The results from this study suggest that the typical older adult undergoing rehabilitation in a non-condition specific RITH program is 78 years of age with at least three co-morbidities and one previous hospital admission within the preceding 12 months. In addition, patients undergoing rehabilitation via RITH had generally high levels of function and cognition, and tended to live alone. Patients in the RITH program had an average of 12 days LOS and achieved an average of 10-point improvement in their FIM motor scores. Approximately one in five patients undergoing RITH had an unplanned ED representation within 30 days of hospital discharge. The need to have previous home modifications and the CCI were two factors that were associated with all three rehabilitation outcomes. These findings suggest that such routinely collected variables may be used to identify patients who are best suited for RITH and in return improve the efficacy of the service.
One of the most surprising findings from our study is the importance of carer status and home environment in influencing outcomes in RITH. The use of these factors may improve patient selection to improve the efficacy of RITH, as has been shown in other rehabilitation settings [32]. While previous studies evaluating the effectiveness of rehabilitation did find living arrangements, such as live-in social supports, to be an independent predictor of rehabilitation outcomes [25, 33], these studies did not breakdown which aspects of living arrangements and home environment impacted on rehabilitation outcomes. As shown in the results from the current study, both social and living environments played an integral role in influencing outcomes of RITH, suggesting that the patient’s social and living environment may be a more sensitive measure than other previously thought factors such as FIM scores. FIM motor scores on admission are frequently identified as an independent factor in influencing length of stay and discharge destination following rehabilitation [24–26, 34, 35]. Contrary to current literature, where FIM motor scores have been found to be associated with length of stay and ED representations following RITH, they were not a significant predictor for these outcomes in the current study. High average FIM scores on admission among patients included in this study may have limited the potential of improving FIM scores on discharge and the impact on length of stay and ED representations, which in turn may suggest that FIM admission scores may not be the most important factor in influencing rehabilitation outcomes for programs like RITH that are already targeting people with high levels of independence in their daily activity. Future studies may evaluate the association of caring arrangements and previous home modifications on rehabilitation outcomes for inpatient rehabilitation services where there is a lower functional baseline of admitted patients.
There are a number of previous studies which have identified that age, CCI and acute hospital LOS were found to be independent predictors of rehabilitation outcomes for people following hip fractures, strokes and traumatic brain injuries [33, 36]. Conversely, our study found that patients with a lower CCI had a longer stay in rehabilitation as compared with patients with higher CCI scores. Patients with a better health status may have stayed longer in rehabilitation in the current study due to having more rehabilitation goals to achieve in order to return back to pre-morbid function. The emergence of such findings also highlights the relevance of length of stay as a measure of outcome for rehabilitation. While length of stay of rehabilitation has been often deemed as an important indicator of the efficiency of a service [38], having a shorter length of stay in rehabilitation may be counterintuitive to the aim of rehabilitation, which is to assist people to regain pre-morbid function. Apart from focusing on length of stay, health executives should also pay attention to patient’s functional gain as measured using objective outcome measures like FIM when evaluating the effectiveness of rehabilitation, even though this can be challenging for many health services.
Other than suggesting the importance of assessing care arrangements and previous home modifications when ascertaining suitability of patients to participate in RITH, the results have also highlighted the importance of supporting an individual at home following rehabilitation. While as many as one in five people who completed RITH had an unexpected ED re-presentation, the reasons for these ED re-presentations were mostly associated with the lack of support available at home. Appropriate case management offered after the conclusion of rehabilitation may reduce the rate of ED re-presentations and the overall health service burden, as supported by the findings of a recently published systematic review which found that community-based case management reduces the likelihood of an ED presentation among older people living in the community [39].
Study Limitations
The large sample size of the study enabled sufficient statistical power for statistical analyses, however, the distribution of data for some of the variables such as FIM cognition subscore at time of admission limited the possibility of adding it to the analysis. This was likely due to the nature of the RITH program where it tends to take on patients of higher level of function, meaning less variances in scores. Further, the heterogeneity of admitted patients improves the external validity to other non-condition specific RITH services. The retrospective nature of this study may also impact on the accuracy of the data extracted from the histories, as it is plausible that missing data from the medical histories were actually collected by the clinician but not documented in the medical history. Lastly, the findings from this study was limited to only one RITH service. Future research should consider including more study sites to evaluate if the findings can be generalised to a wider population.