Use of removable support boot versus cast for early mobilisation after ankle fracture surgery: cost-effectiveness analysis and qualitative findings of the Ankle Recovery Trial (ART)

Objectives To estimate the cost-effectiveness of using a removable boot versus a cast following ankle fracture from the National Health Service and Personal Social Services (NHS+PSS) payer and societal perspectives and explore the impact of both treatments on participants’ activities of daily living. Design Cost-effectiveness analyses and qualitative interviews performed alongside a pragmatic multicentre randomised controlled trial. Setting Eight UK NHS secondary care trusts. Participants 243 participants (60.5% female, on average 48.2 years of age (SD 16.4)) with ankle fracture. Qualitative interviews with 16 participants. Interventions removable air boot versus plaster cast 2 weeks after surgery weight bearing as able with group-specific exercises. Primary and secondary outcome measures Quality-adjusted life years (QALYs) estimated from the EQ-5D-5L questionnaire, costs and incremental net monetary benefit statistics measured 12 weeks after surgery, for a society willing-to-pay £20 000 per QALY. Results Care in the boot group cost, on average, £88 (95% CI £22 to £155) per patient more than in the plaster group from the NHS+PSS perspective. When including all societal costs, the boot saved, on average, £676 per patient (95% CI −£337 to £1689). Although there was no evidence of a QALY difference between the groups (−0.0020 (95% CI −0.0067 to 0.0026)), the qualitative findings suggest participants felt the boot enhanced their quality of life. Patients in the boot felt more independent and empowered to take on family responsibilities and social activities. Conclusions While the removable boot is slightly more expensive than plaster cast for the NHS+PSS payer at 12 weeks after surgery, it reduces productivity losses and the need for informal care while empowering patients. Given that differences in QALYs and costs to the NHS are small, the decision to use a boot or plaster following ankle surgery could be left to patients’ and clinicians’ preferences. Trial registration number ISRCTN15497399, South Central—Hampshire A Research Ethics Committee (reference 14/SC/1409).

In general, it would be very helpf ul if the authors included more inf ormation on the trial in the Methods section.Although the detailed protocol f or the study is included in the Appendix, it is imperative that the paper summarizes the trial so that the readers who only read the manuscript have an overview of what was done.For example, inf ormation on study setting, participants (including inclusion/exclusion criteria) are needed.Also, the authors should describe the clinical results of the trial early on in the manuscript and indicate how that paper dif f ers f rom the current paper (i.e., reporting of clinical versus cost-ef f ectiveness results).
Please describe the sample size in more detail as it is ref erred to elsewhere (it is indicated that the trial was not powered to detect dif f erences in costs, QALYs).If the trial is not powered to detect dif f erences in costs, how does this af f ect the cost -ef f ectiveness analyses?
Why are the results of the trial which was conducted in 2015-18 being reported on now?This is perhaps not necessary to include in the manuscript itself , however, it is important to ref lect on if there have been any practice or intervention changes since 2015-18 that would af f ect the relevance of the results.
Please describe why participants were f ollowed up to 12-weeks (f or example, around line 28 in the Background).
In the Background, please include the rationale f or the trial, summaries of any other cost-ef f ectiveness studies of immobilization interventions f ollowing ankle f racture surgery, and an overview of the literature on the economic impact of ankle f ractures.
Furthermore, inf ormation on the trial registration and the CONSORT checklist should be included.Page 7, line 9 -perhaps reword the sentence as it seems to indicate that ref erence 10 is written by the authors.
For the economic inf ormation, please indicate the year and currency of the results.Also, include a sentence on why discounting was not used.If the study was conducted in 2015-18, why are 2021-22 wage rates used?Were the costs adjusted f or inf lation and if so, how was this done?
For the qualitative study (page 10, lines 35-42), what was the response rate?What was the mean and median age of participants?Discussion -page 15, line 9 -f irst sentence may need rewording "The f ound...".Lines 16-20 -the qualitative results may be overstated as it is dif f icult to compare personal opinions.The sample size f or the qualitative study should be indicated here again (n=...).
Results in context -page 16 -please provide more inf ormation on the studies that have been conducted to date on this topic (e.g., sample sizes, locations, limitations) and how they compare to your results.
Page 16, line 13 -please cite the paper on the clinical results f rom the trial.Please also include more inf ormation in the Background and Discussion on the clinical results that were f ound in the trial.
Page 19 -Patient consent f orm -it states in this section that no person's data are included in the manuscript, however, the manuscript involves quotes f rom participants in the qualitative study with specif ic inf ormation provided (e.g., participant who mentions MTV Clashes in Plymouth).
Table S1 -page 31 -please include year of costs.For hospital procurement, when were the data obtained (e.g., year)?

