Article Text
PDF
Abstract
Objectives To examine the cost-effectiveness of Multi-specialty INterprofessional Team (MINT) Memory Clinic care in comparison to the provision of usual care.
Design Using a Markov-based state transition model, we performed a cost-utility (costs and quality-adjusted life years, QALY) analysis of MINT Memory Clinic care and usual care not involving MINT Memory Clinics.
Setting A primary care-based Memory Clinic in Ontario, Canada.
Participants The analysis included data from a sample of 229 patients assessed in the MINT Memory Clinic between January 2019 and January 2021.
Primary outcome measures Effectiveness as measured in QALY, costs (in Canadian dollars) and the incremental cost-effectiveness ratio calculated as the incremental cost per QALY gained between MINT Memory Clinics versus usual care.
Results MINT Memory Clinics were found to be less expensive ($C51 496 (95% Crl $C4806 to $C119 367) while slightly improving quality of life (+0.43 (95 Crl 0.01 to 1.24) QALY) compared with usual care. The probabilistic analysis showed that MINT Memory Clinics were the superior treatment compared with usual care 98% of the time. Variation in age was found to have the greatest impact on cost-effectiveness as patients may benefit from the MINT Memory Clinics more if they receive care beginning at a younger age.
Conclusion Multispecialty interprofessional memory clinic care is less costly and more effective compared with usual care and early access to care significantly reduces care costs over time. The results of this economic evaluation can inform decision-making and improvements to health system design, resource allocation and care experience for persons living with dementia. Specifically, widespread scaling of MINT Memory Clinics into existing primary care systems may assist with improving quality and access to memory care services while decreasing the growing economic and social burden of dementia.
- health economics
- primary care
- dementia
Data availability statement
Data are available upon reasonable request. Data are available upon reasonable request. The data that support the findings of this study are not publicly available due to them containing information that could compromise participant privacy. Deidentified, limited data will be shared by the corresponding author upon request.
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
Strengths and limitations of this study
This study is an economic evaluation of a multispecialty interprofessional team model of dementia care in Canada for which there is limited economic evaluation data.
This economic evaluation was conducted consistent with best practice methods and suggested that Multi-specialty INterprofessional Team (MINT) Memory Clinic care is less costly and more effective compared with usual care in 98% of the time.
The lack of existing research regarding a comparative usual care group for persons with dementia living in Canada limited us to using available data from different countries and healthcare systems thus comparability between MINT Memory Clinic care and usual care may be limited.
As our data are most relevant to Canada, and in a particular to community care settings, it may be difficult to generalise to other jurisdictions due to differences in healthcare systems.
Introduction
Globally, dementia is one of the major causes of disability and dependency among older persons.1 In addition to the significant impact on the quality of life for individuals diagnosed with dementia and their families, dementia also has significant economic implications for healthcare systems. In Canada, combined healthcare system and out-of-pocket caregiving costs totalled $10.4 billion in 2016 and is expected to increase to $16.6 billion by 2031.2 In 2015, the total societal cost of dementia worldwide in terms of direct medical, social care and informal care costs was estimated to be US$818 billion.1
Primary care clinicians are often the first point of contact for individuals experiencing memory concerns. Given the challenges experienced in diagnosing and managing this complex disorder within the time constraints in busy family practice, persons with memory concerns have historically been referred for specialist care.3 There is increasing recognition of the need for primary care to take on greater responsibility for early diagnosis, management and ongoing dementia care throughout the disease process.4 There is particular interest in strengthening dementia care in primary care with the aim of supporting those with dementia to live at home for as long as possible and to avoid hospitalisation and institutionalisation.4
