Abstract
Objective A shortage of rheumatologists has led to gaps in inflammatory arthritis (IA) care in Canada. Amplified in rural-remote communities, the number of rheumatologists practicing rurally has not been meaningfully increased, and alternate care strategies must be adopted. In this retrospective chart review, we describe the impact of a shared-care telerheumatology model using a community-embedded Advanced Clinician Practitioner in Arthritis Care (ACPAC)–extended role practitioner (ERP) and an urban-based rheumatologist.
Methods A rheumatologist and an ACPAC-ERP established a monthly half-day hub-and-spoke-telerheumatology clinic to care for patients with suspected IA, triaged by the ACPAC-ERP. Comprehensive initial assessments were conducted in-person by the ACPAC-ERP (spoke); investigations were completed prior to the telerheumatology visit. Subsequent collaborative visits occurred with the rheumatologist (hub) attending virtually. Retrospective analysis of demographics, time-to–key care indices, patient-reported outcomes, clinical data, and estimated travel savings was performed.
Results Data from 124 patients seen between January 2013 and January 2022 were collected; 98% (n = 494/504 visits) were virtual. The average age of patients at first visit was 55.6 years, and 75.8% were female. IA/connective tissue disease (CTD) was confirmed in 65% of patients. Mean time from primary care referral to ACPAC-ERP assessment was 52.5 days, and mean time from ACPAC-ERP assessment to the telerheumatology visit was 64.5 days. An estimated 493,470 km of patient-related travel was avoided.
Conclusion An ACPAC-ERP (spoke) and rheumatologist (hub) telerheumatology model of care assessing and managing patients with suspected IA in rural-remote Ontario was described. This model can be leveraged to increase capacity by delivering comprehensive virtual rheumatologic care in underserved communities.
- allied health personnel
- capacity building
- delivery of healthcare integration
- health services needs and demand
- healthcare evaluation: outcome and process assessment
- medically underserved area
Arthritis and musculoskeletal (MSK) disorders are among the most common chronic health conditions in Canada, with a national prevalence of 6 million,1 reflecting 16% of the population. Arthritis reflects a leading cause of morbidity, disability, and healthcare utilization in patients at all stages of life. Of these, the most common forms, and those with the greatest impact on the healthcare system, are osteoarthritis (OA) and rheumatoid arthritis (RA).2,3
Inflammatory arthritis (IA) affects patient health outcomes, productivity, and quality of life, with societal and system-level reverberations. A 2011 report published by the Arthritis Alliance of Canada revealed that the number of patients with RA in Canada doubled between 1990 and 2005 and is expected to double again from 272,000 individuals in 2010 to over 549,000 by 2040, representing 1.3% of the population. Causes of this increase include an aging population, rising rates of obesity, the effect of smoking, and increasing awareness of RA by physicians and patients.2 The Canadian Rheumatology Association (CRA) recommends that all patients with RA be cared for by trained and experienced healthcare professionals,4 but there has been no concomitant effective increase in the number of practicing rheumatologists nationwide. To avert an ever-expanding gap in the delivery of IA care, the current health service delivery model must be substantively adapted.
Access disparities widen the care gap for patients living in rural-remote communities.5 Socioeconomic status and ethnicity (notably, Indigenous peoples of Canada, in whom IA is more prevalent and more severe) further exacerbate this systemic issue.6,7 In Ontario, regional prevalence rates indicate a higher burden of both IA (including RA) and OA in northern communities.4 Populations living in Ontario in regions with decreased access to primary care practitioners (PCPs) also have limited access to rheumatology care.5,8
In 1 study, Indigenous peoples of Canada used IA health services at 50% the rate of non-Indigenous peoples.6,7,9,10 Decreased access to care for people living with IA in rural Ontario, increased burden of disease, escalated travel and accommodation costs, and resultant loss of income for patients and caregivers all reflect deficiencies in an already-fragile healthcare system.
