Twenty-five hospital and community-based ambulatory rehabilitation clinics in Ontario, Canada were recruited through an open call for participants by the RCA. Patients are referred to these publicly funded clinics by a physician or nurse practitioner. They must be either be aged (a) 65 years and older, (b) 19 years and younger, (c) recently discharged from hospital and require rehabilitation for the associated condition, illness, or injury, or (d) eligible for a social assistance program. Clinic treatment protocols, including duration and frequency, are at the discretion of the rehabilitation provider and based on individual patient need, goal attainment, and potential for improvement (21).

Like other interRAI comprehensive health assessments used as minimum data sets, the CRA is designed to be used with all ambulatory rehabilitation recipients. Thus, although participating clinics were asked to pre-commit to a minimum number of assessments for priority conditions such as hip or knee total joint replacement and stroke, the pilot study protocol did not specify patient exclusion criteria. Clinics were provided guidance on the suggested mode of administration of the self-report assessment for cognitively impaired patients. Interviewer assistance was indicated for patients with mild to moderate cognitive impairment (i.e., Cognitive Performance Scale (CPS) score of 2–3 (18)), and completion by and informal caregiver was indicated in the case of moderately severe or worse cognitive impairment (CPS 4+).

Participating clinics were encouraged to implement the CRA in a manner that was congruent with their current clinic workflow. Both patients and clinicians completed paper-based versions of the CRA on the first and last visit of treatment program. Regardless of impairment type, all patients and clinicians were instructed to complete all items on the assessment. Clinicians were instructed to assist patients to complete any unanswered self-report assessment items. In some clinics, volunteers and other staff members were available to assist patients that required assistance to complete the self-report portion of the assessment.

All analyses were performed for the overall sample and two sub-samples composed of individuals receiving rehabilitation following stroke (ICD-10-CA I64, I619, I634 and G459), and hip or knee total joint replacement (ICD-10-CA Z9660 and Z9661). We focused on these conditions because they represent distinct sub-populations in ambulatory rehabilitation that are part of Ontario's Health System Funding Reform model (5, 6).

Baseline patient characteristics, measured using information from both the patient and clinician responses on the assessment, were computed using frequency statistics for ordinal and nominal variables. Several interRAI summary scales can be computed using the information collected from the self-report and clinician rated portions of the assessment. We reported the Cognitive Performance Scale (range 0–6) (18) and the ADL Long Form Scale (range 0–28) (17) at baseline.

To demonstrate the CRA's utility as a measure of functional change, patient and clinician responses were compared between admission and discharge from the ambulatory rehabilitation program. Patients with missing data for a given assessment item on either assessment, most often because a discharge assessment was not completed, were omitted from the comparison. Scale scores were only calculated if all the required items for the scale were completed. Supplementary Material Table S2 compares the baseline clinical characteristics for patients where either a patient-self or clinician assessment was not completed.

Patient-reported difficulty on individual IADLs was measured using a 3-point response scale and compared between assessments using chi-square tests. This analysis was repeated for measures of mobility, including clinician-rated walking independence, self-reported stair difficulty, use of outdoor mobility aide, and farthest distance walked. Lastly, we compared clinician-rated timed 4-meter/13-foot walk test using paired t-tests.

Change in function was measured using the IADL Difficulty Scale at admission and discharge. This scale ranges from 0 to 14 points and is a sum of patient responses on the meal preparation, ordinary housework, managing finances, managing medications, phone use, shopping, and transportation items. Mean change between assessments was computed using paired t-tests. The Cohen's d statistic was also calculated to provide an effect size measure for the mean difference. Finally, we compared frequency distribution of the Pain Scale at admission and discharge.

A total of 709 patients were assessed with the CRA between January 1st, 2018, and December 31st, 2019. Four clinics submitted 50 or more admission assessments, nine clinics submitted 20–49 admission assessments and the remaining twelve clinics submitted fewer than 20 admission assessments.

The median age among participants in the sample was 67 years (IQR 60–74 years). Slightly more than half of the participants were female and almost two-thirds were married or had a partner/significant other. One-quarter of participants lived alone. Two-thirds of participants were retired or unemployed at admission, the majority of whom were not seeking work. A wide range of health conditions were represented in the sample, including patients receiving rehabilitation following total joint replacement (n = 210; knee n = 127, hip n = 83), stroke (n = 82), acquired brain injury (n = 32), spinal cord injury (n = 26), and other orthopaedic (n = 97) and neurological (n = 26) conditions (Table 1).

