PURPOSE: Planning home services for older people requires extensive knowledge about the progression of disabilities. Disability-based case-mix classifications identify meaningful groups of older people; yet transitions between profiles are mostly unknown. METHODS: Disability was assessed annually over four years with the Functional Autonomy Measurement System (SMAF) in 1410 older people at risk of functional decline aged 75 and over and living at home. The SMAF generates a case-mix classification of 14 Iso-SMAF profiles with progressive mean disability levels. Transitions made by older people were analyzed using a continuous-time, multi-state Markov model to estimate the probabilities of annual transitions into and out of each profile as well as the mean sojourn time in each profile. RESULTS: The probability of staying in a profile tended to decrease as profile severity increased. For profiles 5 and above, recovery to mild profiles 1, 2 and 3 was low, while annual probabilities of death and institutionalization were high (>0.10). The lower disability profiles (1 and 2) evidenced a mean profile sojourn time of over two years, contrary to sojourn times of 18 months or less with the other profiles. CONCLUSIONS: The probabilities are identifiable, indicating that a disability-based classification can characterize progression in older people. Since the required resources and costs are known for each profile, these probabilities are very helpful in planning home services for elderly populations.
PURPOSE: Planning home services for older people requires extensive knowledge about the progression of disabilities. Disability-based case-mix classifications identify meaningful groups of older people; yet transitions between profiles are mostly unknown. METHODS: Disability was assessed annually over four years with the Functional Autonomy Measurement System (SMAF) in 1410 older people at risk of functional decline aged 75 and over and living at home. The SMAF generates a case-mix classification of 14 Iso-SMAF profiles with progressive mean disability levels. Transitions made by older people were analyzed using a continuous-time, multi-state Markov model to estimate the probabilities of annual transitions into and out of each profile as well as the mean sojourn time in each profile. RESULTS: The probability of staying in a profile tended to decrease as profile severity increased. For profiles 5 and above, recovery to mild profiles 1, 2 and 3 was low, while annual probabilities of death and institutionalization were high (>0.10). The lower disability profiles (1 and 2) evidenced a mean profile sojourn time of over two years, contrary to sojourn times of 18 months or less with the other profiles. CONCLUSIONS: The probabilities are identifiable, indicating that a disability-based classification can characterize progression in older people. Since the required resources and costs are known for each profile, these probabilities are very helpful in planning home services for elderly populations.
Authors: Christy Cassarly; Renee' H Martin; Marc Chimowitz; Edsel A Peña; Viswanathan Ramakrishnan; Yuko Y Palesch Journal: Commun Stat Simul Comput Date: 2016-09-23 Impact factor: 1.118
Authors: Javier Jerez-Roig; Lidiane Maria de Brito Macedo Ferreira; José Rodolfo Torres de Araújo; Kenio Costa Lima Journal: PLoS One Date: 2017-05-11 Impact factor: 3.240