Trends and Opportunities in Health Economic Evaluations of Prosthetic Care Innovations

Overcoming obstacles to prosthetic fittings requires frequent tryouts of sockets and components. Repetitions of interventions are upsetting for users and place substantial economic burden on healthcare systems. Encouraging prosthetic care innovations capable of alleviating clinical and financial shortcomings of socket-based solutions is essential. Nonetheless, evidence of socio-economic benefits of an innovation are required to facilitate access to markets. Unfortunately, complex decisions must be made when allocating resources toward the most relevant health economic evaluation (HEE) at a given stage of development of an innovation. This paper first, aimed to show the importance and challenges of HEEs of intervention facilitating prosthetic fittings. Next, the main trends in HEEs at various phases of product development and clinical acceptance of prosthetic care innovations were outlined. Then, opportunities for a basic framework of a preliminary cost-utility analysis (CUA) during the mid-stage of development of prosthetic care innovations were highlighted. To do this, fundamental and applied health economic literature and prosthetic-specific publications were reviewed to extract and analyse the trends in HEEs of new medical and prosthetic technologies, respectively. The findings show there is consensus around the weaknesses of full CUAs (e.g., lack of timeliness, resource-intensive) and strengths of preliminary CUAs (e.g., identify evidence gaps, educate design of full CUA, fast-track approval). However, several obstacles must be overcome before preliminary CUA of prosthetic care innovations will be routinely carried out. Disparities of methods and constructs of usual preliminary CUA are barriers that could be alleviated by a more standardized framework. The paper concludes by identifying that there are opportunities for the development of a basic framework of preliminary CUA of prosthetic care innovations. Ultimately, the collaborative design of a framework could simplify selection of the methods, standardise outcomes, ease comparisons between innovations and streamline pathways for adoption. This might facilitate access to economical solutions that could improve the life of individuals suffering from limb loss.


INTRODUCTION
Alfred Nobel (1833-1896) said the following about innovation "If I have a thousand ideas and only one turns out to be good, I am satisfied." In healthcare, the difference between a "good" or a not so good innovation is made during health technology assessment (HTA) and/or health economic evaluation (HEE). 1 As defined in APPENDIX 1, these evaluations aim at understanding what is the value for money of a treatment. Simply put, payers want to make sure they get a bang for their buck! This is tough question because the answer is rarely black and white. Nonetheless, addressing any concerns with socio-economical value of an intervention is a prerequisite to warrant access to market. Great but unaffordable treatments have little prospect of being adopted by healthcare policymakers.
The paper deals with issues of health economic assessments specific to prosthetic care innovations as presented in Figure 1. First, the importance and challenges of HEEs of interventions facilitating prosthetic fittings are highlighted. Next, the main trends in HEEs of new healthcare technologies are outlined with particular emphasis on specific HEEs to consider during the course of development of innovations. Then, opportunities for a basic framework of preliminary assessments during the mid-stage of development of prosthetic care innovations are suggested. Finally, the paper concludes with some calls to action to further develop preliminary assessments.

IMPORTANCE OF HEALTH TECHNOLOGY EVALUATIONS
This introductory section highlighted (A) the needs for solutions facilitating prosthetic fittings and (B) the current challenges to produce relevant health economic evaluations of prosthetic care innovations.

Role of prosthetic care
Because the everyday ability of individuals suffering from limb loss to use an artificial limb is critical to their quality of life, clinical teams made bespoke recommendations intending to maximize comfort, stability and mobility of prosthetic fittings. 2, 3 Ultimately, this process incorporates all personalized interventions performed by a prosthetist around the choice and alignment of prosthetic components as well as the management of prosthetic attachment to the residuum including design, manufacture and adjustment of socket or osseointegrated implant. 4 Outcomes of prosthetic fitting depends largely on the performance of prosthetic components. 5-11 where Dillingham et al (2001) noted that 60% of amputees are satisfied with prosthetic characteristics such as weight, aesthetics and functionality (e.g. servicing, how easy the prosthesis is to use) and 57% of the traumatic lower limb amputees in the study expressed some dissatisfaction with prosthetic comfort. 12 Since, studies showed that the use and satisfaction of prosthetic lower limb could be significantly improved when using advanced components such microprocessor-controlled knees compared to a nonmicroprocessor-controlled knees. 9,13,14