Collaborative, multidisciplinary team approaches to healthcare represent a significant opportunity to provide patient-centred care, improve health outcomes and patients’ experience with care.5 6 The Multi-specialty INterprofessional Team (MINT) Memory Clinic care model (formerly Primary Care Collaborative Memory Clinics) aims to improve assessment, diagnosis and management of dementia in primary care.7 Integrating specialist and community care for the most complex of cases, this model supports person-centred care that is experienced by patients and caregivers as comprehensive, coordinated, timely and accessible from one location, close to home.7–10 Memory clinics are usually located within the same location as their family physician. Within this care model, patients with memory concerns are referred by their family physician to the MINT Memory Clinic for comprehensive assessment and care planning conducted by an interprofessional team consisting of specially trained family physicians, nurses and other healthcare professionals (eg, social workers, pharmacists, occupational therapists), and representatives from local community services (Alzheimer Society, home care, behavioural support services) as available.11 Assessments are conducted with all team members working together in a coordinated and collaborative manner to complete the assessment at the same visit, formulate a diagnosis and develop an integrated, individualised care plan based on patient and caregiver preferences and needs. Using a shared care approach, MINT Memory Clinic team members work with the patient’s own family physician over the course of the disease to ensure that changes in care needs are identified and met. Key model components include integration of geriatric specialists to provide consultative support, ongoing capacity building support and team integration and coordination of community support services.6
The MINT Memory Clinic model exists in over 100 primary care settings across Ontario and is currently being expanded to other provinces across the country. Published evaluative studies have demonstrated improved clinical practice and quality of dementia care, improved access to health and social services, enhanced care experiences for patients and their caregivers, healthcare provider satisfaction with dementia care and improved collaboration among health professionals.6 8–10 To assess the quality of care provided in MINT Memory Clinics, two geriatricians independently reviewed 50 medical charts from five Memory Clinics using a chart audit tool developed by the Ontario of College of Physicians and Surgeons of Ontario.12 This chart audit revealed a high level of agreement among the geriatricians (kappa coefficient=0.86) with the diagnosis and management provided by the clinics, verifying the quality of care provided.10 A significant healthcare system outcome associated with this care model has been the highly efficient use of limited available specialist resources with a less than 10% referral rate to specialists, reduced pressure on specialist wait lists and delayed institutionalisation.7 10 13 14 The purpose of this study was to examine the cost-effectiveness of the MINT Memory Clinic care model in comparison to the provision of usual dementia care in Ontario, Canada.
Methods
Study design
We developed a Markov-based state-transition model to determine the cost-effectiveness of MINT Memory clinics for patients with cognitive impairment (CI) in Ontario, Canada using cost-utility analysis. We adopted a public payer perspective (provincial Ministry of Health), used a lifetime time horizon, and a 1.5% discount rate for our analysis based on Canadian economic evaluation guidelines.15 An overview of our methodology is presented as follows and additional information can be found in online supplemental material.
Supplemental material
Patient and public involvement
None.
Interventions
Two different care strategies were evaluated for their cost-effectiveness:
Usual (non-MINT Memory Clinic) care: patients are initially seen by their family physician for symptoms of CI and then referred to a geriatric specialist to determine a formal diagnosis and a treatment plan.
MINT Memory Clinic care: as described above, this care model provides team-based interprofessional collaborative dementia care, in a shared care approach with patients’ family physicians and with access to consultative specialist support for complex issues.6 7 10 If a family physician has access to a MINT Memory Clinic, any adult with memory concerns can be referred. MINT Memory Clinics exist in a variety of primary care settings across Ontario in rural, urban, remote and underserved communities. When there is no access to a MINT Memory Clinic, patients are likely to receive usual care.