Synchronous telehealth is an important tool used to deliver care. The coronavirus disease 2019 (COVID-19) pandemic has accelerated the development and uptake of this resource. Telerheumatology is an established model of care for adults and children with suspected IA living in rural-remote communities.11,12 One model involves an academic rheumatologist, located remotely (hub site), and a community-embedded clinician (spoke site) who collaborate virtually to deliver care to patients.13 Similar models have been successfully implemented in the care of other chronic diseases, such as diabetes mellitus14 and heart failure.15 Efforts have been made to increase capacity by engaging PCPs and allied health providers with Project ECHO (Extension for Community Healthcare Outcomes). However, this educational program does not provide direct patient care.16,17
A review of the adult and pediatric literature suggests that telerheumatology could provide substantial benefit, especially in rural-remote areas where access to rheumatology care is limited.12,13,18 For telerheumatology to work optimally, the spoke site practitioner should be trained, experienced, and competent in performing a comprehensive health history and physical examination specific to patients with rheumatic diseases.13 A recent patient and provider survey noted a high level of satisfaction with telerheumatology among patients.19 However, providers of care expressed dissatisfaction with suboptimal communication with patients, an inability to conduct a physical examination, and the technical limitations of telehealth.19
The Advanced Clinician Practitioner in Arthritis Care (ACPAC) program was developed in 2005 to promote the extension of roles of allied arthritis healthcare professionals in Canada.20 This University of Toronto certificate program has trained nearly 150 experienced healthcare practitioners (including physical and occupational therapists, nurses, and chiropractors). These ACPAC-trained extended role practitioners (ACPAC-ERPs) work interprofessionally and in collaboration with rheumatologists in 2 streams of arthritis care: triage and ongoing management.21,22
Our study describes a longitudinal hub-and-spoke telerheumatology shared-care model in rural-remote communities of North East Ontario as a proof of concept, including (1) patient demographics and clinical features of adults triaged with suspected IA, (2) the impact of the hub-and-spoke model on time-to-access-to-care, and (3) Ontario Ministry of Health (MOH)–funded travel savings.
METHODS
This study is a retrospective descriptive observational single cohort study of a shared-care virtual clinic using an ACPAC-ERP located in Espanola, Ontario, Canada (spoke site), from its inception in January 2013 to January 2022. Health care in Ontario over that period was managed by 13 Local Health Integrated Networks (LHINs). Telerheumatology consultation was provided to patients living in the underserviced North East (NE) LHIN by a rheumatologist located at St. Michael’s Hospital, Toronto (hub site).
Setting. The NE LHIN services approximately 557,000 people distributed across a rural-remote region of 400,000 km2 (150,000 mi2).23 The rheumatologist (RS) has been practicing for 38 years in a large urban academic health center. The ACPAC-ERP (JM) is an occupational therapist and an employee of Arthritis Society Canada with 21 years of experience in this role. The ACPAC-ERP practices with multiple rheumatologists at various sites within the NE LHIN, which includes the Espanola & Area Family Health Team. The spoke site was situated 465 km from the hub site. Administrative support was provided locally, through the rheumatologist’s office and Arthritis Society Canada, Ontario Division. Patients were seen in-person by the ACPAC-ERP and subsequently presented to the rheumatologist via the Ontario Telehealth Network. This clinic ran 4-6 hours/month, 11 months of the year. Across the interval, comprehensive virtual follow-up patient care was provided.
The ACPAC-ERP was trained by the rheumatologist (RS, ACPAC Program Director) to take a complete health history and perform a relevant rheumatology physical examination. Prior to initiating this shared-care telerheumatology clinic, the ACPAC-ERP and rheumatologist co-developed a well-defined protocol to collect objective health outcomes and patient-reported outcome measures (PROMs).
The ACPAC-ERP triaged patients referred with suspected IA/CTD. Ontario requires referrals for all specialty consultations by a physician or nurse practitioner. After triaging a referral, the ACPAC-ERP organized an in-person consultation, recorded the patient’s health history, performed a focused physical examination, and ensured that appropriate laboratory and imaging tests were completed in preparation for the telerheumatology visit. Patients triaged with non-IA were returned to the care of their PCP or referred to other healthcare professionals (eg, physiotherapists, Arthritis Society Canada, or Arthritis Rehabilitation and Education Program) for further management.