Health economic evaluations of innovations
Ijzerman and Steuten (2011) identified that in order for societal benefits to be maximized three things must occur: 1) governments need more data on benefits arising when public resources are spent, 2) companies need more data to effectively manage their product development portfolios and 3) research programs at universities may need to be actively encouraged in this direction. 37 Policymakers in healthcare organizations around the world adopt a reasoning more or less utilitarian when making decisions about medical care expenses. 37 However, healthcare administrators are often obligated to confirm the value for money of interventions prior approval (e.g. fee-forservice, fee-for-value The burden of HEE of an innovation also falls onto developers and manufacturers of technological solutions including attachments (e.g., liners, sockets, implants), artificial limb components (e.g., elbow, wrist, knee, ankle) and protective device (e.g., shock absorbers, failsafe). 38 Steven et al (2019) suggested that solution developers must understand the value created by their interventions and act quickly on them to provide some forms of evidence of costeffectiveness of their innovations. 48 Failing to do so could seriously hinder access to market and adoption of their innovations. O'Malley (2010) indicated that the most common reason for the Australian Medical Services Advisory Committee to not recommend funding for new technology was not only insufficient clinical evidence but also the lack of proven cost-effectiveness presented during early stage of the examination process. 54

Making decisions about economic evaluations
Steven et al (2019) stated that HEE can be approached in a number of ways. They identified a range of approaches to compare the costs of health care services and possible cost savings which observe the consequences of an intervention and the effectiveness of that same intervention through a lens of outcomes that are valued patients, payers and providers, or which align with widely used global utility measures. 48 They specified that the value of a prosthetic care intervention could be assessed using a range of costbenefit, cost-consequence, cost-effectiveness and costutility analyses considering valuations of costs (e.g., monetary units) and a range of benefits. Ijzerman and Steuten (2011) specified that no single method will produce the right information for all decision makers. Each method has advantages and disadvantages and work for specific applications, as opposed to all applications. 37 They suggested that a toolbox of methods must be used. Issue

CURRENT TRENDS IN HEALTH TECHNOLOGY EVALUATIONS
This second section (A) reviewed generic pathways to assess health economic consequences of a new treatment at a given stage of product development and clinical acceptance and (B) highlighted selected studies that followed these pathways to assess prosthetic care interventions.

Key concepts of health economic evaluations
As described in APPENDIX 1, HEE include, but not limited to, cost-effectiveness analyses (CEA) or cost-utility analyses (CUA). These terms are often used interchangeably although they are technically looking at different types of utilities. CEAs are concerned with a particular functional outcome of a treatment (e.g., walking speed). CUAs rely on self-reported quality of life status measured using standard surveys such as EQ-5D or 36-Item Short Form Survey (SF36). CUAs comparing usual and new treatments involve incremental cost-utility ratio (ICUR) based on incremental costs and utilities over time that could be compared to cost-effectiveness (CET) or, more often, willingness-to-pay (WTP) thresholds. Two studies were of particular interest because they can assist promoters to make an educated decision when choosing an HEE accordingly to the level of innovation development. Ijzerman and Steuten (2011) systematically described that early, preliminary and full CUAs can be conducted at the early, mid and late stage of clinical acceptance of any medical treatment, respectively. 37 More recently, new insights were provided by Kannenberg and Seidinger (2019) who explained how these three types of CUAs should also be performed by prosthetic manufacturers at early, mid and late phase development of a prosthetic product. 38 The authors indicated that CUA during the product's life cycle is beneficial in three ways. It allows potential cost-effectiveness to be estimated and included in investment decision processes and mitigates the risk of investing in technology unlikely to be costeffective. It helps to prioritize between competing costeffective concepts or technologies. It facilitates the identification of parameters having the largest impact on the likely cost-effectiveness of the product to be identified in order to best manage limited research funds. 38 Figure 2 gives an overview synthesizing both approaches. Decision uncertainty and strength of evidence were suggested for early, preliminary and full CUAs during early, mid and late phase of product development (manufacturer's perspective) and clinical acceptance (healthcare's perspective) of prosthetic care innovations, respectively.
Next, the general principle, expected capacity to address Consolidated Health Economic Evaluation Reporting Standards (CHEERS) and Consensus Health Economic Criteria (CHEC) extended checklists, typical strengths and weaknesses as well as selected examples of prostheticfocused CUAs. 75-77 is briefly described. The decision was made to present the CUAs as they historically gained recognition starting from full, to preliminary and early CUAs rather than following the sequential timeline of their implementation. Appraisal of each type of CUAs using the CHEERS and CHEC-extended checklists were detailed in Supplementary material.