Cohort
This study focused on older adults with memory concerns who were referred to receive usual care or MINT Memory Clinic care. Our cohort was based on data from a sample of 229 patients from the Centre for Family Medicine MINT Memory Clinic in Kitchener, Ontario. Patients were seen between January 2019 and January 2021. For inclusion, patients had to have had at least one clinic visit that documented standardised scale scores for cognition (Montreal Cognitive Assessment, MoCA)16 and quality of life (EQ5D-5L, a preference-based health status scale that is a valid and reliable measure of quality of life).17 The MoCA and EQ5D-5L are administered to patients as part of the Memory Clinic’s comprehensive assessment. We excluded patients who were unable or unwilling to provide consent or lack of capacity (as judged by patient’s physician). Patient characteristics are presented in table 1. The mean age of the cohort was 80 years; 52% were women. A total of 376 MoCA scores were collected from the sample of 229 patients. To account for the varying level of care required for patients during their disease progression, patients were classified into four CI states based on their MoCA scores: little to no CI (scores of 20–30); mild CI (scores of 16–19); moderate CI (scores of 11–15); and moderate-severe CI (scores of 2–10). The majority of patients (61%) had MoCA scores classified as little to no CI state (in this group, the average MoCA score was 24/30). It is important to note that while all patients referred to Memory Clinics have some cognitive symptoms or concerns, some will have subjective cognitive decline (SCD), which involves normal cognitive testing scores.18 Like mild cognitive impairment, SCD is an at-risk state for future Alzheimer’s disease and other dementias19; current Canadian Consensus guidelines recommend appropriate investigations and monitoring of persons with SCD because of the risk of progression to dementia.20 With cognitive test scores being within normal limits, persons with SCD were included in the little or no CI state. The identical cohort as described above was used for both the usual care intervention and the MINT Memory Clinic intervention in the cost-utility analysis.
Multi-specialty INterprofessional Team Memory Clinic patient characteristics
Model
A Markov-based state transition model was created to represent the progression of CI to dementia throughout a patient’s care journey (figure 1); a detailed model is presented in online supplemental figure 1. In our simulations, cohort members move between predefined health states in yearly cycles until all members die. In each yearly cycle, there are transition possibilities associated with a patient progressing to the next disease stage or remaining in their current health (CI) state. At each stage, changes in use of healthcare resources (emergency department (ED), hospital) were tracked. In our model, six main health states were: little to no CI; mild CI; moderate CI; moderate-severe CI; long-term care (LTC) admission; and death.
Markov-based state transition model for usual care and Multi-specialty INterprofessional Team Memory Clinics. *Little to no cognitive test findings; includes mild cognitive impairment and subjective cognitive decline. MoCA, Montreal Cognitive Assessment.
Data
Our model assumed that all patients started their journey within the little to no CI health state, and followed them over time until death. Transition probabilities related to disease progression, ED visits, hospitalisation and transition into LTC, were either derived from the MINT Memory Clinic data, an independent provincial evaluation of the Memory Clinics commissioned by the Ontario Ministry of Health,14 or other published literatures (table 2).21–24
Model parameters: transition probabilities, costs and utility
Disease progression probabilities
To calculate the annual disease transition probabilities, we used medical record data from the MINT Memory Clinic to build a disease history for each patient that began at their first assessment visit. The transition probability of patients moving between CI state groups within the next year was calculated using only data from patients who had at least two visits. Transition probabilities for disease progression are presented in table 2. Identical transition probabilities were used for both the usual care and Memory Clinic patients since we conservatively assumed that Memory Clinic care will not affect the progression of CI.
ED visit probabilities, hospitalisation probabilities and frequency of visits
The annual probability of a person in the little to no CI, mild CI and moderate CI states who have at least one ED visit is 26.2%.21 For the moderate-severe CI person, an annual probability of 45.5% was used.23 Among those who have had at least one ED visit, our model assumed that 22% of individuals visited the ED once, 24% visited two times and 54% visited three times based on published data.25 According to the provincial evaluation, 65% of MINT Memory Clinic patients returned to the community after a short-term hospital stay, compared with 61% of usual care patients.14
Transition into LTC homes
The probabilities of entering nursing homes were 1.2% for patients in the mild CI state and 3.5% for patients in the moderate CI state.22 For patients in the moderate-severe CI state, the transition probability was reported as 37.7%.24 Since patients in the little to no CI group were mostly younger and did not show many symptoms of CI, the model assumed no transition into LTC homes.
