Abstract
Objectives 1) To compare the average cost of an emergency department (ED) visit for various minor musculoskeletal disorders between two models of care (physiotherapist and ED physician or ED physician alone); 2) To evaluate the incremental cost-effectiveness ratio (ICER) of these two models of care over a 3-month period post-initial visit; and 3) To estimate the ICER of three ED models of care (physiotherapist and ED physician, ED physician alone, physiotherapist alone) over a two-year period.
Methods A randomized clinical trial was conducted among individuals (n=78) aged 18 to 80 years presenting with a musculoskeletal disorder at a Quebec City (Canada) hospital in 2018-2019. Two models of care were compared: management by a physiotherapist and an ED physician versus usual management by an ED physician. Participants follow-ups were conducted at 1 and 3 months post-initial ED visit. Obj.1: The health care costs incurred by the two groups during their ED visit will be calculated using the Time-Driven Activity-Based Costing (TDABC) method and compared using generalized linear models. Obj. 2: The cost-utility analysis over a 3-month time horizon will combine economic and clinical variables (estimated through quality-adjusted life years) using a Societal perspective. The results of the cost-utility analysis will be reported using an ICER. Obj. 3: The ICER will be estimated using a hybrid decision tree (0-3 months post-visit) and Markov model (3-24 months post-visit); the analysis will be conducted from a Societal perspective over a two-year time horizon.
Conclusion This study will help to determine which model of care is most efficient for the management of individuals who come to the ED with minor musculoskeletal disorders. The increased involvement of various health professionals in the management of patients in the ED paves the way for the development of new avenues of practice and more efficient organization of services.
Introduction
The emergency department (ED) serves as the main gateway and the preferred resource when primary care services are not available, for example in cases of lack of affiliation with a primary care source or inability to see a physician within a reasonable time frame.[1–5] Although pain conditions for which patients decide to go to the ED are varied, they are oftentimes related to a musculoskeletal disorder (MSKD).[6–8]
According to the World Health Organization, MSKDs are characterized by “pain (often persistent) and limitations in mobility, dexterity and general functioning”.[9] MSKDs can affect joints, bones, muscles, spine and multiple regions of the body.[9,10] The prevalence of these disorders is reported to be significantly higher in women, older people and people with low socio-economic status.[11–15] When they do not receive timely and appropriate care, people with MSKDs tend to make greater use of health care services and resources.[16–22] MSKDs account for up to 12.6% of a country’s total health care costs each year [15] and this figure is expected to rise with the increase in obesity, physical inactivity and the aging of the population.[11,23] It is therefore essential to study the costs and clinical effectiveness of interventions aimed at managing MSKDs in order to choose the most efficient ones, including in the ED.
Various models of care have been implemented in the ED and studied in recent years to optimize the management of people presenting with MSKDs. These models of care aim to optimize the flow of patients to and in the ED in three distinct phases: “input” (i.e., flow of patients deciding to come to the ED), “throughput” (i.e., flow of patients while in the ED), and “output” (i.e., flow of patients upon discharge from the ED).[24] Such models of care include for instance fast-track corridors for patients with minor injuries or rapid assessment teams.[25] Some models include the addition of ED nurse practitioners and a variety of health professionals with a usual or extended scope of practice, such as the primary contact physiotherapist or advanced practice physiotherapist.[25]
The addition of primary contact physiotherapists in the ED is an emerging model of care that aims to optimize patient flow while in the ED.[25] Several studies conducted in recent years have shown that this model of care is associated with reduced time waited before receiving care, and reduced length of stay in the ED, as well as fewer unnecessary consultations with various health professionals, and less prescriptions of imaging tests and medication, including opioids, and over-the-counter medication.[8,26–29] In addition, this model of care was associated with fewer repeat visits to the ED for a similar condition for up to one month after the initial ED visit.[29] Thus, management by a primary contact physiotherapist appears to be associated with decreased service and resource use, both at the ED and up to several weeks later. However, very few studies having investigated primary contact physiotherapist care in the ED have looked at its cost-effectiveness.
Indeed, despite evidence of clinical benefits associated with the presence of a primary contact physiotherapist in the ED (effectiveness), scientific evidence remains rather scarce regarding the cost-effectiveness of this model of care. Two studies conducted in primary care settings (primary care clinic and private clinic) report that primary contact physiotherapist management is associated with a slight increase in health-related quality of life and a decrease in total costs compared to usual management by a family physician.[30,31] In addition, early physiotherapy management was associated with a decrease in total MSKD-related costs for up to two years after initial management.[17,19,32] Two cost-minimization studies conducted in Great Britain looked specifically at the costs associated with the integration of a primary contact physiotherapist in the ED compared to usual management by an emergency physician.