Full cost-utility analyses
Traditionally, mainstream HEEs involved comprehensive or "full" CUAs essentially produced when innovations are gaining clinical acceptance after commercialisation. Full CUAs can be conducted from societal and/or healthcare perspectives. These CUAs usually rely on primary costs extracted from financial records expressed in monetary units as well as utilities measured by quality of life surveys expressed in QALY for cohorts of participants over an extended period of time (APPENDIX 1). 48 It was postulated that conventional full CUAs should address strongly all items of the CHEERS and CHECextended checklists ( Table 1,Table 2).
Modelling CUAs can be comprehensive because of the breadth (e.g., scenarios) and depth (e.g., time horizon) of their analysis. Furthermore, uncertainty and sensibility of outcomes, shown by the size of the errors around the point estimates due to data sources (e.g., sample size) and/or to the process of evaluation (APPENDIX 1), tend to be well worked out and, possibly, relatively low compared to early and preliminary CUAs. 82 Therefore, full CUA provide strong evidence supporting robust recommendations considered by decision makers (e.g., approval for funding).
However, modelling CUAs require substantial resources. Building models is labour intensive (e.g., determine scenarios, test assumptions). More importantly, Kannenberg and Seidinger (2019) noted the necessity of requiring the inclusion of outcome parameters, like healthrelated quality of life, in these models. 38 This means that full CUAs produce their best outcomes when sufficient costs and utilities are known for large cohorts over an extended length of time in a given jurisdiction (e.g., within-trial and beyond-trial horizon studies). 83 Evidence-based Several studies used a full CUA to assess consequences of the provision of socket based solution including advanced prosthetic components such as microprocessor-controlled knees and energy storing and return feet as well as socketfree solutions including bone-anchored prostheses. 21,61,63-

Preliminary cost-utility analyses
The issue of timeliness of full CUAs could be addressed by performing preliminary CUAs of innovations that could take place sometimes around the mid-stage of product development when clinical usage is still limited to small cohorts. Preliminary CUA is an option "in-between" early and full CUAs that considered innovations with a broad range of development status. Therefore, preliminary CUAs can be conducted using a wide spectrum of methods. They can involve primary data of actual (e.g., financial records) and/or simulated (e.g., purposely created schedules) costs expressed in monetary units as well as measured (e.g., quality of life surveys) and/or guesstimated (e.g., literature) utilities expressed in QALY for cohorts of participants over a somewhat lengthy time horizon. 48,50-52,62,78,79 The assumption was made that typical preliminary CUAs have a weak and moderate capacity to address 9 (33%) and 8 (30%) of items in the CHEERS checklist, including 7 (44%) and 6 (38%) of items in the Methods as well as 2 (40%) and 2 (40%) of items in the Results sections, respectively ( Table 1). It was estimated that preliminary CUAs should be capable to address 11 (58%) of items in the CHEC-extended checklists ( Table 2).
Resources needed to conduct preliminary CUAs could varied depending on the sources of data considered. Estimating costs from schedules and utilities from literature might require less resources than extracting costs from financial systems and utilities from a survey for a cohort of convenient sample size. Preliminary CUAs can provide some indications of probable consequences of innovations. Practically, preliminary CUAs can generate primary information, in part or in whole, useful for modelling CUAs (e.g., costs and utilities estimates, scenario drafting).
However, preliminary CUAs are usually built around substantial assumptions based on best-estimates of costs and utilities at the time.

