The implementation of a precision case management model in a Canadian inpatient rehabilitation center: The 12-months post-implementation findings of a quality improvement project

Abstract

Despite recommendations, few have reported on quality improvement initiatives to implement length of rehabilitation stay benchmarks, while actively monitoring functional outcomes. This article describes the development, implementation, and evaluation of a precision case management model across all inpatient rehabilitation client groups in a Canadian facility. To develop the length of rehabilitation-stay (LoRS) benchmarks, patient data was retrospectively analyzed. A severity specific method was used to stratify median length of stay. A target reduction on 8.6 days in LoRS was established. Functional discharge targets were also set and monitored at specific intervals via the Functional Independence Measure (FIM®). The implementation used an incremental quality improvement phased approach. Following 12-months, a statistically significant reduction in mean LoRS of 13.2 days was achieved, along with a small increase in FIM® change across all rehabilitation client groups. A similar pattern was seen across the three main client groups, where a LoRS reduction greater than the target was achieved, along with important improvements in LoRS efficiency. This study demonstrates how the implementation of a precision case management model can assist a facility in markedly reducing LoRS across inpatient groups, without compromising functional change or community discharge rates and begin its transformation to a value-based organization.

Keywords

Benchmarks length of stay rehabilitation quality improvement health systems

Introduction

There is a growing need for rehabilitation resources, with an estimated 2.41 billion people globally having conditions that could benefit from rehabilitation.¹ Inpatient rehabilitation health care costs are closely linked to patient length of stay.² Moreover, length of rehabilitation stay (LoRS) has been utilized to monitor patient flow through the care continuum, as a direct indicator of healthcare delivery efficiency and as a proxy for the intensity of rehabilitation service delivery in some patient populations.^3,4 According to the Canadian Institute for Health Information (CIHI), length of rehabilitation stay (LoRS) is defined as the number of days between a patient’s admission to and discharge from an inpatient rehabilitation facility.

In North America, the average LoRS for inpatient rehabilitation has become progressively shorter amongst various patient populations.^3–7 Benchmarking strategies are important in managing LoRS, holding clinicians accountable for the efficiency of resources consumed during care provision throughout the rehabilitation stay, and the discharge planning process.⁵ Length of stay has long been established as a key indicator of efficiency in acute care hospitals with many research and policy efforts focusing on its appropriate reduction.^8,9

However, it is important to acknowledge that significant differences exist between acute and rehabilitation patient care management. Rehabilitation facilities have greater control over patient admission, which means that if benchmarking strategies focus solely on mean LoRS reduction, the admission of less severely impaired patients could be encouraged; readily achieving the LoRS targets, distorting efficiency, and discriminating against those who recover more slowly.⁵

To minimize this risk, alternative strategies have been developed focused on severity specific LoRS benchmarking.^2,5 This strategy promotes reduction in LoRS stratified based on patient’s level of severity and using data already routinely collected.⁵ In Canada, the Canadian Institute for Health Information (CIHI) National Rehabilitation Reporting System (NRS) assigns patients to one of 17 Rehabilitation Client Groups (RCG) based on the primary reason for admission to inpatient rehabilitation (e.g., Stroke, Orthopedic Conditions, Spinal Cord Dysfunction). Upon admission to inpatient rehabilitation, the Functional Independent Measure (FIM®), a measure of disability, is administered.¹⁰ Composed of 18 items (i.e., 13 motor and 5 cognitive) scored on a 7-point Likert scale (i.e., 1 = complete dependence to 7 = complete independence), the FIM® provides a score ranging from 18–126, with higher scores indicative of greater functional independence.^10,11FIM® score also provides an estimate of the burden of care (i.e., hours of care required to assist a person with basic activities of daily living in the home setting).¹⁰Based on a combination of their RCG, admission motor and/or cognitive of FIM® scores and age, patients can be stratified to one of 83 Rehabilitation Patient Groups (RPG), which can then be used to establish severity specific LoRS benchmarks.²