According to the study by Richardson et al. (2005, n=766 patients with non-fracture MSKDs), the presence of a primary contact physiotherapist in the ED results in costs equivalent to usual management (emergency physician).[33] Similarly, according to McClellan et al. (2013, n=372 patients >16 years of age with a peripheral MSKD), management by a primary contact physiotherapist results in costs at least as high as usual management (emergency physician).[34] Nevertheless, in addition to having been conducted exclusively in Great Britain several years ago, these two studies only measured the costs of the two models of care compared and not their effectiveness, the authors assuming that the two models compared were equivalent in terms of clinical effectiveness. These studies are thus not considered to be formal economic evaluation according to current guidelines, but rather a costing exercise, in that a cost-effectiveness analysis accounts for the uncertainty associated with the effects of the interventions being compared.[35] To our knowledge, no other study has examined the cost-effectiveness of primary contact physiotherapy in the ED. Furthermore, no study has assessed whether involving primary contact physiotherapists in the ED have a long-term impact on use of health system services and resources for persons with minor MSKDs. Consequently, further evidence is needed on the efficiency of integrating a primary contact physiotherapist in the ED compared to usual management by an emergency physician.
Therefore, the general objective of this project is to evaluate the efficiency of different models of care for the management of minor MSKDs in the ED. More specifically, the objectives are to:
Compare the average costs of an ED consultation and care for various MSKDs, according to two models of care:
Usual management by an emergency physician
Primary contact physiotherapist management + emergency physician management
Evaluate the incremental cost-effectiveness ratio (ICER), from a Societal perspective, of these two ED models of care for the management of MSKDs over a three-month period post-initial ED visit.
Estimate the ICER between three ED models of care for MSKD management over a two-year period from a Societal perspective:
Usual management by an emergency physician
Primary contact physiotherapist management + emergency physician management
Primary contact physiotherapist management alone
Materials and Methods
Study design and costing approaches
Objectives 1 and 2 will be achieved using data collected through a two-arm pilot pragmatic randomized clinical trial (RCT) conducted in the ED of the CHUL, one of the five sites of the CHU de Québec - Université Laval (UL) (Quebec City, Canada) from September 2018 to March 2019. This trial aimed to compare the effects of management by a primary contact physiotherapist to usual care provided by an emergency physician for persons presenting with a minor MSKD on their clinical course (pain and pain interference) and the use of resources at ED discharge and after 1 and 3 months post-visit.[29] Two groups of participants were compared: one group managed by a primary contact physiotherapist and an emergency physician and one group managed by an emergency physician alone. The costing approach used for Objective 1 will be Time-Driven Activity-Based Costing (TDABC), which involves determining the per-minute costs associated with each care process included in a care pathway by multiplying the cost per minute of each care process by its duration. Details on this costing approach and its application to the ED have been described by one of the authors elsewhere.[36] Objective 2 will be achieved through a cost-utility analysis approach in which health care costs at the ED visit and those reported at the 1- and 3-month follow-ups will be compiled and combined with the utility scores obtained at the same measurement times, from a Societal perspective. Cost-utility analysis is favored in Canada since it uses a generic outcome measure allowing comparison of the health gains associated with several different interventions, such as different models of care.[35]
The ICER between the three ED models of care for the management of MSKDs over a two-year period (Objective 3) will be estimated using a cost-effectiveness analysis via a hybrid mathematical model. This model will consist of a decision tree covering the period from the initial ED visit up to three months post-initial visit, and a Markov model starting three months post-initial ED visit and ending 24 months (two years) after the ED visit. The decision tree provides a simple and clear illustration of a patient’s possible short-term care pathways following a new intervention.[35,37] In addition to reporting the different interventions used, the decision tree also allows for the inclusion of adverse events following the initial intervention, such as a new ED visit for the same condition, and for repeating an intervention over time as needed (e.g., new visit in the ED a few days after the initial visit and then a new visit two months later for the same condition).[35,37] It also permits to determine the proportion of disability associated with each of the three ED models of care.
Several considerations guided the choice of the time horizon for the Markov model. First of all, to be considered chronic, a musculoskeletal disorder must be present for at least three months.[38] Moreover, approximately 30% of people presenting with MSKDs report pain and functional disability lasting more than 12 months after the onset of their condition. Furthermore, studies on MSKD care in primary care or the ED have had follow-up periods ranging from six to 24 months (e.g. [26,39–42]). Thus, the Markov model will cover a 24-month period. It will include two-week cycles in order to capture the clinical evolution of the patients included.