Early cost-utility analyses
Preliminary CUAs can provide timelier assessment than full CUAs. Nonetheless, there is a current trend in health economic literature arguing that preliminary CUAs are yet to provide sufficiently timely assessment of innovations. Hence, the promotion of early CUAs, also called "iterative economic evaluations" or "very early HTA" by Ijzerman and Steuten (2011), which pointed out that attempts have already been made, using "horizon scanning systems", to include new, emerging technologies into health policy as it is developed. Other authors have referred to this as the use of "early warning systems". 37 Early CUAs tend to be constructed like preliminary CUAs but they rely more heavily on sparser costs and utilities data as well as sketchier assumptions. These analyses tend to be based on best guestimates of most likely costs and utilities collected with case-series studies and/or extracted from the literature often produced outside the relevant jurisdiction.

Background and objectives 3
Provide an explicit statement of the broader context for the study. Present the study question and its relevance for health policy or practice decisions.  For each intervention, report mean values for the main categories of estimated costs and outcomes of interest, as well as mean differences between the comparator groups. If applicable, report incremental cost-effectiveness ratios.

Strong Strong Strong
Characterising uncertainty 20a Single study-based economic evaluation: Describe the effects of sampling uncertainty for the estimated incremental cost and incremental effectiveness parameters, together with the impact of methodological assumptions (such as discount rate, study perspective).

Weak Moderate Strong 20b
Model-based economic evaluation: Describe the effects on the results of uncertainty for all input parameters, and uncertainty related to the structure of the model and assumptions.

Weak Weak Strong
Characterising heterogeneity 21 If applicable, report differences in costs, outcomes, or cost effectiveness that can be explained by variations between subgroups of patients with different baseline characteristics or other observed variability in effects that are not reducible by more information.

Discussion
Study findings, limitations, generalisability, and current knowledge 22 Summarise key study findings and describe how they support the conclusions reached. Discuss limitations and the generalisability of the findings and how the findings fit with current knowledge.

Strong Strong Strong
Other

Source of funding 23
Describe how the study was funded and the role of the funder in the identification, design, conduct, and reporting of the analysis. Describe other non-monetary sources of support.

Conflicts of interest 24
Describe any potential for conflict of interest of study contributors in accordance with journal policy. In the absence of a journal policy, we recommend authors comply with International Committee of Medical Journal Editors recommendations. It was assumed that usual early CUAs have a weak capacity to address 15 (56%) of items in the CHEERS checklist including 11 (69%) of items in the Methods as well as 4 (80%) of items in the Results sections ( Table 1). Preliminary CUAs might be incapable to address up to 9 (47%) of items in the CHEC-extended checklists ( Table 2).

Strong Strong Strong
Early CUAs are affordable and timely. They could help to reduce or validate assumptions subsequently used in preliminary or modelling CUAs. Perhaps, the most valuable return on investment of early CUAs is to provide insight into the viability of the product and worthiness of the clinical introduction on an innovation, as described by Kannenberg and Seidinger (2019). 38 As expected, outcomes of early CUAs are likely to have high uncertainty and sensibility leading to low level of evidence and only tentative recommendations. Early evidence of potential CUA might fast-track on-going innovation development. Limited prospects of CUA might raise questions about further allocation of resources to a product that has, ultimately, minimal chance to meet payer's expectations.

GAPS IN EARLIER HEALTH ECONOMIC EVALUATIONS
This last section (A) presented the current consensus and knowledge gaps around earlier HEEs and (B) highlighted opportunities for developments of a basic framework of preliminary CUA.

