Rates of malnutrition identified during hospital admissions and related length of stay for people with mental illness: A population-wide data linkage study

BACKGROUND & AIMS: People with mental illness experience numerous risk factors for malnutrition, however, the rate and impact of malnutrition in this population group have not been examined. We aimed to determine the rates of malnutrition identified during hospital admissions and the associated length-of-stay in mental health service users compared to non-mental health service users. METHOD: In a population-wide data linkage study, malnutrition diagnoses made during hospital admissions and associated bed-days were examined for mental health service users compared to non-mental health service users between 2017 and 2019. Primary and secondary outcomes were crude and standardized rates of malnutrition and associated bed-days, respectively, presented as incidence and days per 100,000 person years. Sensitivity analyses assessed the contributions of: (i) eating disorder diagnoses, (ii) organic mental disorders, and (iii) admissions for mental health care. RESULTS: Mental health service users were admitted with malnutrition 3526 times per 100,000 person years. This rate was higher than for non-mental health service users, who experienced 593 admissions per 100, 000 person-years. In the fully adjusted model (aIRR), mental health service users had 7.76x (95% CI 7.74-7.77) higher rates of malnutrition being identified during an admission and 11.44x (95% CI 11.43-11.44) higher bed-days for those admissions compared to non-mental health service users. Excluding people with an eating disorder diagnosis had little effect on incidence of malnutrition (aIRR = 7.35, 95% CI 7.34-7.37) or bed-days (aIRR = 10.80, 95% CI 10.79-10.80). Excluding hospitalizations for organic mental disorders produced similar estimates (malnutrition aIRR = 7.65, 95% CI 7.63-7.67; bed-days aIRR = 11.24, 95% CI 11.23-11.24). Excluding malnutrition diagnosed during a non-organic mental health care admission reduced the incidence rates of malnutrition diagnoses being identified during hospital admissions (aIRR = 6.33, 95% CI 6.31-6.35) and bed-days (aIRR = 8.08, 95% CI 8.07-8.08), though rates remained significantly higher compared to non-mental health service users. CONCLUSION: Mental health service users are at higher risk for malnutrition being identified during a hospital admission, irrespective of whether the admission was for an episode of mental or physical health care. Strategies for prevention and early-identification are needed for this population group.