Mortality
All-cause mortality was calculated using life tables developed by Statistics Canada.26 Dementia-related mortality for both Memory Clinic and usual care patients in the hospital was 0.2% based on the provincial evaluation.14 Once patients were admitted to LTC, the annual mortality was assumed to be 30% based on the literature.27 28
Cost
Cost values in this model were derived primarily from the provincial memory clinic evaluation reported in 2017, in which a retrospective costing analysis based on health administrative data was conducted between patients receiving MINT Memory Clinic care and usual care from 2006 to 2015.14 Online supplemental table 1 presents a detailed summary of the daily costs of healthcare services for Memory Clinic and usual care patients. The cost of Memory Clinics was based on the conservative assumption that clinics operate 1 day a month and see four patients per day. The daily costs of healthcare services involved in both interventions were converted to yearly costs in order to determine the annual health state cost for both interventions. The total annual health state cost for Memory Clinics was calculated to be $C14 438 and $C21 020 for usual care. The one-time direct training cost involved in setting up the Memory Clinics was estimated at $C23 000 per clinic; this implementation cost is paid by the Ministry of Health. Using the same assumption as in the provincial evaluation,14 with each Memory Clinic operating once per month with a minimum four of patients per clinic day, the one-time training cost is estimated to be $C479 per patient ($23 000/12 months/4 patients) for the first year of operation.
For hospitalisation costs, inpatient hospital stays and mental health hospital stays costs reported in the provincial evaluation were combined, using an average length of hospitalisation stay of 10 days.29 The overall annual cost of hospitalisation was estimated at $C877 for usual care patients and $C416 for Memory Clinic patients. Similarly, annual nursing home costs were estimated at $C12 213 for usual care patients and $C9902 for MINT Memory Clinic patients. Table 2 provides an overview of all cost values used in our model.
Utility
Effectiveness was measured in quality-adjusted life years (QALY), calculated based on the quality of life of patients in given CI states. Utility scores were obtained from EQ-5D-5L surveys completed by 229 Memory Clinic patients, and a published study for purposes of comparative effectiveness for the usual care cohort.30 A detailed summary of the utility values used for both intervention groups is presented in table 2. The total effectiveness of care is presented as a sum of the QALY throughout the patient transition.
Analyses
A base case analysis was conducted first to estimate the incremental cost-effectiveness ratio (ICER) between the Memory Clinics and usual care based on a probabilistic analysis using Monte Carlo simulation for 5000 iterations. A full deterministic one-way sensitivity analysis was then performed on all model parameters over the plausible ranges using the reported 95% confidence interval, if available, or ±25% of the reference value, for parameters where estimates of uncertainty were not available. Further, two scenario analyses were conducted by (1) assuming the utility scores in each CI state remain the same for both the Memory Clinic patients and the usual care patients and (2) using the utility scores in each CI state from a published study (mild CI 0.9; moderate CI 0.68; severe CI 0.45).31 All analyses were conducted using TreeAge Pro 2021 (TreeAge Software, Williamstown, Massachusetts, USA).
Results
Base case analysis
The cost-effectiveness results between MINT Memory Clinics and usual care are presented in table 3 and online supplemental figure 2. The total average cost for a patient receiving MINT Memory Clinic care and usual care in MINT Memory Clinics is $C145 805 (95% CrI $C42 594 to $C244 574) and $C197 301 (95% CrI $C59 539 to $C331 406), throughout their entire care journey, respectively. The cost difference between Memory Clinic and usual care is $C51 496 (95% CrI $C4806 to $C119 367), indicating that MINT Memory Care is cost-saving in comparison to usual care. In addition, MINT Memory Clinics care is a more effective intervention in terms of total QALY (7.86 (95% CrI 2.34 to 12.86) QALY), in comparison with usual care (7.43 (95% CrI 2.31 to 7.56) QALY), which translates to a gain of 0.43 (95% CrI 0.01 to 1.24) QALY for MINT Memory Clinic care over usual care. In this probabilistic analysis (online supplemental figure 2), MINT Memory clinics were the superior option (less costly and more effective) in 97.7% of the 5000 Monte Carlo simulations.