To our knowledge, only two studies have developed severity specific LoRS based on severity specific stratified data. Meyer and his colleagues,⁵ used 3 years of retroactive facility specific inpatient stroke admission data, stratified by NRS established RPG, calculated mean FIM® gain, FIM® efficiency (FIM® gain by LoRS), piloted LoRS targets using a target FIM® efficiency of 0.75–1.0, and adjusted the median benchmarks one-year post implementation based on actual LoRS being achieved by greater than 40% of patients.⁵ Durand and her team derived their LoRS targets based on data from a specific retrospective cohort (2013–2014) of stroke patients admitted to their inpatient rehabilitation unit. Instead of using the CIHI NRS RPGs to stratify, patients were categorized into 4 severity subgroups based on FIM® admission scores (FIM® score > or = 100; FIM® 80–99; FIM® 60–79; FIM® < 60),¹² and median LoRS values for each patient were then calculated, and used as LoRS and discharge date targets.⁶ Additionally, Durand and colleagues also showed that the rehabilitation care team was 87% accurate in predicting which patients would achieve established discharge target dates. Despite the differences in methodology, these studies demonstrated that the implementation of a severity-specific benchmark strategy identifying discharge target dates for stroke patients admitted for inpatient rehabilitation facilities resulted in shorter LoRS (reduction of 5.9 days⁵ vs 11 days⁶), and decreased system costs, without compromising patients functional gains, sensorimotor outcomes, or rate of discharge to the community.^5,6

These studies also suggests that in order to be effective, benchmarking strategies should be combined with process improvement initiatives, to ensure adequate care continues to be provided and to encourage clinicians to continually improve the care they provide.⁵ In additional to its capability to accurately predict a patient’s functional outcomes upon completion of inpatient rehabilitation, functional rating thresholds for community discharge have been also been established for the FIM®.¹⁰ These FIM® ratings thresholds can be used by clinicians to establish goal targets critical to achieve for patient’s return to the community. These functional goal targets can be monitored through interim FIM® scores collected at pre-determined time intervals (e.g., 14 days post-admission) during patient’s admission.¹⁰ Known as precision case management, this model provides the opportunity to track patient’s progress toward these specific goals and estimated discharge date, evaluate efficiency of rehabilitation team effort (i.e., a gain of at least one point per day is expected), allows for real-time adjustments (e.g., LoRS, intervention dosing) to care and discharge plan based on the patient’s rehabilitation recovery trajectory and available supports (e.g., hours of assistance required).¹⁰ Unfortunately, evidence of the implementation and evaluation of a precision case management model is lacking in the literature.

The purpose of this study was to: (1) describe the methodology used to develop and implement the LoRS and functional (i.e., goal FIM®) benchmarks for the three most common Rehabilitation Client Groups admitted to our facility; and (2) to share the evaluation result of key indicators (i.e., LoRS, FIM® change, and LoRS efficiency) one year-post implementation of a precision case management model.

Methods

Setting

This quality improvement project was undertaken at a provincial stand-alone tertiary rehabilitation facility in Eastern Canada. The facility services a population of approximately 520,000 people, including many rural and remote communities. Admission rate is approximately 400 patients per year. Although other conditions are admitted, persons with strokes, spinal cord dysfunction, orthopedic conditions (e.g., hip fractures) make up the facility’s three main inpatient populations. The facility is a CIHI NRS participant organization, where sociodemographic and descriptive data about patients admitted to inpatient rehabilitation, along with FIM® scores and discharge destination information are collected at admission and discharge and submitted quarterly.