Inclusion criteria. New consultations seen by the ACPAC-ERP and the rheumatologist with suspected IA/CTD (ie, RA, systemic lupus erythematosus, polymyalgia rheumatica [PMR], vasculitis, psoriatic arthritis [PsA], reactive arthritis, gout, ankylosing spondylitis [AS]) and their subsequent follow-up visits were included.
Exclusion criteria. Patients who were assessed by the ACPAC-ERP as not suspected of IA/CTD, those who did not attend the telerheumatology visits with the rheumatologist (no shows), and patients who were previously known to the rheumatologist from her clinical practice at St. Michael’s Hospital were excluded.
Clinical data collection included. The following clinical data were collected:
Demographic information: Age at telerheumatology assessment, sex, ethnicity, employment, marital status, and the first 3 digits of the home postal code (census forward sortation areas).
Clinical information: Diagnoses made by the rheumatologist as per first and final Ontario Health Insurance Plan billing codes, comorbid conditions, physical examination (number of swollen and tender joints), function (grip strength), laboratory results (rheumatoid factor [positive once], antinuclear antibodies [positive at any visit], anticyclic citrullinated peptide antibodies [positive once]), smoking status at first visit, family health, and social history.
PROMs: Pain visual analog scale (VAS), sleep VAS, fatigue VAS, patient global assessment (PtGA) VAS, morning stiffness in minutes, and Multidimensional Health Assessment Questionnaire (MHAQ) scores were collected at each visit.
Referral, assessment, and follow-up information:
Number of follow-up visits: Total number of visits and follow-up visits with the rheumatologist, including the number of visits conducted in-person.
Time-to–key clinical indices: Days from PCP referral to date of initial ACPAC-ERP assessment (T1); number of days between initial ACPAC-ERP assessment and first consultation with the rheumatologist via telerheumatology (T2); and time from telerheumatology consultation to prescription for a disease-modifying antirheumatic drug (DMARD) for those determined to have IA and not currently on treatment (T3; Figure 1).
Travel information: Geospatial information for individual patients was obtained using regional postal codes collected through demographic data. Calculations of hypothetical distances between these locations and St. Michael’s Hospital were based on the most efficient and cost-effective driving route (ie, no use of toll routes or ferries) using Google Maps. Aggregate data were used to calculate distances not traveled due to the virtual care delivery model.
Time-to–key clinical indices. ACPAC-ERP: Advanced Clinician Practitioner in Arthritis Care–trained extended role practitioner; DMARD: disease-modifying antirheumatic drug; PCP: primary care practitioner.
Estimation of the cost savings of this telerheumatology shared model of care was based on MOH travel grants that would have been reimbursed to patients should they have traveled for in-person care. Travel grant data from the MOH and Long-Term Care Northern Health Travel Grant (NHTG)24 was used as a reference to predict average costs associated with a 2-day visit to Toronto to see a specialist. NHTGs are applicable for a patient’s travel distance of at least 200 km to the closest specialist or ministry-funded healthcare facility that can provide the required services. Travel by car is reimbursed at CAD $0.41/km, with a deductible of 100 km per trip, and CAD $100 lodging per night.
Data extraction from electronic health records was performed by the research coordinator (TI). Additional data extraction from physical charts was performed by the rheumatologist (RS) and ACPAC-ERP (JM) when data were not available electronically. The research coordinator had open communication with the rheumatologist and ACPAC-ERP across the data collection period. Data were anonymized for analysis and entered into a Microsoft Access database. SAS 9.425 was used for all analyses. Descriptive statistics were generated.
This project was formally reviewed by the institutional authorities of Unity Health Toronto and Arthritis Society Canada. The project was deemed a quality improvement initiative, and thus did not require research ethics board approval or written informed consent from participants.