Cost effectiveness of MINT Memory Clinics versus usual care: base case analysis and scenario analysis results
Scenario analysis and sensitivity analysis results
When we assumed the utility scores in each CI state remain the same for both the Memory Clinic patients and the usual care patients in the analysis, MINT Memory Clinic care remained to be a cost-saving option in comparison with usual care (table 3). Similarly, when we used the utility scores in each CI state from a published study31 in the analysis, the conclusion remained unchanged (online supplemental table 2). One-way sensitivity analysis (online supplemental figure 3) revealed that patients’ intervention starting age had the largest effect on the results. Patients with a lower starting age provided further cost-saving than the base case and showed improved quality of life compared with patients who entered usual care at the same age. Level of cost-saving was affected by the lower health service usage in MINT Memory Clinic care compared with usual care and the lower utility values for the usual care CI states, which created a greater difference in utility values between the groups and affected the level of cost-saving. Further, the cost of care for Memory Clinic patients in the little to no CI state group also affected the level of cost-saving. However, the conclusion remains favourable for MINT Memory Clinics when such uncertainty is considered.
Discussion
This study demonstrated that MINT Memory Clinic care is cost saving compared with the provision of usual dementia care in Ontario. Despite the minimal difference in utility values, MINT Memory Clinics greatly reduce overall healthcare costs as demonstrated in the lower costs for system resources such as LTC and ED visits.14 Variation in intervention starting age was found to have the greatest impact on ICER; patients may benefit from MINT Memory Clinic care more if they begin care at a younger age. When patients were identified with CI at a younger age and underwent usual dementia care services, they used more resources, which increased overall costs significantly. Even when considering the variation of all factors and a deviance in the normal values in our model, MINT Memory Clinic care was still shown to be cost saving. Moreover, as demonstrated in the probabilistic analysis, MINT Memory Clinics provided superior treatment over usual dementia care 98% of the time.
We have used a model-based approach to conduct the cost-effectiveness analysis for MINT Memory Clinic care; a similar approach has also been used to evaluate the cost of illness associated with dementia,32 33 and the cost-effectiveness of health interventions for people with dementia.31 Although no other studies have compared care models similar to MINT Memory Clinic care to usual dementia care services, cost-effectiveness of other dementia care interventions has been studied with positive results.30 31 34 35 A community health intervention that supported informal caregivers with systematic collection and sharing of patient health data with medical providers, was reported to be cost-effective under three of the four scenarios presented.31 The cost-effectiveness of a community-based, nurse-led collaborative dementia care management intervention that aimed to support persons with dementia and their caregivers through coordination of optimal care with their family physician was found to be a potentially cost-effective strategy for treating dementia due to improving quality of life (+0.05 QALY) at lower costs (−€569) compared with usual care services.30 Based on main cost-per-QALY analysis, care provided by an integrated multidisciplinary diagnostic facility was deemed cost-effective.34 Finally, an economic evaluation comparing the cost-effectiveness of 1-year dementia follow-up care by specialist-led memory clinics versus usual care provided by general practitioners showed that memory clinics were on average €1024 cheaper but had a decrease of 0.025 QALY compared with usual care,35 which may be attributable to the short follow-up time period. A 1-year follow-up period may not be sufficient to capture the effects of living with a progressive illness with significant sequalae that can negatively impact quality of life. A strength of our economic analysis is our larger sample size and longer EQ-5D-5L data collection time period.