With the arrival of a new program leadership team and a vision focused on value-based health care framework, examination of NRS data revealed a worrisome historical trend of LoRS far exceeding those of other Canadian specialty rehabilitation facilities (46.7 days compared to 30.8 days for peer facilities). To address this, we embarked on a quality improvement project, hypothesizing that the implementation of a precision case management model (i.e., LoRS benchmarks and functional gain targets) for every rehabilitation client group admitted to the facility would reduce mean LoRS, without compromising FIM® change score, and result in improved LoRS efficiency. A continuous quality improvement process was adopted^13,14 and guided by Lewin’s 3-step Model of Change.¹⁵ One manager was designated as the project lead to identify project timelines and monitor activities.

Procedures

A Gantt chart of the main project implementation activities and timeline is presented in Figure 1 and described below.

Figure 1.

Gantt chart of the incremental implementation process and associated timelines.

Planning the improvement intervention

Problem analysis and staff education

Key indicators were calculated based on facility specific NRS data and included LoRS, FIM® change, and LoRS efficiency information at the program, RCG, and RPG levels, contrasted with those of peer facilities, and presented to staff on each unit. These presentations aimed to improve staff’s understanding of the concepts, our national standing, and targeted areas requiring improvement.

Benchmark development

To determine our LoRS targets, our NRS data for the previous five-year period (2014–2015 to 2018–2019) was reviewed, stratified based on client groups (i.e., Stroke, Spinal Cord Dysfunction, and Orthopedics). The median and average LoRS for each RPG for our facility as well as other similar specialty rehabilitation facilities across Canada were then calculated. Our aim was to set a realistic and achievable LoRS targets for our context, while also aligning with peer facilities. Managers, nursing and/or allied health staff were engaged in reviewing the LoRS calculation for each RPG to ensure appropriateness, which also helped educate them to the foundational concepts of the precision case management model.

Data collection and monitoring system

Learnings from a prior benchmarking project, which proved unsustainable due to reliance on manual processes for recording and analyzing data, led to the engagement with the NRS software vendor [3M™ Health Data Management (HDM) Rehab Module] in an automated solution. Leveraging staffs’ existing familiarity with recording relevant data using this software, we worked with the vendor to build a calculator to determine the RPG and to project a discharge date, as informed by benchmark development process. The software was then customized to allow staff to enter a goal FIM® score and multiple interim FIM® scores. The goal FIM® was the functional target for discharge based on burden of care, discharge location, and available supports. Interim FIM® scores were to be completed to track progress every 14 days after admission, until the patient was ready for discharge. A goal LoRS efficiency was calculated based on the change in FIM® score per day required to reach the goal FIM® by the projected discharge date. Interim LoRS efficiencies were calculated based on the improvement in FIM® score per day at each interim assessment.

To be able to integrate precision care management model into practice, effectively monitor implementation, and provide regular feedback to staff, a custom report was required, which could easily be generated and integrated into weekly unit interdisciplinary rounds. This report contained key implementation indicators such as: admission date, RPG, admission FIM® score, admission burden of care, projected discharge date, interim FIM® score, FIM® gain since admission, interim LOS efficiency, interim burden of care, goal FIM®, goal LOS efficiency, and goal burden of care. The customized report was integrated into weekly interprofessional rounds on each unit and served to provide immediate feedback on patient progress, goal attainment, and accuracy of LoRS targets. By assessing gain and efficiency at each interim assessment, the team could determine whether the patient was progressing as expected for functional goal achievement and would be ready for discharge on the projected date. For patients with a slower than expected trajectory (see Figure 2), teams were asked to consider the following: (1) Was there a medical issue preventing progress?; (2) Was a change in intervention (type, frequency and/or intensity) required?; (3) Did the projected discharge date and/or discharge plan need to change?. For patients progressing faster than expected (see Figure 3), earlier than the anticipated discharge date was to be considered.

Figure 2.

Example of patient progressing slower than anticipated. FIM® = patient’s actual rehabilitation trajectory; projected FIM® = patient’s rehabilitation trajectory to meet goal FIM® within target LOS; Goal FIM® = the functional level identified at admission as goal for discharge.

Figure 3.

