sample performance.
Results: We included 1,149 patients. Median age was 86 years (IQR 79–91); 790 (68.8%) were female and 359 (31.2%) were male. There were 107 deaths within 30 days (9.3%), and 332 patients (28.9%) had died by 1 year. MUST was available as a numeric score in 1,116 patients. Thirty-day mortality was 6.6% in patients with MUST 0, compared with 11.5% for MUST 1 and 14.3% for MUST ≥2. The same pattern was seen at 1 year, when mortality was 23.6%, 35.4% and 42.0%, respectively.
After adjustment for age, sex, calendar year and ASA, patients with MUST ≥2 had higher odds of death at 30 days than patients with MUST 0 (OR 2.06, 95% CI 1.25–3.38; p=0.004). This was also seen at 1 year (OR 2.36, 95% CI 1.68–3.31; p<0.001). For MUST 1, the association was significant at 1 year (OR 1.78, 95% CI 1.09–2.91; p=0.021), but not at 30 days (OR 1.90, 95% CI 0.91–3.96; p=0.087). The analysis using MUST as a continuous score gave similar findings. For each one-point increase in MUST, the OR was 1.51 for death within 30 days and 1.49 for death within 1 year (both p<0.001). Adding MUST to the adjusted model changed the AUC from 0.77 to 0.79 for 30-day mortality and from 0.71 to 0.74 for 1-year mortality. The likelihood-ratio tests also favoured models containing MUST (p=0.011 at 30 days and p<0.001 at 1 year).
Conclusion: Patients with higher MUST scores had higher mortality after hip fracture, even after adjustment for ASA, age, sex and calendar year. MUST added some information to the models, although the change in AUC was small. These findings show an association and do not establish MUST as a mortality prediction tool. Incomplete ascertainment of deaths outside the Trust and the use of complete cases should be considered when interpreting the results.
Keywords: Hip fracture, Mortality, Malnutrition, MUST, ASA, Nutritional Screening