Benefits of earlier health economic evaluations
There is consensus around the weaknesses of full CUAs (e.g., lack of timeliness, resource-intensive) and strengths of early and preliminary CUAs, summarised in Table 3.
Earlier CUAs have the potential to assist promoters to: • Identify evidence gaps and headroom for improvements that generate insights into potential capacity of an innovation to alleviate the financial burden of prosthetic fittings. 37,50-52 • Educate the design of primary and modelling studies including the planning (e.g., calculate statistical power, determine of sample size, obtain ethics approval), collection (e.g., mine data from financial records, design

Obstacles to earlier health economic evaluations
Ijzerman and Steuten (2011) pointed out that the emerging field of HTA research will likely gain prominence as it will help navigate the increasingly complex trade-offs that must be considered when making investments in medical product development and ensuring access to those products. 37 However, earlier CUAs are far from being widely considered when developing new prosthetic solutions. Several obstacles must be overcome before earlier and, more particularly, preliminary CUAs of prosthetic care innovations would be routinely carried out by promoters.
One critical obstacle is the abundance of methods. Ijzerman and Steuten (2011) listed ten quantitative methods that could be used in earlier HTA (e.g., payback from research analysis, strategic business cases, health impact assessment, multi-criteria decision methods, choice-based preference methods, real options analysis, early health economic modelling, horizon scanning systems, clinical trial simulation, value-of-information analysis). 37 Another obstacle is the multiple pathways for HEE relying on the same level of clinical evidence of utilities (Figure 2). Logically, early and full CUAs are indicated at early stage and after clinical acceptance, respectively. Initial clinical evidence provided by proof of utility and case-series could be used to perform an early and preliminary CUAs. Stronger evidence gathered during cohort study and clinical trial might be deemed sufficient to conduct a preliminary or full CUAs.
Disparities of methods and constructs of earlier CUAs (e.g., perspective, time horizon, discount, uncertainty, sensibility) have ripple effects limiting implementation of earlier CUAs. Cross-comparing outcomes of earlier CUAs between innovations might be challenging to interpret. Generalization of outcomes across healthcare organisations might be limited. Earlier CUAs might show a broad level of quality when appraised with standard CHEERS and CHEC-extended checklists, primarily designed for full CUAs ( Table 1, Table 2). Altogether, disparity of outcomes also makes earlier CUAs scoring modestly in these checklists less likely to be published. The result of this is a sparsity of publications in prostheticfocused scientific journals, let alone heath economics journals, the latter of which are inclined to consider that socio-economic research in prosthetics is for a niche audience. Literature review and meta-analyses of health economic evaluations failing to stratify publications accordingly to the three types of CUAs might appraise unfavourably the contribution of earlier CUAs. 84, 85 Therefore, this review might skew the perception on the overall quality of health economic evaluations of prosthetic care. Earlier CUAs might score less not because they are poorly done but because they are dealing with more unreliable datasets.

Opportunities for basic framework of preliminary CUA
On a one side, every innovation is different. Each healthcare organisation has particular expectations. Promoters might choose a specific pathway for a given CUA depending on their confidence to make valid assumptions. Therefore, a preliminary CUA of an innovation could be unique.
One the other side, provision of prosthetic care follows a rather standardized process. Reimbursement are often made for categories of components (e.g., microprocessorcontrolled knees. 14 Prosthetists performed series of wellidentified specific tasks related to prosthetic fitting (e.g., fitting of socket, choice of components, alignment of prosthesis), assessment of outcomes (e.g., comfort, stability, mobility) and reporting to payers (e.g., reimbursement claims). 47 Indeed, each of these tasks is sufficiently codified to be individually supported by healthcare organisations (e.g., L-Codes). This means that most preliminary CUAs relying on estimated rather than primary costs could apply a template of schedule of allowable expenses. This typical matrix can present costs at the intersections of list of tasks in rows and timeline of interventions in columns (APPENDIX 1). Ideally, disruptive and economical innovations changing best prosthetic care practice should affect a schedule by reducing the price tag and/or the frequency of one or more tasks.
Furthermore, standard assessments are commonly used to quantify outcomes of prosthetic fittings using self-reported satisfaction (e.g., Orthotics and Prosthetics User's Survey, Quebec User Evaluation of Satisfaction with Assistive Technology, socket prosthetic comfort score), physical tasks (e.g., Berg Balance Scale, timed get-up and go, walking speed, 2-minute walk, 6-minute walk, functional ambulation profile, amputee mobility predictor with prosthesis) as well as specific (e.g., Questionnaire for Persons with a Transfemoral Amputation) and generic (e.g., EQ-5D, SF36) health-related quality of life with an innovation. 12,13,29,86 Altogether, organisation of the delivery and assessment of prosthetic care might be sufficiently transferable across innovations to consider a more uniform approach to preliminary CUAs. [50][51][52] This creates opportunities to explore the development of a basic framework including set Issue  constructs (e.g., perspective, time horizon, discount) and practical recommendations (e.g., funding cycles) specific to preliminary CUAs of the prosthetic care innovations (APPENDIX 1). This new approach to a preliminary CUA has the potential to simplify the selection of methods, standardise outcomes, ease comparisons between innovations and streamline pathways for adoption while facilitating the production of a body of literature on prosthetic health economics.