Example of patient progressing faster than expected. FIM® = patient’s actual rehabilitation trajectory; projected FIM® = patient’s rehabilitation trajectory to meet goal FIM® within target LOS; Goal FIM® = the functional level identified at admission as goal for discharge.

Implementation plan

An incremental phased approach was utilized for the implementation. The stroke inpatient unit, which recently completed a quality improvement initiative, was chosen to pilot the implementation. The aim of the pilot was to identify and solve unanticipated issues prior to a broader site implementation. The pilot implementation started on April 1, 2020, with the initiation of unit staff to the new software. This initial phase lasted 3 months, prior to being scaled to the other units. The second phase incorporated the establishment of the goal FIM® and the goal efficiency concepts, which aligned with the precision case management model. The phase began in October 2020, with site wide implementation completed as of January 2021. As per Durand and colleagues protocol,⁶ the FIM® measurement was to be completed within the first 72 h.⁶ We established a goal FIM® identification timeframe of within 4 days of a patient’s admission.

Throughout the first year of this project, quarterly NRS data updates were shared with the teams using various formats (e.g., All Program Staff Meetings, Quality and Safety Leadership Meetings, Discipline and Unit Specific Meetings). As well, a program policy was developed to ensure all staff were aware of roles, responsibilities, and expectations.

Evaluation method and analysis

To evaluate the impact of the precision case management model, indicators derived from our facility aggregated NRS data from all patients with a completed admission and discharge FIM® seen as inpatients between April 1, 2019, to March 31, 2020 (designated as the Baseline Year) (n = 403) were compared to those seen from April 1, 2020, to March 31, 2021 (identified as Year 1) (n = 315). The indicators included: LoRS, FIM® change, and LoRS efficiency. LoRS (Active length of rehabilitation stay) included the total number of days from admission to discharge from our facility, excluding service interruptions (days that service is suspended due to a change in the patient’s health status) and days waiting for discharge (days after the patient was deemed ready for discharge). FIM® change represented the total FIM® score change from admission to date ready for discharge. The local Health Research Ethics Authority was consulted, yet as this evaluation was considered a quality assurance and program evaluation project it was exempt from review.

Descriptive statistics (i.e., frequencies and percentages) were used to examine the characteristics of the two cohort. Cohorts were then compared using a paired t-test for group differences based on the following characteristics: age, FIM® admission score, onset to admission interval (i.e., days from initial diagnosis to admission to rehabilitation facility). Where data was insufficient for statistical analysis (e.g., sex, goals achieved, identified discharge location), percentages were contrasted. These characteristics were also compared in the three main Rehabilitation Client Groups between cohorts. To assess whether the benchmarking strategy had the desired effect of reducing LoRS, without compromising functional gains paired t-tests were also used to compare the key indicators (LoRS, FIM® change, and LoRS efficiency) for all patients admitted to the facility in both years, for our three main Rehabilitation Clinic Groups (e.g., Stroke, Spinal Cord Injuries, Orthopedics), and subgroup RPGs (i.e., level of severity within a client group).

Results

Once set for all RPGs, the new targets represented an average LoRS reduction of 8.6 days over the most recent year’s (2018–2019) median LoRS. The data presented in Table 1 represent the indicators for the facility 1 year after initiating Phase 1 of the implementation plan. Overall, there were 22% fewer admissions in 2020–2021 compared to the Baseline year (i.e., 2019–2020). Patients admitted in both years were similar with respect to age, admission FIM^®, and onset to admission interval. There was a higher percentage of female patients in 2020–2021 compared to the previous year. Patients were also similar at discharge in terms of the percentage of patients meeting their rehabilitation goals and the identified discharge destination. Most patients meeting their goals were discharged home after inpatient rehabilitation in both years.

Table 1.

Patient demographics.