GENERAL COMMENTS
It is a well written manuscript with originality and comprehensible review of the literature.We believe that this is an interesting topic f or f oot and ankle surgeons and that this knowledge is usef ul f or clinical practice.It could be benef icial f or the scientif ic community as it is a prospective study with clear message.
More specif ic comments are the f ollowing:

INTRODUCTION
In the introduction section, there has been done a meticulous work to provide a short but adequate background of the study.Besides, the importance of the study is also designated through this section and the purpose is also clearly cited.

METHODS
The research methodology is detailed and well written.

RESULTS
The outcomes of the study are presented comprehensively.However, the results section is long, please make it more concise.

DISCUSSION
In the DISCUSSION, the content is concise and well written.Additionally, the clinical relevance of the study is well presented and the limitations of the study are plainly analyzed.

FIGURES
No comment.

TABLES
No comment.
While Cost-ef f ectiveness analysis is a great methodology, the results (estimates) obtained f rom it are quite unstable and can easily change depending on assumptions.I am not that f amiliar with statistics so we should check with a statistician to see if there are any problems.
VERSION 1 -AUTHOR RESPONSE Reviewer: 1 Dr. Iwona Bielska, McMaster University, Jagiellonian University Comments to the Author: Thank you f or giving me the opportunity to review this manuscript, which describes the results of a trial examining two types of immobilization f ollowing ankle f racture surgery.This is an important area of study and there is limited inf ormation on the cost -ef f ectiveness of various interventions.The authors supplement the quantitative inf ormation with qualitative data collected f rom a subset of participants, which provides a patient perspective on the use and type of immobilization.
We thank the reviewer f or the positive f eedback on the paper.R.1.1.In general, it would be very helpf ul if the authors included more inf ormation on the trial in the Methods section.Although the detailed protocol f or the study is included in the Appendix, it is imperative that the paper summarizes the trial so that the readers who only read the manuscript have an overview of what was done.For example, inf ormation on study setting, participants (including inclusion/exclusion criteria) are needed.
Thank you f or this suggestion.We have added more inf ormation of the trial in the Methods section, e.g.inclusion and exclusion criteria, registration number.
'A health economic evaluation and qualitative component were included in the ART pragmatic multi-centre RCT to provide a more inclusive picture.The trial was conducted across eight UK NHS secondary care trusts between July 2015 and November 2018 and compared casts with removable support boots, combined with weight bearing as able, as methods of ankle f racture management f rom two weeks post-surgery.Most participants at the 6-week f ollow-up timepoint were expected to remove their plaster cast or boot and start to mobilise f reely.Questionnaires were completed at randomization (2 weeks af ter surgery), then at 6-, 7-, and 12-weeks post-surgery.The primary clinical endpoint was the Olerud and Molander ankle score (OMAS) 14 measured at 7-weeks post-surgery, when participants should have adapted to mobilising without plaster cast or boot.The ART trial was approved by the South Central -Hampshire A Research Ethics Committee (ref erence 14/SC/1409) and was prospectively registered (ISRCTN: 15497399)…' (See in Methods/ Trial design and ethics approval) 'All adults (>16 years old) undergoing open reduction and internal f ixation (ORIF) surgery f or an unstable ankle f racture were eligible f or the study.Exclusion criteria were: open ankle f racture; concern about quality of f ixation/wound integrity; requiring f urther stabilisation in/around the ankle (e.g.syndesmosis); active leg ulceration; poor skin condition at operating site; serious concomitant disease, diabetic/other sensory neuropathy; non-ambulatory prior to injury; inability to understand/complete outcome questionnaires; enrolment in other interventional research which may conf ound data collection; concomitant injuries which may af f ect rehabilitation.'(See in Methods/ Participants) R.1.2.Also, the authors should describe the clinical results of the trial early on in the manuscript and indicate how that paper dif f ers f rom the current paper (i.e., reporting of clinical versus costef f ectiveness results).
The clinical paper will report the results on the primary outcome measure of the trial, the Olerud and Molander Ankle Symptom Score (OMAS) at week 7, and on secondary outcomes, such as OMAS scores at week 12, EQ-5D VAS scores at week 7 and 12, secondary mechanistic measures such as ankle ROM (dorsif lexion, plantarf lexion, inversion and eversion), calf circumf erence, and ankle swelling at 6 weeks , as well as weight-bearing status, use of walking aids, return to driving and work, impact on daily activities at weeks 6 and 12. Complications and serious adverse events are also reported at 6 and 12 weeks in the clinical paper.These can be also f ound in the Final report to the Funder which is available f rom the authors upon request.
