| dc.description.abstract | Purpose: The purpose of this dissertation was to characterize rehabilitative care in Ontario Complex Continuing Care hospitals and to examine the association of patient, structure, and process factors on functional outcomes and health state transitions.
Methods: First, a retrospective cross-sectional study of patients admitted to Ontario Complex Continuing Care hospitals between March 31st, 2011 and March 31st, 2016 (n = 100,778) was conducted to characterize rehabilitative care service utilization in this health service setting. The MDS 2.0 comprehensive health assessment that is completed at admission to this post-acute care service setting was used as the primary source of patient health status and service utilization data. A series of zero-inflated negative binomial regression models were fit to study the association of patient, facility and system-level factors on physical, occupational, and speech-language pathology therapy receipt and intensity.
Second, a retrospective study of patients admitted to Ontario Complex Continuing Care hospitals between January 1st, 2010 and March 31st, 2015 (n = 30,924) who were subsequently re-assessed with an interRAI assessment in either a Complex Continuing Care hospital, residential long-term care facility, or home care service setting was completed. This study aimed to describe patterns of functional gain following rehabilitation in Complex Continuing Care. The MDS 2.0 assessment that is completed at admission to Complex Continuing Care was used as the baseline measure of physical function, and was compared to measures collected with the next available MDS 2.0 or RAI-HC assessment completed in hospital, long-term care, or community care. A series of multivariate linear regression models were fit to study the association of patient, process, facility, and system-level factors on functional gain following rehabilitative care.
Third, a retrospective study of patients admitted to Ontario Complex Care hospitals between January 1st, 2010 and March 31st, 2015 (n = 76,132) that were either discharged from hospital or re-assessed with an interRAI assessment was performed to study factors associated with health transitions immediately following rehabilitative care in Complex Continuing Care hospital. The MDS 2.0 assessment that is completed at admission to Complex Continuing Care was used the primary source of patent health information and was used to stratify the sample into three baseline functional states. A multistate transition model was fit to study the association of patient, process, facility and system-level factors on health transitions at follow-up. Possible transition states included functional improvement and decline, discharge to community care, discharge to residential long-term care, discharge to acute care, and death.
Finally, a Markov chain multistate transition model was fit for a sample of Ontario Complex Continuing Care patients that were discharged to community care between January 1st, 2010 and January 1st, 2014 and assessed with a RAI-HC assessment within 105 days of discharge (transitions = 12,824). This analysis aimed to describe the effect of hospital-based rehabilitation therapy intensity on health state transitions after community discharge. Patients were classified into two initial states using functional measures from the RAI-HC assessment. Possible transition states of interest included functional improvement and decline, hospital admission, residential long-term care facility admission, death, and discontinuation of home care services.
Results: Overall, 79% of Ontario Complex Continuing Care patients received physical therapy, 69% received occupational therapy, and 16% received speech-language pathology therapy. The mean therapy intensity was 103 (SD = 92) minutes per week for physical therapy, 75 (SD = 87) minutes per week for occupational therapy, and 11 (SD = 37) minutes per week for speech-language pathology therapy. Patient-level factors including age, diagnosis group, baseline functional and cognitive status, medical instability, and rehabilitation potential were predictive of both receipt and intensity of therapy across each provider type. However, these associations were stronger for the receipt-component of the model, suggesting that after determining eligibility for rehabilitation, providers allocate therapy time based on other factors. After adjusting for patient characteristics, facility size, facility rurality, and region were significant factors across the models. This indicates that there are likely inter-facility and inter-region differences for rehabilitation service utilization.
On average, Complex Continuing Care patients improved by 3.24 points (Cohen's d=0.36) on the ADL-Long Form Scale between admission and follow-up. Statistically significant functional gain was observed for most activities of daily living; however, patients that were discharged to community care achieved greater gains than patients that were in hospital or residential long-term care at follow-up. Across baseline functional levels, the multivariate regression models explained between 19% and 23% of the variance in functional gain. Patient-level factors associated with functional outcomes included age, diagnosis group, cognitive status, and rehabilitation potential. Receipt of physical therapy was associated with functional gain; however, small amounts of additional physical therapy time were generally not associated with additional functional gain. Additionally, physical therapy time beyond 135 minutes per week did not result in additional gains in function. Receipt of occupational therapy resulted in gains in function for the least functionally impaired patients. However, more intensive occupational therapy was not associated with greater functional gains.
Within 105 days admission to a Complex Continuing Care hospital, 43% of patients were discharged to a community care setting, 11% were discharged to a long-term care facility, 8% were discharged to an acute care hospital, and 22% died. Among the 17% of patients that remained in Complex Continuing Care, 8% transitioned to a more impaired functional state and 16% transitioned to less impaired functional state. After adjusting for patient, facility, and system-level factors, patients that received more intensive physical therapy were generally more likely to be discharged to community care and hospital, and were less likely to die. Among patients that were not discharged, those that received more intensive physical therapy were generally more likely to transition to a less impaired functional state. Greater occupational therapy intensity was generally associated with greater odds of community discharge and lower odds of discharge to a residential long-term care facility.
Among patients that were discharged from Complex Continuing Care to home care, 13% of state transitions resulted in re-hospitalization, 5% resulted in residential long-term care admission, and 6% resulted in death. Among non-absorbing state transitions, 7% resulted in functional improvement and 6% resulted in functional decline. After adjusting for both patient and system-level factors, physical and occupational therapy intensity in Complex Continuing Care was not associated with greater odds of experiencing most health state transitions after discharge.
Conclusions: This dissertation represents the first comprehensive study of rehabilitation service patterns and outcomes for patients admitted to Ontario Complex Continuing Care hospitals. Through the use of national administrative health databases with near census-level coverage, this dissertation succeeds in answering research questions that span multiple health service settings along the continuum of care. The models that were developed in this dissertation lend support for the quality of rehabilitative care in Ontario Complex Continuing care hospitals; however, they suggest that there are opportunities to better allocate rehabilitation therapy for certain patient populations. Additionally, the results of this dissertation indicate that greater therapy intensity in Complex Continuing Care is associated with positive health state transitions; however, it is not protective over the long-term for patients that return to the community. | en |