The positive outcomes in this economic analysis are likely attributable to the unique features of the MINT Memory Clinic model, which differentiates it from other dementia care models and usual care. The MINT Memory Clinic model is effective because dementia care is provided at a primary care level, true coordination and collaboration between primary care, specialist and community care, and ongoing access to full dementia care service from one location that facilitates the comprehensive care needed to support healthy and safe living within the community as the disease progresses. Moreover, there is enhanced and ongoing nationally accredited training for the multi-disciplinary team members that was created and delivered by primary care-based clinicians, making it highly relevant to primary care practice, and involves best teaching practices.11 36 Timely diagnosis, person-centred care and early access to support and coordinated care for each patient and caregiver dyad compared with patients receiving usual care may reduce healthcare costs in the long term by decreasing frequency of ED visits and delaying institutionalisation. The fact that MINT Memory Clinic care demonstrated a slight increase in QALY in the face of a progressive neurodegenerative condition can be viewed as positive as it may reflect the significant impact that early support can have on helping persons with dementia live fulfilling and independent lives for as long as possible. Current evidence demonstrates the potential of interventions focused on earlier management of CI and/or dementia in yielding economic benefits.37
Similar to all studies that use convenience sampling, our results may have underestimated or overestimated the cost-effectiveness of MINT Memory Clinic care due to selection bias associated with our sampling method and a relatively small sample size.38 The lack of existing research regarding a comparative usual care group for persons with dementia living in Canada limited us to using available data from different countries and healthcare systems. As such, the comparability between MINT Memory Clinic care and usual care may be limited since all of the data used was not collected from within the Canadian healthcare system. Despite this limitation, key values such as transition probabilities and cost values were taken directly from the MINT Memory Clinic patient database and Canadian administrative databases (ICES). Further research is needed to collect utility values for persons living with dementia in Canada in the usual care setting. This data would play a key role in future economic analyses of dementia care programmes in Canada. Further, we are not able to investigate the impact of the type of dementia in relationship to our results due to existing data limitations. In addition, we conducted our analysis using a health system perspective rather than a societal perspective, thus we may have underestimated or overestimated the benefit of MINT Memory Clinics as costs associated with patient and caregiver time and out-of-pocket expenses were not included in our analysis.39 40
Another limitation was the exclusion of costs of space and administration costs in the calculation costs for MINT Memory Clinics. As MINT Memory Clinics are often operated within existing family practice sites, there is no additional cost for space in most cases. We conservatively estimated new MINT Memory Clinic capacity at four newly diagnosed patients with dementia per month among the patients with other cognitive diagnoses being made. As more mature clinics may have greater capacity, our results may underestimate cost-efficiency for some clinics. The estimated cost for salaries used in our study is a gross overestimation as most health professionals are already employed within the primary care site and their work in the clinic is infrequent, in some cases just 1 day per month, given the efficiencies of a shared care model with the patients’ own family physicians. Finally, as our data are most relevant to Canada, and in a particular to community care settings, it may be difficult to generalise to other jurisdictions due to differences in healthcare systems.
Conclusion
As there is a growing need for high-quality, cost-effective, dementia care within the context of limited healthcare resources, information about the economic impact of the MINT Memory Clinic care can inform health service design and resource allocation. Our study adds to the growing body of literature demonstrating that dementia care interventions in primary care can have significant positive impacts on healthcare system resource use.41 Our study showed that as compared with usual care, patients receiving MINT Memory Clinic care had much lower healthcare costs and modestly improved quality of life. Based on the results of this study, the MINT Memory Clinic model has a very high likelihood (98%) of reducing healthcare costs and improving healthcare over usual care. Implementation of this care model across primary care systems may assist with improving quality and access to memory care while decreasing the growing economic and social burden of dementia.
Data availability statement
Data are available upon reasonable request. Data are available upon reasonable request. The data that support the findings of this study are not publicly available due to them containing information that could compromise participant privacy. Deidentified, limited data will be shared by the corresponding author upon request.
Ethics statements
Patient consent for publication
References
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Footnotes
Contributors LL and WWW were involved in study conceptualisation and implementation. LL, WWW, SW, CL and TP were involved in study design. WW, SW and CL completed the data collection and analysis. WWW is the guarantor. All authors (WWW, LL, SW, CL, LH, APC, SS) were involved in data interpretation and manuscript preparation and final approval.
Funding This work was supported by the Centre for Family Medicine Family Health Team.
Competing interests SW and CL are employed by the Centre for Family Medicine Family Health Team.
Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.