We now f ully specif y the trial outcomes in page 7 of the methods: 'Results on clinical ef f ectiveness (such as primary outcome measure and secondary outcomes, such as OMAS at week 12, EQ-5D VAS scores at week 7 and 12, secondary mechanistic measures at 6 weeks, weight-bearing status, use of walking aids, return to driving and work, impact on daily activities at weeks 6 and 12, as well as complications and serious adverse events at 6 and 12 weeks) can be f ound in the f inal report to the Funder 13 and will be reported in a separate paper.' (See Methods/ Trial design and ethics approval) This cost-ef f ectiveness paper results on costs f rom the NHS+PPS and social perspective, on QALY-s, and on incremental net monetary benef it statistics (INMB) measured 12-weeks af ter surgery, f or a society willing-to-pay £20,000 per QALY.
We have also added the f ollowing summary of the main results of the clinical paper in the Background section: 'Clinically there were no observed dif f erences in ankle f unction between treatment groups at any time point.Overall early weight bearing in both the boot or cast was saf e with low wound complication rates (7%) that, although more f requent in the boot group, were minor.' (See Background section, 4 th paragraph) R1.3.Please describe the sample size in more detail as it is ref erred to elsewhere (it is indicated that the trial was not powered to detect dif f erences in costs, QALYs).If the trial is not powered to detect dif f erences in costs, how does this af f ect the cost -ef f ectiveness analyses?
The study was powered to the primary outcome, the OMAS at 7 weeks post -surgery as the primary outcome measure based on an unpaired t-test with two-sided signif icance level (α) of 0.05 and 90% power (β), standard deviation of 21.9 and mean between group dif f erence of 10 points on the OMAS (minimum clinically important dif f erence).To achieve this, a total sample size of 204 (102 per group) was required, The study aimed to over-recruit by 20% to accommodate non-responders and missing data not to contribute data to the main analysis, giving a recruitment target of 246.
RCTs are usually not powered to detect dif f erences in costs .As per Briggs 2000 [1]: Given the skewed nature of cost data and consequently higher variance of cost variables than of clinical outcomes as well as the heterogeneity of costs, the comparisons of treatment cost as well as cost-ef f ectiveness ratios would require greater sample sizes than the corresponding clinical comparison [1].If RCTs would be powered to the economic outcomes, they would most likely be overpowered with respect to the clinical outcomes.This might be problematic f rom the ethical point of view as it would probably be inappropriate to continue a trial beyond the point at which clinical superiority has been determined [1].
Not powering the trial f or economic outcomes mean that we might not detect signif icant dif f erences, costs, QALYs, or the economic result (incremental net monetary benef its) and our estimates would result in larger conf idence intervals around cost -ef f ectiveness results.Thus, we have to be caref ul when interpreting such results as we cannot be sure if dif f erences do not exist or just not detected by the study.In this evaluation, the dif f erences we f ound in QALYs are so minor, that we do not expect any meaningf ul dif f erence even if we increased sample size.Dif f erence in costs was signif icant f rom the NHS+PPS perspective, and we do not expect our conclusions to change with increasing sample size.It may be possible that increasing sample size would render the smaller dif f erence in costs f rom the societal perspective signif icant, by narrowing the conf idence intervals of INMB results, making the boots the more cost-ef f ective option.
We added some of this explanation to the Limitation section: 'The ART trial was powered to detect a dif f erence in the primary outcome (OMAS score at 7 weeks), and not dif f erence in costs, QALYs, or INMB statistics, consequently there is more uncertainty in the economic results, especially f rom the societal perspective.' (See in Methods/ Trial design and ethics approval) R.1.4.Why are the results of the trial which was conducted in 2015-18 being reported on now?This is perhaps not necessary to include in the manuscript itself , however, it is important to ref lect on if there have been any practice or intervention changes since 2015-18 that would af f ect the relevance of the results.
We thank the reviewer f or raising this point.There was indeed some delay in publishing the results of this trial due to career breaks (maternity and sick leaves and the COVID pandemic).The f inal report of the randomised controlled trial was submitted to the Funder on 18 March 2022, af ter the of f icial end date of the trial, 30 April 2019.Although there may have been changes in the surgical treatment of ankle f ractures af ter the results of the AIM trial [2], management of post-surgical treatment has not changed noticeably and our study and its f indings are still relevant today.
We agree with the reviewer that this is a limitation that needs to be mentioned, thus we added the f ollowing sentence to the Limitations section: 'Despite data collection taking place several years ago, we believe our results are relevant and transf erable to current NHS practice.'(See Discussion/Limitation) R.1.5.Please describe why participants were f ollowed up to 12-weeks (f or example, around line 28 in the Background).
