Targeting the underlying causes of undernutrition. Cost-effectiveness of a multifactorial personalized intervention in community-dwelling older adults: A randomized controlled trial: A randomized controlled trial

Background & aims: Undernutrition in old age is associated with increased morbidity, mortality and health care costs. Treatment by caloric supplementation results in weight gain, but compliance is poor in the long run. Few studies targeted underlying causes of undernutrition in community-dwelling older adults. This study aimed to evaluate the cost-effectiveness of a multifactorial personalized intervention focused on eliminating or managing the underlying causes of undernutrition to prevent and reduce undernutrition in comparison with usual care. Methods: A randomized controlled trial was performed among 155 community-dwelling older adults receiving home care with or at risk of undernutrition. The intervention included a personalized action plan and 6 months support. The control group received usual care. Body weight, and secondary outcomes were measured in both groups at baseline and 6 months follow-up. Multiple imputation, linear regression and generalized estimating equation analyses were used to analyze intervention effects. In the cost-effectiveness analyses regression models were bootstrapped to estimate statistical uncertainty. Results: This intervention showed no statistically significant effects on body weight, mid-upper arm circumference, grip strength, gait speed and 12-Item Short-Form Health Survey physical component scale as compared to usual care, but there was an effect on the 12-Item Short-Form Health Survey mental component scale (0-100) (β = 8.940, p. =0.001). Borderline significant intervention effects were found for both objective and subjective physical function measures, Short Physical Performance Battery (0-12) (β = 0.56, p. =0.08) and ADL-Barthel score (0-20) (β = 0.69, p. =0.09). Societal costs in the intervention group were statistically non-significantly lower than in the control group (mean difference -274; 95% CI -1111; 782). Cost-effectiveness acceptability curves showed that the probability of cost-effectiveness was 0.72 at a willingness-to-pay of 1000 €/kg weight gain and 0.80 at a willingness-to-pay of 20,000 €/quality-adjusted life year gained. Conclusions: This multifactorial personalized intervention showed a statistically non-significant effect and was not cost-effective on body-weight compared to usual care. We observed consistently beneficial treatment effects in the intervention group on all outcomes measures.