	2019–2020 (N = 403)^a	2020–2021 (N = 315)^a	p-value (95% CI)
Age, mean (SD)	69.9 (14.0)	71.1 (13.9)	0.3 (−3.3 to 0.9)
Sex, N (% female)	197 (48.9)	172 (54.6)	N/A
Admission FIM, mean (SD)	77.7 (20.2)	78.1 (19.0)	0.8 (−2.5–3.3)
Onset to admission interval days, mean (SD)	32.7 (34.6)	45.3 (244.3)	0.3 (−11.6–36.8)
Patients meeting rehab goals, N (%)	353 (87.6%)	277 (87.9)	N/A
Discharge destination, N (%)
Home without services	187 (46.4%)	154 (48.9%)	N/A
Home with services	85 (21.1%)	68 (21.6%)	N/A
Other community discharge	31 (7.7%)	21 (6.7%)	N/A
Residential care	38 (9.4%)	34 (10.8%)	N/A
Acute care	55 (13.6%)	35 (11.1%)	N/A
Other	7 (1.7%)	3 (1.0%)	N/A

^aThis total represents the total number of patients admitted to the facility within each of these years.

With respect to the main indicators measured across all client groups during Year 1 of the project (2020–2021) compared to the Baseline year (2019–2020), a statistically and clinically significant reduction in mean active LoRS of 13.2 days was achieved (see Table 2). A small increase in FIM® change was also noted. When combined, this resulted in a significant increase in LoRS efficiency. The same pattern was observed with patients admitted following a stroke; in this subgroup the mean decrease in LoRS was 12 days. Although the increase in FIM® change was not statistically significant, the increase in LoRS efficiency did achieve statistical significance. When the Stroke subgroup was further analysed at the RPG level, the three lowest functioning subgroups (RPG 1100, RPG 1110, RPG 1120) each showed statistically significant reductions in LoRS (48.7 vs 74.9, 46.7 vs 59.2, 27.5 vs 47.5, respectively). Furthermore, two subgroups (1110, 1120) also demonstrated increases in LoRS efficiency (0.65 vs 0.37, 0.87 vs 0.56, respectively) which reach statistical significance.

Table 2.

Main indicators for year 1 compared to baseline year.

	2019–2020 (Baseline)	2020–2021 (Year 1)
All patients	(n = 353)^a	(n = 278)^a	p-value (95% CI)
LOS in days Mean (SD)	46.7 (34.5)	33.5 (20.71)	0.0001 (−17.8 to −8.6)*
FIM change mean (SD)	22.1 (13.7)	23.8 (13.1)	0.12 (−0.42–3.8)
LOS efficiency mean (SD)	0.67 (0.55)	0.87 (0.54)	0.001 (0.11–0.29)*
Stroke	(n = 141)	(n = 109)	p-value (95% CI)
LOS in days Mean (SD)	46.7 (30.5)	34.7 (20.0)	0.0004 (−18.6 to −5.4)*
FIM change mean (SD)	19.6 (13.5)	22.8 (13.8)	0.07 (−0.2–6.6)
LOS efficiency mean (SD)	0.59 (0.55)	0.78 (0.52)	0.006 (0.055–0.32)*
Orthopedic conditions	(n = 100)	(n = 86)	p-value (95% CI)
LOS in days Mean (SD)	39.0 (28.1)	27.3 (19.0)	0.001 (−18.8 to −4.6)*
FIM change mean (SD)	24.8 (10.5)	23.4 (10.2)	0.4 (−4.4 to 1.6)
LOS efficiency mean (SD)	0.95 (0.57)	1.09 (0.59)	0.1 (−0.028–0.31)
Spinal cord dysfunction	(n = 37)	(n = 21)	p-value (95% CI)
LOS in days Mean (SD)	64.5 (43.5)	47.0 (23.7)	0.09 (−38.1 to 3.1)
FIM change mean (SD)	24.7 (15.5)	33.2 (16.8)	0.06 (−0.2–17.2)
LOS efficiency mean (SD)	0.50 (0.47)	0.69 (0.34)	0.1 (−0.044–0.42)

*p < 0.05.