At the six-week appointment participants had their cast or boot removed, and their recovery (without the boots and cast) were f ollowed up f or 6 weeks, by the time they were expected to return to their normal activities.This corresponds with f inding of a previous study on the economic burden of ankle f ractures, where patients returned to work af ter 11.2 weeks on average [3].We have added this explanation to Method section: 'At the six-week appointment participants had their cast or boot removed (as appropriate), and were f ollowed up f or an additional 6 weeks, by which time they were expected to return to normal activities.' (See in Methods/ Intervention and usual care) R.1.6.In the Background, please include the rationale f or the trial, summaries of any other costef f ectiveness studies of immobilization interventions f ollowing ankle f racture surgery, and an overview of the literature on the economic impact of ankle f ractures.
We thank the reviewer f or raising this point.According to our knowledge, there are no other cost-ef f ectiveness studies have been published so f ar that compares the boot and cast f or ankle f racture.We added to the text the f ollowing: 'Systematic reviews conclude that well-designed, prospective studies are still needed to determine which immobilization technique of fers the most benef it f or treating ankle f ractures. 712 .Furthermore, evidence on cost-ef f ectiveness of boots compared to casts is missing f rom the literature. 12' (See in Background 3 rd paragraph) We added some sentences on the economic impact on ankle f ractures to the Background section as well: 'According to a recent systematic review, the direct costs of ankle f ractures ranges f rom $1,908 to $19,555 worldwide. 5In England, the cost of inpatient hospital care f or ankle f ractures was estimated to exceed £63.1 million in 2016/17. 1 Costs f rom the social perspective are even higher due to indirect costs associated with f ractures, such as productivity loss and inf ormal care costs, that contribute to almost half of total costs. 6' (See in Background 1 st paragraph) R.1.7.Furthermore, inf ormation on the trial registration and the CONSORT checklist should be included.
The study was approved by the Health Research Authority (Ref : 14/SC/1409) and prospectively registered (ISRCTN: 15497399).The ref erence is added to the text.(See in Methods/ Trial design and ethics approval) Since we do not report the ef f ectiveness results of the randomised controlled trial, we are using the CHEERS 2022 checklist f or reporting results of an economic evaluations [4] which also applies to the reporting of economic evaluations alongside clinical trials.We added a ref erence to the checklist in the paper and attached the checklist as supplementary material.(See in Methods/ Overview of economic evaluation) R.1.8.Page 7, line 9 -perhaps reword the sentence as it seems to indicate that ref erence 10 is written by the authors.
The sentence reads: "All analyses f ollowed the pre-specif ied health economics analysis plan [10]…" The economic analysis plan f or the ART trial was indeed written by some of the authors of this paper at the beginning of the trial.Please f ind the ref erence to the HEAP below: Thomas P, Marques EMR, Docherty S, Sangar S: Statistical and Health Economics Analysis Plan: ART Ankle Recovery Trial version 1.1.University of Bristol.28th Nov 2018.Available at: https://research-information.bris.ac.uk/en/publications/statistical-and-healtheconomics-analysis-plan-art-ankle-recoveryAccessed: 26/01/2023 Peter Thomas retired in 2020 and therefore is not an author on this paper.R.1.9.For the economic inf ormation, please indicate the year and currency of the results.Also, include a sentence on why discounting was not used.If the study was conducted in 2015 -18, why are 2021-22 wage rates used?Were the costs adjusted f or inf lation and if so, how was this done?Detailed unit costs (including currency and years f or all item) are presented in Table S1 in the Appendix.We agree with the reviewer that some of these data should be added to the main paper as well.We present the results in 2020/21 prices in GBP.Primary care services, physiotherapy and community based services were valued using the unit cost database of health and social care prof essionals 2020/21 (i.e, 2020/2021 prices).The latest available unit costs (National Schedule of NHS Costs 2019/20) that were available at the time of analysis were used to evaluate resource use f or secondary care services: secondary care services (Accident and Emergency, Day case admission or procedure, Outpatient appointment).These costs were inf lated to the 2020/21 values using the NHS Cost Inf lation Index (NHSCII) f or Pay & Prices.Also, private expenses by patients reported at the time of the survey were inf lated to 2020/21 prices.We used wage data f rom 2021 to match the date of the direct cost data.We added these details to Methods/Valuing resource use to derive costs where the years were missing.R.1.10.For the qualitative study (page 10, lines 35-42), what was the response rate?What was the mean and median age of participants?Mean age of participants was 49.19, median age was 50.5 Of the 19 patients that were approached 3 declined to participate.We have added these details to the paper.(See Results/Sample) R.1.11.Discussion -page 15, line 9 -f irst sentence may need rewording "The f ound...".