Study population
Inclusion and exclusion criteria of the previous pragmatic RCT are described in Box 1. A more detailed version of the study population can be found elsewhere.[29]
Inclusion and exclusion criteria for the pilot pragmatic randomized clinical trial in the ED
Data collection
Data necessary to achieve Objectives 1 and 2 were collected at the initial ED visit and at the one- and three-month post-visit follow-ups. While more details of the data collection procedures can be found in our previous paper [29], any person presenting to the ED who met the inclusion and exclusion criteria was seen by a member of the research team who confirmed eligibility, obtained informed consent, and ensured completion of baseline questionnaires. The participant was then randomized to either study group: primary contact physiotherapist + emergency physician management or usual management by the emergency physician alone. After the ED visit was completed, participants were contacted at 1 and 3 months either by phone or email to complete post-visit follow-ups.
The study population for the model of care consisting of primary contact physiotherapist management and discharge from the ED (Objective 3) was not observed during the pilot pragmatic RCT. Therefore, the parameters needed to represent it (probabilities, costs, measures of effectiveness) will be taken from a literature review, an approach regularly used in economic evaluation.[43] However, the studies from which the metrics will be derived will need to have a sample that meets the same inclusion criteria as those presented in Box 1. Data extracted from the literature will be validated with members of the RCT’s research team and with experts in the field of emergency medicine, MSKDs and rehabilitation if necessary during the construction of the hybrid model.[35]
Study outcomes
Primary outcomes used to measure the average cost of an ED visit (Objective 1) will be the costs of care processes and the time associated with each care process. This method of costing is routinely used by some members of the research team.[36,44,45] The costs related to ED management (medical and non-medical staff, imaging, medication, consumables, maintenance, etc.) were obtained via a formal request made by a member of the research team to the CHU de Québec – UL Finance Department. The time associated with each care process was calculated by a member of the research team using estimates provided by the CHUL medical and non-medical staff that were validated during an observation period in the ED.[36,44]
As part of the cost-utility analysis (Objective 2) and hybrid mathematical model (Objective 3), the efficiency of the ED models of care will be assessed using an incremental cost-effectiveness ratio (ICER). The resulting ICER will be reported in terms of incremental cost per quality-adjusted life years (QALY) gained, between the models of care. The total cost of the ED visit for each individual obtained within Objective 1 will be added to the individual follow-up costs recorded via the self-administered follow-up questionnaires completed by participants at 1 and 3 months during the pilot pragmatic RCT and averaged to obtain an average 3-month individual cost for each model of care (Objective 2 & 3 – decision tree). The questionnaires provided data on resources used by each participant during follow-up such as ED re-visits for the same condition, number of consultations with other health professionals, etc. Costs associated with each of the resources used will be drawn from data from the Régie de l’assurance-maladie du Québec (RAMQ) (fees of the emergency physician and other physicians consulted, drugs, imaging tests) and from a search of the grey literature (costs of the fees of the primary contact physiotherapist and other health professionals consulted, and of technical aids).[46] Mean utility scores were obtained at the initial visit [47] and at 1 and 3 months using the EQ-5D-5L, a generic standardized questionnaire designed to measure health status in an economic and clinical evaluation.[48] The EQ-5D-5L has been found to be reliable, valid, and sensitive to change.[49,50] The efficiency values and the costs from 3 to 24 months required to run the Markov model (Objective 3) for the three models of care will be taken from the literature.
Data analysis and interpretation of results
As part of Objective 1, a mapping of the care pathways encountered will be completed for each type of MSKD encountered in our study population (i.e., low back pain, neck pain, upper limb, lower limb) (Fig 1). The unit cost of each of the resources, consumables and indirect costs required in each process of care of the care pathway will be calculated and multiplied by the duration of each process to obtain the cost related to each process of care present in the care pathway. The costs associated with each process will be summed to obtain the total cost of the ED care pathway specific to each MSKD and each model of care (i.e., emergency physician or primary contact physiotherapist management). A generalized linear model with a Gamma distribution and log link will be used to test whether there is a significant difference in the costs of managing equivalent MSKDs between the two models of care.[51,52]
The decision tree (Objective 3) will include all the interventions and services possibly used by a participant following the initial visit to the ED for each model of care considered (Fig 2). The conditional probability of ending up in each of the terminal nodes of the decision tree will be used to calculate the proportion of disability associated with each model (Fig 2).The disability proportions obtained for each model of care will be used to determine the number of individuals in each state at entry in the Markov model (Fig. 3). The Markov model will then be used to calculate the long-term costs and effectiveness over 2 years of each of the model of care based on the level of disability estimated in the decision tree.[35,37]
Both the cost-utility analysis (Objective 2) and the hybrid mathematical model (Objective 3) will be conducted from a Societal perspective. Results obtained via the EQ-5D-5L at 1 and 3 months (Objective 2 & 3 – decision tree) will be converted to utility scores using the Canadian conversion algorithm developed by Xie et al.[53] As the 3-month retention rate for the pilot pragmatic trial was 80% [29], some participants’ data are missing (service and resource use, costs, utility scores). Missing data will be imputed using the Missing not at random (MNAR) multiple imputation method.[54] Uncertainty in cost and effectiveness measures for the costutility analysis (Objective 2) will be obtained using non-parametric bootstrap resampling with replacement. Uncertainty in the hybrid model parameters (probabilities, costs, and efficiency) (Objective 3) will be obtained via a probabilistic sensitivity analysis performed using a Monte Carlo simulation. Both uncertainties will be represented visually using a cost-effectiveness diagram, cost-effectiveness acceptability curve, and cost-effectiveness acceptability frontier.[35] Subgroup analyses of men and women and by MSKD category (spine, upper extremity, and lower extremity) will also be performed for both objectives (p<.05).