RESULTS
One hundred twenty-eight patients were considered for the study. Three patients did not attend their telerheumatology visit with the rheumatologist and were excluded. One patient, previously in the care of the rheumatologist prior to attending the telerheumatology clinic, was also excluded. Data from 124 patients were included in the final dataset.
The average age of patients at the time of their first telerheumatology visit was 55.6 (SD 16.2) years, 75.8% were female, and 21% identified as Indigenous peoples of Canada (Table 1). In terms of diagnosis, 65% (n = 80) were confirmed to have IA/CTD during their first telerheumatology visit (36 RA, 16 CTD, 10 PsA/reactive arthritis, 7 gout, 7 AS, 4 vasculitis/PMR). Non-IA diagnoses at the first telerheumatology visit included OA, frozen shoulder, disc disease, tendonitis/tenosynovitis, and MSK disorders not yet determined. The cohort presented with the following comorbid conditions: hypertension (32.3%), depression (19.4%), diabetes mellitus (14.5%), anxiety (13.7%), cancer (8.9%), and kidney disease (5.7%); 32.3% of patients reported they were smokers at the time of the first visit. At the initial visit, PROMs revealed moderate levels of pain, morning stiffness, PtGA, fatigue, and sleep disturbance (Table 2). At the final visit recorded for this study, 76.4% had the same rheumatologic diagnosis as at the first visit (Table 2).
Demographic description of cohort (n = 124).
Clinical characteristics.
A total of 504 visits were included in the dataset, with a range of 1-20 visits per patient (mean 4.0 visits/patient). Of the 504 visits, 98% were conducted virtually, with only 10 (2%) in-person visits. Date of initial referral to the ACPAC-ERP (T1) was available for 117 patients. Mean time from PCP referral to ACPAC-ERP assessment was 52.5 days (75% seen within 76 days), and mean time from ACPAC assessment to the telerheumatology (T2) visit was 64.5 days (75% seen within 85 days). Of the patients seen via telerheumatology, 21 were deemed appropriate for new DMARD therapy. Mean time-to-DMARD initiation in these patients was 52.0 days (75% received treatment within 70 days; Table 3).
Referral, assessment, and follow-up visits.
All patients included in this study were eligible for a NHTG based on the distance calculated between their home and St. Michael’s Hospital, Toronto. Individual patient data were used to calculate the hypothetical (most efficient and cost-effective driving route using Google Maps) return distance traveling by car for an in-person visit with the rheumatologist. The number of telerheumatology visits per patient was multiplied by these hypothetical distances. The aggregate sum of distance not traveled was 493,470 km (based on 504 total encounters with the rheumatologist minus 10 in-person visits = 494 telerheumatology visits). The loci where > 5 patients resided is represented in Figure 2, with the remainder distributed in other communities across the NE LHIN.
NE LHIN-based study population accessing health services via a shared-care telerheumatology model. NE: North East; LHIN: Local Health Integrated Networks.
A theoretical estimate of the NHTG cost (in CAD) was calculated based on the total number of patient return-trip visits that would have been incurred should these patients have been followed by the rheumatologist in-person. This was calculated by subtracting the requisite deductible from the sum of the number of kilometers not traveled, multiplying this number by the rate reimbursed per kilometer, and then adding the base cost allotted for 2 nights lodging in Toronto as follows:
DISCUSSION
To our knowledge, this is the first description of a shared hub (rheumatologist) and spoke (ACPAC-ERP) telerheumatology model of care in patients living in rural-remote North East Ontario. The NE LHIN has 0.2 rheumatologists/100,000 population—the lowest in Ontario—resulting in a lack of access to arthritis specialty care.26 In the present study, patients with arthritis and MSK issues were referred to the ACPAC-ERP for triage and those clinically suspected of IA/CTD were presented to the rheumatologist in the telerheumatology clinic.