These totals include all patients representative of all of the RCGs, who were admitted and discharged from our facility with a completed admission and discharge FIM®.

For patients admitted with an orthopedic condition, a significant decrease in mean active LoRS of 11.7 days was achieved. In this subgroup, the mean FIM® change was lower in the project year, but the difference was not statistically significant. Although LoRS efficiency did increase, it was not statistically different from Baseline Year.

Patients with spinal cord dysfunction were the smallest RCG examined in this study, with only 21 patients discharged in Year 1. For this client group, a 17.5 days decrease in mean LoRS was observed, along with an 8.5-point increase in mean FIM® change, resulting in a much-improved LoRS efficiency. Unfortunately, despite their magnitude, these changes did not achieve statistical significance.

Discussion

The purpose of this quality improvement study was to develop LoRS benchmarks and functional outcome targets, as per a precision case management model, and implement them across all rehabilitation clinical groups admitted to our facility, with the anticipated outcome of a reduction in mean LoRS without compromising patient functional outcomes or discharge destination. To date, no standardized process to develop LoRS benchmarks for inpatient rehabilitation clinical groups has been established across Canadian facilities, and only two facilities have published their methodologies which they applied only to persons admitted with Stroke. Similar to Meyer and colleagues,⁵ we chose to calculate severity-specific LoRS benchmarks from our facility NRS historical data and use the well-accepted Canadian stratification method. Functional items (i.e., goal FIM®) based on a patient’s community discharge, which also translates into the number of hours required by another person to support them in the basic activities of daily living, were also added to the strategy to help establish patient’s rehabilitation trajectory. To the best of our knowledge, the impacts of such a model have not previously been reported.

Following the first year of benchmark implementation, limited case-mix differences were observed in patients admitted as compared to the previous year. Yet a 22% reduction in admission rate was observed. It should be noted that this benchmark implementation was initiated just prior to the COVID-19 pandemic. The measures later imposed (e.g., staff redeployment, unit closure, capacity limitations, personal protective equipment rationing) impacted inpatient rehabilitation admission to our facility. Similar trends have been reported in other specialty rehabilitation centres in Canada and around the world.¹⁶ However, according to our NRS peer comparisons, the inpatient rehabilitation LoRS for these specialty rehabilitation centres did not decrease between 2019 and 2020 (30.8 days) and 2020–2021 (32.0 days).

While peer facilities did not experience a reduction in LoRS during this timeframe, at our facility the precision case management model implementation achieved the desired effect of reducing the mean LoRS across Rehabilitation Client Groups without compromising patient outcomes, as we hypothesized. In fact, our main findings indicate that the established 8.6 days reduction LoRS target was exceeded across all client groups, with a mean LoRS reduction of 13.2 days over the previous year, closing the LoRS gap between our facility and peer facilities from 15.9 days in 2019–2020 to 1.5 days in 2020–2021. A similar LoRS reduction was seen in our three main client groups (i.e., Stroke = 12 days, Orthopedic = 11.7 days, and Spinal Cord Dysfunction = 17.5 days). Based on a one-day reduction in mean LoRS translating to a CAD 2 million healthcare cost savings,² this study would have generated important cost savings (i.e., CAD 26.4 million). Furthermore, a small increase in FIM® change was also noted across certain client groups, with the Stroke and Spinal Cord Dysfunction groups demonstrating larger FIM® change. The important reductions achieved in mean LoRS and the small FIM® change resulted in significant increase in LoRS efficiency across client groups. Our mean LoRS reduction findings mirror those of previous published studies,^5,6 yet we achieved larger reduction and included a larger, more varied patient population. Like previous publications, not all findings in this study were statistically significant. The relatively small sample size, in particular in some groups (e.g., Spinal Cord Dysfunction) may explain the lack of statistical significance observed.

