Can we predict which hospitalized patients are likely to be admitted to institutional long-term care facilities?

The rate of discharge from hospital to institutional long-term care facilities is highly variable, but the median is approximately 15%. The existing data on predictors are generally of poor quality and suffer from selective reporting. This study reinforces the observation most of us have made: The decision to institutionalize elderly patients is complicated. (LOE = 1a-)

Harrison JK, Walesby KE, Hamilton L, et al. Predicting discharge to institutional long-term care following acute hospitalisation: a systematic review and meta-analysis. Age Ageing 2017;46(4):1-12.  [PMID:28444124]

Systematic review

Other

Other

These authors systematically reviewed MEDLINE, EMBASE, and CINAHL to identify studies of hospitalized patients (in medical, surgical, or older adult wards) who were subsequently admitted directly to long-term care facilities. Two authors independently evaluated studies for inclusion, with discussion with a third author in the event of disagreements. Although the authors formally assessed the risk of bias of the included studies using a validated tool, they don't report how this was implemented. They also don't describe a process for identifying unpublished studies. The authors had several preplanned subgroup analyses: residential care versus nursing care, country of origin, younger than 65 years versus 65 and older, timing of assessment of predictors, dementia, and delirium. Ultimately they included 23 studies with more than 350,000 patients in Europe, North America, and Hong Kong (range: 94 to 262,345 patients). The duration of the studies varied from 3 months to 10 years. The authors found significant variability in terminology and in which variables were reported, and none of the studies provided operational definitions of long-term care institution. The rate of discharge to long-term care varied from 3% to 77% (median 15%). Eleven of the studies (12,642 patients) were of moderate to low quality and reported data in a manner conducive to performing meta-analysis. Long-term care placement was associated with increased age (odds ratio [OR] 1.02 per year of age; 95% CI 1.00 - 1.04), female sex (OR 1.41; 1.03 - 1.92), dementia or cognitive impairment (OR 2.14; 1.24 - 3.70), and functional dependency (OR 2.06; 1.58-2.69). The presence of delirium was not significantly associated with long-term care placement. The subgroup analyses were not very enlightening, mainly because of sparse reporting of relevant data.