Patients (n = 124) were followed for 9 years in the telerheumatology clinic. All initial consultations were done virtually. Of the 504 visits, 98% were virtual, with 10 visits (2%) seen in-person by the rheumatologist for follow-up in Toronto. Reasons for in-person visits included expedited joint aspiration/steroid injections and clarification of diagnosis in complex cases. PROMs collected include MHAQ, VAS for pain, fatigue, sleep, PtGA, and morning stiffness. The intent was to provide self-reported patient population characteristics at initial visit. Further studies could address changes in these variables over time. A data linkage study of primary care electronic medical records and administrative data of 2430 patients referred to rheumatologists in Ontario between 2000 and 2013 had a mean age of 53.0 years and 69.2% were female, similar to the findings in this study’s cohort.26 In Canada, over time, rheumatologists are seeing a higher number of patients presenting with inflammatory diseases.27 The population cared for in our telerheumatology shared-care model is representative of that seen in a large population-based study.
Of our patient cohort, 65% were confirmed to have IA/CTD and were followed longitudinally in this clinic. Studies have shown high correlation between rheumatologists and ACPAC-ERPs in assessment (joint counts) and IA diagnosis.28,29 A community-based ACPAC-ERP triaged new referrals for suspected IA and, when compared to a rheumatologist assessment, had a 90.7% positive predictive value and 100% negative predictive value, without missing a single patient with IA.28 They concluded that a well-trained and experienced ACPAC-ERP could shorten the time-to–rheumatologist assessment and time-to–treatment decision for patients with suspected IA. In 2021, a solo rheumatologist/ACPAC-ERP co-management model was noted to increase capacity in community rheumatology clinics for patients newly diagnosed with IA while maintaining confidence and satisfaction with their care.30
Our study population is culturally unique, with 21% of patients identified as Indigenous peoples of Canada, compared to 5% of Canada’s overall population31 and 13.6% of the NE LHIN population.23 Our model shows a reach that has been inclusive to a population known to be underserved and overrepresented in chronic disease prevalence including IA.
The Wait Time Alliance released consensus-based rheumatology wait-time benchmarks for IA.32 These benchmarks were developed and endorsed by the CRA and Arthritis Alliance of Canada. An observational study involving electronic medical records from Ontario PCPs (to provide accurate dates of referral requests) linked with health administrative data (to provide accurate dates of encounters with rheumatologists) was conducted between 2000 and 2013.26 Median wait time from referral to rheumatologist consultation was 74 (IQR 27-101) days; it was 66 (IQR 18-84) days for systemic autoimmune rheumatic diseases. In our study, the mean number of days from PCP referral to ACPAC-ERP was 52.5 (SD 43.5), with 75% seen within 76 days.
Time from ACPAC-ERP assessment to rheumatology consultation was 64.5 days (SD 57.7), with 75% of patients seen within 85 days. This wait time exceeds the CRA benchmark for IA assessment by a rheumatologist. Reasons for this include (1) JM was the sole ACPAC-ERP triaging and servicing the NE LHIN, which comprises 30.6% of Ontario’s land mass23; (2) this clinic ran only 4 to 6 hours per month, 11 months of the year; and (3) patients frequently missed clinics because of regional weather-related or adherence issues—these patients were subsequently followed up and rescheduled, thus resulting in longer wait times. This reflects a proof of concept for a single rheumatologist–ACPAC-ERP dyad. Wait times could be substantially improved by integrating a network of rheumatologists working effectively within an administrative structure, which currently does not exist in the NE LHIN. Augmentation and amplification of this model presents an opportunity for increased capacity for care.
Satisfaction is high for patients receiving rheumatology care via telehealth.33 However, in 1 study, rheumatologists expressed serious concern about the quality of care delivered virtually, with only 17.6% expressing satisfaction and 82.4% citing inferiority of telehealth over in-person assessments due to an inability to perform a physical examination.19 Our model addresses this deficiency as the ACPAC-ERP effectively fulfills this critical component of the patient’s assessment. Through this synchronous interprofessional telehealth model employing ACPAC-ERPs, rheumatologists can provide virtual comprehensive care to patients in a rural-remote setting.