The implementation costs incurred were in the development of the benchmarks and the methodology, which proved easy to implement once the calculation and reporting software was created. The utilization of data already being collected, and a familiar software system further minimized the burden on clinicians. Although data collection and feedback are recognized as indispensable in quality improvement projects, it is imperative that systems fit the purpose of the project, be available from the start, and avoid imposing excessive burden or other unintended consequences on the staff.¹⁷

Throughout the initial implementation, skepticism existed among some rehabilitation team members in relation to the methodology used. In addition, adherence to timelines for the FIM® completion upon admission, interim FIM®, and goal FIM® required careful monitoring and reinforcement. Previous authors have also identified initial skepticism from rehabilitation staff.^5,6 However, skepticism seems to dissipate for most within a year when evidence emerges as to the relative ease of target attainment and other tangible clinical advantages are recognized such as: improved discharge planning, focus for discussion on timely goal attainment, facilitating targeted discharge date discussions, and regular team timely feedback to assess their own efficiency.^5,6 Further integration of the benchmarking strategy into inpatient rehabilitation unit operational structures (e.g., FIM® goal targeted interprofessional rounds and family meetings) is anticipated in Year Two. These changes will provide an opportunity for ongoing monitoring and evaluation, to explore if our results prove sustainable over time.

With the growing trend of health care organizations embarking on value-based healthcare journeys,^18,19 where value is defined as the health care outcome achieved per dollar spent,²⁰ a strong focus on generating maximum value for patients by assisting them to achieve the best possible outcomes and by doing it in the most cost-efficient manner is emerging, and the importance of meaningful comparison among facilities is increasing.²¹ Although not a perfect measure of severity, the RPG methodology does offer a meaningful representation of a person’s functioning. The implementation of a precision case management model ensures that clinicians do not simply aim for LoRS benchmarks, but instead are focused on achieving patients’ optimal functional performance (i.e., outcome) within an established time frame (i.e., cost proxy).²²

Furthermore, this study demonstrates how an organization can transform its approach from a data collection focus to a data-driven decision-making model focused on value-based healthcare in a short period of time. Keys to the success of this study were understanding the data available, the facility’s standing relative to peers, consistent feedback on progress made throughout the journey, and leveraging this new knowledge to drive change.^14,15 The additions of a project lead and program policy helped ensure that the implementation proceeded according to planned timelines and provided a resource to guide practice and monitor performance.

This study was designed to advance the process for developing LoRS benchmarks, within a precision case management model, and report on the findings of its implementation. However, its pragmatic nature does gives rise to some limitations. Due to its design, a direct cause-and-effect relationship between the intervention and the observed outcomes cannot be proven. Furthermore, our relatively small sample size and use of aggregate data (measures of variance were not available for some variables) limited our analysis of key indicators at the RPG level, for RCGs other than for the Stroke subgroup. Additionally, Meyer and colleagues identified that the RPG methodology may have some limitations and the inclusion of other factors specific to client groups into the algorithms may make for a more meaningful stratification based on severity.^3–5,23 Finally, the LoRS benchmarks are based on information from a single facility and thus limit their generalizability. However, it is hoped that the detailed methodology description provided in this article allows further testing of the implementation of precision case management and LoRS benchmarking strategies in other Canadian jurisdictions.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Disclaimer

Parts of this material are based on data and information compiled and provided by CIHI. However, the analyses, conclusions, opinions and statements expressed herein are those of the author, and not necessarily those of CIHI. The 18-item FIM® instrument referenced herein is the property of Uniform Data System for Medical Rehabilitation, a division of UB Foundation Activities, Inc.

ORCID iD

Michael Chislett

References

Cieza

Causey

Kamenov

, et al. Global estimates of the need for rehabilitation based on the global burden of disease study 2019: a systematic analysis for the global burden of disease study 2019. Lancet 2021; 396: 2006–2017.