CONCLUSION
This work showed that promoters must make complex decisions when attempting to establish the socio-economic values of prosthetic care innovations. It is commonly acknowledged that a unique type of CUA could not be applied at every stage of development of an innovation. Preliminary CUAs of innovations at the mid-stage of development is particularly valuable but challenging. Boundaries delineating preliminary CUAs from early and full CUA might be blurry pushing promoters to consider a wide range of methods.
The outcomes suggest that there are opportunities for collective design of a basic framework of a preliminary CUA of prosthetic care innovations. However, reaching consensus around a framework can be challenging because there is no formal forum capable to organise discussions outside of usual scientific peer-review channels. There is a need for an ad-hoc reference group involving promoters and heath economists specialized in prosthetics and medical aids. Ideally, this working group should be hosted by international (e.g., World Health Organisation Standards for Prosthetics and Orthotics Service Provision, International Society for Prosthetics and Orthotics) or national (e.g., American Orthotic and Prosthetic Association, Center for Orthotic and Prosthetic Learning and Outcomes/Evidence-Based Practice) governing bodies. Its missions could be to develop guidelines and, possibly, standards of HEEs of prosthetic care interventions including preliminary CUAs frameworks (e.g., set constructs, practical recommendations).
Ultimately, a wide adoption of a this collegial preliminary CUA framework will, hopefully, contribute to promote the routinely used preliminary CUA. It is anticipated that this framework should facilitate access to economical prosthetic care solutions improving the life of individuals suffering from limb loss worldwide.

CALL TO ACTION
• Gather an ad-hoc reference group capable of (A) monitoring the current trends in HEEs of new healthcare technologies, (B) develop guidelines and, possibly, standards of HEEs of prosthetic care interventions, (C) promote the adoption of these guideline (e.g., publications of position papers, presentations at conferences).
• This working group could facilitate discussions between promoters of prosthetic care innovations around the use and validation of preliminary CUAs frameworks.
• Practically, these discussions should focus on the development of basic framework of a preliminary CUAs, more particularly set constructs and practical recommendations. Issue

AUTHOR SCIENTIFIC BIOGRAPHY
Dr Laurent Frossard is a bionic limbs scientist who is passionate about developing groundbreaking prosthetic solutions to improve the lives of individuals suffering from limb loss. He is internationally recognized as a researcher and an independent expert for his unique expertise in bionic limbs. He approaches bionic solutions from a holistic perspective, by integrating the prosthetic biomechanics, clinical benefits, service delivery, and health economics. Dr Frossard has over 25 years of experience, both in academia and in private industries in Australia, Canada, and Europe. He has collaborated with over 100 organizations worldwide. He is currently a Professor of bionics at the Griffith University, the Director and Chief Scientist Officer at YourResearchProject Pty Ltd, and Adjunct Professor at the Queensland University of Technology and the University of Sunshine Coast in Australia.