Currently in Canada, there are an insufficient number of rheumatologists employed to optimally care for patients with IA/CTD.34 Indeed, a recent paper published by Kulhawy-Wibe et al35 calculated an additional 194 full-time rheumatologists would be required to meet the CRA’s workforce benchmark of 1 full-time equivalent rheumatologist to support a population of 75,000. This is unlikely to be realized in the foreseeable future, given the inequities of care due to regional disparities and nationwide rheumatology workforce issues.35
The estimated $280,868.70 cost savings in this study represents a minimal fiscal value of this telerheumatology model of care. This estimate was based on direct MOH costs, specifically with respect to travel grant utilization for a 2-day visit to Toronto to see a specialist. This grossly underestimates costs related to travel by car as well as basic accommodation in a large urban center such as Toronto. Air travel would involve significantly higher costs and the North East region of Ontario is not well-served by alternate means of transportation. Patients living in rural-remote communities are disadvantaged by being required to materially supplement their travel expenses in order to access in-person specialty care. There are other tangible and significant cost savings, both direct and indirect, that need to be further captured in a more rigorous economic analysis.
Approximately half a million kilometers (493,470 km) of travel was avoided for patients involved in this telerheumatology study, representing a significantly reduced carbon footprint and environmental impact relative to that incurred through in-person care. The estimated cost savings cited in the study were realized working with a single ACPAC-ERP for 11 clinics per year over 9 years. Although this ACPAC-ERP worked with additional rheumatologists via telehealth, these data are not available to us. If taken into account, these virtual clinics would significantly augment the estimated savings. The cost savings by this ACPAC-ERP shared model of care provides a solid fiscal argument to support ACPAC-ERP roles as well as the scaling up of this model.
We acknowledge there are limitations to this study. As this retrospective cohort study presented descriptive data without a comparator, drawing meaningful conclusions is challenging. Our goal was first to describe this novel model of care. We look forward to designing prospective analyses in the future as the model is further scaled up.
One limitation to this study is that it represents 1 interprofessional dyad of an experienced rheumatologist and an ACPAC-ERP who have worked collaboratively for years. A further limitation involves the triage process, which could have excluded other patients with IA/CTD. We believe that this is minimal due to previously published high level of concordance of ACPAC-ERP and rheumatologists’ assessments.28,29,36
Another limitation is that we did not have 2 independent reviewers to extract data, although we did have multiple meetings to review the data and establish consensus.
Finally, our cost savings extrapolation was limited to publicly available data on the NHTG. To capture direct and indirect cost savings, a full economic analysis of this model is warranted.
In conclusion, we present an ACPAC-ERP–rheumatologist interprofessional shared-care telerheumatology model that delivered longitudinal care to 124 patients with suspected IA/CTD in rural-remote Ontario. The rheumatologist, in collaboration with the ACPAC-ERP, followed 98% of the patients virtually, with only 2% of patients requiring in-person follow-up visits. The ACPAC-ERP–rheumatologist shared-care model provided comprehensive care to the right patients, in the right place, and at the right time, thus embodying the tenets of quality care.
A dearth of rheumatologists in Northern Ontario is likely to persist. The described hub-and-spoke telerheumatology shared-care model was successful by many metrics, and thus reflects a viable strategy to address this widening care gap. We have demonstrated estimated cost savings, which should be further borne out with more robust economic modeling. Our clinic operated in a resource-limited setting providing essential rheumatology care in a vastly underserviced region. Expanding and sustaining this model of care with continued generation and fiscal support of highly trained ACPAC-ERPs would provide improved local access to rheumatology care to patients living in rural-remote Ontario.
Footnotes
This work was supported by an investigator-initiated grant from Pfizer Canada.
The authors declare no conflicts of interest relevant to this article.
- Accepted for publication May 21, 2024.
- Copyright © 2024 by the Journal of Rheumatology
This is an Open Access article, which permits use, distribution, and reproduction, without modification, provided the original article is correctly cited and is not used for commercial purposes.