Ethical considerations and data management
Ethical approval for the collection of the necessary data was obtained from the Research Ethics Committee of the CHU de Québec - UL (approval number: MP-20-2019-4307). The randomized clinical trial was also registered with the US National Institutes of Health (#NCT04009369). Each participant signed a written consent form prior to participation. All data collected will be kept in a secure repository and destroyed thereafter. All members of the research team signed a confidentiality agreement.
Discussion
The overall aim of this project is to evaluate the costs of different models of care for the management of MSKDs in the ED. This will be achieved through three specific objectives: 1) to compare the average costs of an ED consultation and care for various MSKDs; 2) to evaluate the incremental cost-effectiveness ratio (ICER) of two ED models of care for the management of MSKDs over a three-month period post-initial ED visit; and 3) to estimate the ICER between ED models of care for the management of MSKDs over a two-year period.
Until now, there has been no formal economic evaluation of the inclusion of a primary contact physiotherapist in the ED compared with usual practice (emergency physician). The only studies that have been done on the subject have assumed that the effectiveness of the primary contact physiotherapist’s management in the ED is equivalent to that of usual care by the emergency physician. However, several studies have reported that primary contact physiotherapist management can reduce the use of services and resources during the ED stay.[8,26–29,55] This research project will fill an important need in the literature by providing an in-depth analysis of the costs and efficiency of the considered models of care. Indeed, this project will help identify the most efficient ED model of care. These models of care also have the potential to improve the quality of services offered to people with MSKDs, their clinical evolution and their quality of life. The increased use of various health professionals in the management of patients in contexts such as the ED can pave the way for the development of new avenues of practice and potentially more efficient organization of services that will benefit the population.
This study is associated with some potential limitations. First of all, the data needed to carry out Objectives 1 and 2 were mostly obtained from a pilot pragmatic randomized clinical trial. Therefore, the results obtained should be interpreted with caution. The small sample size (n=78) could possibly limit analyses on the number of plausible branches in the final decision tree as well as the amount of subgroup analysis that will be performed. In addition, although high, the retention rate at the 3-month follow-up of the randomized clinical trial was 80% [29], which implies that some data related to the use of services and resources, costs and health-related quality of life will be missing. However, this limitation will be mitigated using multiple imputation methods.[54] Sensitivity analyses will also be performed to assess the robustness of the results obtained. Finally, it may be difficult to obtain some of the data on medium- and long-term costs and measures of effectiveness for the ED models of care studied in Objective 3 from the scientific literature. Nevertheless, estimates can be obtained by soliciting the opinions of experts in the fields of MSKD management and emergency medicine, as this method is regularly used in modeling.[35]
As for knowledge translation, following the project, formal presentations will be made to all key stakeholders at the CHU (emergency physicians, physiotherapists, nurses, orderlies, patient representatives and administrators) on site or remotely to present the results of the study and discuss lessons learned and future avenues. The results of this project will also be shared with provincial stakeholders (professional associations, patient associations and governments). They will also be disseminated at national and international scientific conferences on economics, health services organization and emergency services. Four manuscripts will be published in peer-reviewed journals. If successful, this project will help guide economic evaluations for a large-scale, multi-center trial aiming to improve the management of people presenting with a MKSD in the ED.
Data Availability
All data produced in the present study are available upon reasonable request to the authors.
Acknowledgements
The authors would like to thank the following persons for their contributions: project participants, Antony Barabé, PT, physiotherapist at the Centre Hospitalier de l’Université Laval (CHUL), the entire team of managers at the Direction des services multidisciplinaires of the CHU de Québec – Université Laval (Marie-Christine Laroche, Catherine Van Neste, Marie-Claude Brodeur and Stéphane Tremblay) for their support throughout the implementation of the project and its realization.
References
