Hospitalization as a turning point for sleep medication use in older adults: prospective cohort study.

BACKGROUND: Use of sleep medications as a result of hospitalization among older adults is common and has been shown to result in chronic use and increased risks for adverse effects such as falls and cognitive decline. However, few studies have explored in-hospital sleep medication use or disuse as a possible factor related to subsequent home use.
OBJECTIVE: The aim of this study was to examine changes in sleep medication use pre- to post-hospitalization as a function of in-hospital use.
METHODS: The study was designed as a prospective cohort study, and included 485 acute medical patients aged 70 years and older, hospitalized in a large, Israeli, teaching medical centre. Sleep medication use was assessed by patient interviews regarding patterns of use prior to, during and at 1 and 3 months after discharge. Post-discharge using patterns were assessed as a function of in-hospital discontinuation or initiation of sleep medications; background demographic and clinical characteristics were assessed as well. Logistic regressions were modelled separately for discontinuation and initiation of sleep medication use at each follow-up.
RESULTS: Of those patients who used sleep medications prior to admission, 37 (18% of 206 prior users) discontinued use during the hospital stay. Non-use of sleep medications during hospitalization was the main significant characteristic associated with post-hospitalization discontinuation among prior users, when comparing patients who continued with those who discontinued using sleep medications in bivariate analyses. Discontinuation was associated with an adjusted odds ratio (AOR, adjusted for cognitive status) of 3.91 (95% confidence interval [CI] 1.64, 9.30) for non-use at the 1-month follow-up. Of those who did not use sleep medications prior to admission, 39 (14% of 279 non-prior users) initiated use during hospitalization. Again, sleep medication initiation at time of hospitalization was the main correlate of change in post-hospitalization medication use status, when comparing post-discharge users and non-users, among the non-prior users. Hospital initiation of sleep medications was associated with an AOR (adjusted for levels of education and morbidity, readmission, and functional status) of 4.65 (95% CI 1.95, 11.09) for post-discharge use. Similar results were obtained for the 3-month follow-up, reaching significance levels only for the discontinuation group.
CONCLUSIONS: Though overall prevalence rates of sleep medication use pre- and post-hospitalization are fairly similar, rigorous scrutiny of the findings demonstrates that in-hospital sleep medication use and disuse may be a significant turning point both for initiation and discontinuation of sleep medications, especially in the short post-discharge time frame. Thus, in-hospital sleep medication prescribing policies should acknowledge the potential for changes in the post-discharge sleep medication regimen.