The usual prognosis for long-term disabled elderly is that they are no
t likely to improve or regain independent functioning. But computed tr
ansition probabilities reveal a significant proportion of improvement.
Based on the National Long-Term Care Surveys of 1982 and 1984, this s
tudy identifies the personal characteristics that are associated with
changed functional status between these 2 years and estimates how much
each characteristic contributes to this change. The study distinguish
es four mutually exclusive statuses of disabled older persons: improve
d, remaining unchanged, deteriorated-alive, and deteriorated-dead. It
shows that cognitive impairment reported in 1982, high-risk medical ev
ents that occurred during the year before the 1982 interview, and hosp
italization between 1982 and 1984 predict an increased risk of deterio
ration by 1984. However, similar estimates do not predict improvement.
Yet, data compatible with reported improvement suggest that acute med
ical problems might have caused a temporary worsening of functional st
atus. When the recovery from this status takes more than 3 months, the
ensuing change may be recorded as a long-term disability in the initi
al survey and as an improvement over time. However, long-term disabled
elderly seldom improve to such an extent that they regain complete in
dependence in physical functioning. In this study, most of the elderly
who improved ended with a functional status similar to that of person
s whose ADL disabilities remained unchanged over time.