If patients are not readily available, family members or others someti
mes are used as substitute informants when evaluating consumer satisfa
ction. Little is known, however, about the extent to which responses o
f patients and proxies are interchangeable. In this study, patients (N
= 225) or significant others (N = 115) provided an overall rating of
discharge plans, as well as information on psychosocial, health, and p
ost-hospital service related factors, at 3-4 weeks post-discharge. Whi
le no difference was found between mean plan ratings for the two group
s, substantial differences were found in factors predicting satisfacti
on for patients and for proxies. Results indicate that care should be
taken in using proxy ratings in the place of patient ratings, particul
arly when developing strategies for enhancing consumer satisfaction. F
indings are discussed as they relate to continuity of care initiatives
and programs for enhancement of desirable health utilization behavior
s, in the context of managed care.