Thanks f or pointing out this typo, we have corrected it."We f ound.." R.1.12.Lines 16-20 -the qualitative results may be overstated as it is dif f icult to compare personal opinions.The sample size f or the qualitative study should be indicated here again (n=...).
We have added the sample size and rephrased the sentence: 'Qualitative interviews with sixteen participants complemented the quantitative analysis to better understand patients' perspectives.Lower productivity losses and inf ormal care need in the boot group were ref lected in the interviews, and some participants in the boot group highlighted that they were more satisf ied with their treatment option, f elt more independent, and were able to return to their usual activities sooner."(See Discussion/ Statement of principal f indings)' R.1.13.Results in context -page 16 -please provide more inf ormation on the studies that have been conducted to date on this topic (e.g., sample sizes, locations, limitations) and how they compare to your results.
According to our knowledge, this is the f irst cost -ef f ectiveness evaluation alongside a trial comparing boots with cast.Nevertheless, a f ew other studies have also compared resource use, return to work or to normal activities between f unctional treatment and cast groups, or reported EQ-5D-5l results.We have added the f ollowing to the Results in context section.(See Discussion/Results in context) 'Nevertheless, a f ew other studies have also compared resource use, return back to work or to normal activities between f unctional treatment and cast groups.For example, Egol et al (2000) reported earlier return to work in the brace group f or patients who underwent internal f ixation 29 , however these patients were all instructed to avoid weight -bearing on the af f ected side f or six weeks.Honignmann (2007) f ound that f or patients who had malleolar f ractures f ollowed by open reduction and internal f ixation, p atients in the orthesis group with prescribed f ull weight bearing returned to work sooner, however the dif f erence was not signif icant 30 .Simansky et al (2006) and Lehtonen et al (2003)  In our study, no dif f erences were observed in how soon participants returned to driving or f ull pre-injury work duties.The boot group indicated at the 6-week post-operative time point that their injury was having a greater impact on daily activities; however, this dif f erence was no longer present 4 weeks later.
Similarly to our results, Kearney et al. (2021) andHaque et al. (2023) f ound no signif icant dif f erence between brace and cast groups in terms of EQ 5D -5L scores at any time point 32 33.' R.1.14.Page 16, line 13 -please cite the paper on the clinical results f rom the trial.Please also include more inf ormation in the Background and Discussion on the clinical results that were f ound in the trial.
The main results of the clinical paper are now summarised in the Background section(see Response to comment R.1.1.and R.1.2.) The clinical paper has been draf ted but has not been submitted f or publication yet, but it is aimed to be submitted in 2023.We ref er to the Final report submitted to the Funder instead.R.1.15.Page 19 -Patient consent f orm -it states in this section that no person's data are included in the manuscript, however, the manuscript involves quotes f rom participants in the qualitative study with specif ic inf ormation provided (e.g., participant who mentions MTV Clashes in Plymouth).
We have removed those ref erences to the place and event f rom the quote, to make sure the person is not identif iable.(See Table 4) R.1.16.Table S1 -page 31 -please include year of costs.For hospital procurement, when were the data obtained (e.g., year)?
The year of costs is included in the 2 nd column f or the items separately.For hospital procurement, we added the year (2022) when the data was obtained.
Reviewer: 2 Dr. Kentaro Amaha, St Luke's International University Comments to the Author: It is a well written manuscript with originality and comprehensible review of the literature.We believe that this is an interesting topic f or foot and ankle surgeons and that this knowledge is usef ul f or clinical practice.It could be benef icial f or the scientif ic community as it is a prospective study with clear message.

Iwona Bielska
McMaster University, Department of Health Research Methods, Evidence, and Impact REVIEW RETURNED 12-Nov-2023

GENERAL COMMENTS
Thank you f or the opportunity to review the revised version of the manuscript.I would like to acknowledge that the authors caref ully addressed all of my questions and comments f rom the f irst review and provided detailed answers and additions to the manuscript, that greatly strengthened it.I have no f urther major comments.
Minor comments: -Page 15, line 34: the word "f ound" seems to be missing in the Simansky sentence.
-For the "Patient consent f orm" section on page 18, the authors write "No individual person's data are included in the manuscript."However, quotes f rom the qualitative interview are included in the Appendix.As such, I am not sure if this sentence should be revised or lef t as is.
Thank you once again f or giving me the chance to review your paper.
no signif icant dif f erences in return to work comparing f unctional treatment versus cast patients who underwent open reduction and internal f ixation 11 31.