Approximately one-quarter of the total sample reported that their overall health at admission was “fair”, and an additional 4.8% reported that it was “poor”. Nearly three-quarters of the sample were cognitively intact at admission (CPS 0), with most of the other participants exhibiting mild cognitive impairment (CPS 1–2). Similarly, at admission, 70.8% of the sample were independent in all basic ADLs including personal hygiene, toilet use, locomotion and eating. Overall, 94.8% of participants received care from a physiotherapist, while only 31.6% and 13.6% of patients received care from an occupational and speech-language pathology therapists, respectively. The median number of sessions with each provider type ranged between 7 and 9 visits.

Overall, 557 patients (78.6%) completed a self-report discharge assessment, and for all but 15 patients (2.7%), an accompanying clinician discharge assessment was also completed. Clinicians completed discharge assessments in 30.9% of cases where a patient self-report discharge assessment was not completed. In cases where a discharge assessment was not completed, it was most often because the patient was discharged prematurely from rehabilitation due to a decline in health, the patient cancelled or failed to attend their final visit(s), or the rehabilitation program extended beyond the length of the study. Patients with a missing discharge self-report assessment were more likely to rate their overall health as “fair” or “poor”, report symptoms associated with depressive mood disorders, and experience shortness of breath at admission (Supplementary Material Table S2).

At discharge, nearly 80% of the overall sample reported that they achieved their rehabilitation goals or made progress in most areas. Compared to patients in the total joint replacement sub-group (39.9%), fewer patients in the stroke sub-group achieved all their rehabilitation goals (13.1%).

There was substantial improvement in IADLs. Relative to at admission, the percentage of patients in the overall sample and the stroke sub-group that reported some or great difficulty at discharge was significantly lower for all activities. For patients in the total joint replacement sub-group, this was only true for meal preparation, ordinary housework, medications, and shopping activities (Table 2). Among the overall sample, the mean improvement in the IADL Difficulty Scale (range 0–14) was 2.8 points (SD = 3.1, P < 0.001), which corresponds to a large effect size (Cohen's d = 0.9). This difference was small for the stroke sub-group (mean = 1.6 points, SD = 3.3, P < 0.001, Cohen's d = 0.4) and large for total joint replacement sub-group (mean = 4.3 points, SD = 3.2, P < 0.001, Cohen's d = 1.6). Figure 1 presents admission and discharge scores for this scale.

Between admission and discharge, patients in the overall sample walked an average of 3.2 s faster (SD = 3.2, P < 0.0001) on the timed 4-meter/13-foot walk test, which corresponds with a medium effect size (Cohen's d = 0.6). Patients receiving rehabilitation following stroke improved by an average of 0.9 s (SD = 4.6, P < 0.001, Cohen's d = 0.2 (small effect size)) and patients receiving rehabilitation following total joint replacement improved by an average of 4.4 s (SD = 5.1, P < 0.001, Cohen's d = 0.8 (large effect size)).

At discharge, fewer patients in the overall sample and total joint replacement sub-group reported difficulty with stairs and need for a mobility aide when outdoors. Participants in the overall sample and both diagnosis sub-groups were able to walk further at discharge, with more than one-third of participants capable of walking more than one kilometer at discharge. Similarly, participants reported that they left their residence more frequently at discharge and that their fear of falling was reduced, such that they were able to engage in more activities. This was true for patients receiving rehabilitation for both stroke and total joint replacement. Fear of falling limited indoor walking and other activities at the same frequency at admission and discharge among the stroke sub-group (Table 3).

At admission, 73.6% of the overall sample reported that they experienced daily pain compared to 54.5% of patients at discharge. Further, at admission, 25% of patients in the overall sample reported that the daily pain they experienced was severe or excruciating. The percentage of patients reporting pain at these intensities improved to 9.7% at discharge. Although patients in both the stroke and total joint replacement sub-samples experienced less frequent and intense pain at discharge, the difference was greatest for patients receiving rehabilitation following total joint replacement. For example, 32.7% reported severe or excruciating pain at admission compared to only 7.6% at discharge (Figure 2).