Sutherland

Walker

. Challenges of rehabilitation case mix measurement in Ontario hospitals. Health Policy 2008; 85: 336–348.

Burns

Yee

Flett

, et al. Impact of benchmarking and clinical decision making tools on rehabilitation length of stay following spinal cord injury. Spinal Cord 2013; 51: 165–169.

Catharine Craven

Kurban

Farahani

, et al. Predicting rehabilitation length of stay in Canada: it’s not just about impairment. The Journal Spinal Cord Medicine 2017; 40: 676–686.

Meyer

Britt

McHale

, et al. Length of stay benchmarks for inpatient rehabilitation after stroke. Disabil Rehabilitation 2012; 34: 1077–1081.

Durand

D'Amours

Giroux

, et al. Benchmarking length of stay for inpatient stroke rehabilitation without adversely affecting functional outcomes. J Rehabil Med 2020; 52: 1–10.

McGilton

Campitelli

Bethell

, et al. Impact of dementia on patterns of home care following inpatient rehabilitation discharge for older adults after hip fractures. Arch Phys Med Rehabil 2021; 102(10):1972–1981.

Borghans

Kool

Lagoe

, et al. Fifty ways to reduce length of stay: an inventory of how hospital staff would reduce the length of stay in their hospital. Health Policy 2012; 104: 222–233.

Buttigieg

Abela

Pace

. Variables affecting hospital length of stay: a scoping review. J Health Organization Management 2018; 32: 463–493.

10.

Uniform Data System for Medical Rehabilitation . The FIM® Instrument: Its Background, Structure, and Usefulness. Buffalo, NY: UDSMR, 2012.

11.

Hamilton

Uniform Data Set for Medical Rehabilitation . The Data Management Service of the Uniform Data System for Medical Rehabilitation and the Center for Functional Assesment Research. New York, NY: State University of New York, 1990.

12.

Ween

Alexander

D’Esposito

, et al. Factors predictive of stroke outcome in a rehabilitation setting. Neurology 1996; 47: 388–392.

13.

Van Bokhoven

Kok

Van der Weijden

. Designing a quality improvement intervention: a systematic approach. BMJ Qual Saf 2003; 12: 215–220.

14.

Varkey

Reller

Resar

. Basics of quality improvement in health care. In Mayo Clinic Proceedings. Elsevier, 2007, pp. 735–739.

15.

Lewin

. Field Theory In Social Science: Selected Theoretical Papers. In: Cartwright

(ed), The Johns Hopkins University Press, 1951, pp. 522–524.

16.

Whiteson

Escalón

Maltser

, et al. Demonstrating the vital role of physiatry throughout the health care continuum: Lessons learned from the impacts of the COVID‐19 pandemic on inpatient rehabilitation. Pm R 2021; 13: 554–562.

17.

Dixon-Woods

McNicol

Martin

. Ten challenges in improving quality in healthcare: lessons from the Health Foundation’s programme evaluations and relevant literature. BMJ Quality Safety 2012; 21: 876–884.

18.

Colldén

Hellström

. Value-based healthcare translated: a complementary view of implementation. BMC Health Services Research 2018; 18: 1–11.

19.

Elf

Flink

Nilsson

, et al. The case of value-based healthcare for people living with complex long-term conditions. BMC Health Serv Res 2017; 17: 1–6.

20.

Porter

Teisberg

. Redefining Health Care: Creating Value-Based Competition on Results. Harvard business press, 2006.

21.

Steinmann

van de Bovenkamp

De Bont

, et al. Redefining value: a discourse analysis on value-based health care. BMC Health Services Research 2020; 20: 1–13.

22.

Evans

III Hwang

Nagarajan

. Physicians’ response to length-of-stay profiling. Med Care 1995; 33: 1106–1119.

23.

Sajobi

Lucyk

, et al. Systematic review of risk adjustment models of hospital length of stay (LOS). Med Care 2015; 53: 355–365.