Development of a government continuous quality improvement procedure for assessing the provision of bone anchored limb prosthesis: A process re-design descriptive study

BACKGROUND. Evidences of sustainable clinical benefits of bone-anchored prosthesis (BAP) using osseointegrated fixation over typical socket-suspended prostheses are becoming more probing. This influx of individuals to be fitted with BAP has pressed government organisations to adjust their policies. However, the appraisal of consumer’s experience for the provision of BAP founded by government organisation is yet to be developed. This descriptive study shares the experience gained by a government organisation, namely the Queensland Artificial Limb Service (QALS), while developing a specific BAP-inclusive continuous quality improvement (CQI) procedure. OBJECTIVE(S). The primary objective was to present the methods and outcomes of key steps required to plan and create this CQI procedure. The secondary objective was to highlight key barriers and facilitators of the transition from a socket-focused to the Abstract There is evidence supporting the long term clinical benefits of bone-anchored prosthesis (BAP) using an osseointegrated fixation over typical socket-suspended prostheses. The increasing number of individuals treated with osseointegrated fixation has pressed government organisations to adjust their policies for fair and equitable provision of prosthetic care. However, the appraisal of consumer’s experience for the provision of BAP by government organisation is yet to be developed. This descriptive study has fulfilled this need by sharing the experience gained by a QALS while developing a specific BAP-inclusive continuous quality improvement (CQI) procedure. This study revealed that government organisations can design a CQI procedure for comprehensive appraisal of the provision of prosthesis that could be inclusive of BAP and affordable while minimally time-consuming for prosthetists. The transition from a socket-focused to the proposed minimally disruptive BAP-inclusive CQI procedure was facilitated by prior knowledge of BAP treatment, early identification of the stakeholders and adaptation of current CQI procedure.


Health services delivery of bone-anchored prosthesis
This influx of individuals fitted with osseointegrated fixation has pressed government organisations, like the Queensland Artificial Limb Service (QALS), to adjust their policies for fair and equitable provision of BAP. [18,29] Indeed, QALS established such procedure allowing financial assistance for consumers choosing BAP that involves seven processes costing AUD$3,300 for 22 hours of labour per patient during the treatment. [29] Furthermore, cost cross-comparing and costeffectiveness demonstrated health economic benefits of BAP over SSP from government perspective. [30][31][32] For instance, provision of BAP costed 21±41% more but increased quality-adjusted life-year by 17±5% compared to SSP leading to an indicative incremental cost effectiveness ratio of approximately $17,000 per quality-adjusted life-year. Despite a partial compensation of the cost by qualityadjusted life-year, the provision of BAP was deemed costeffective since the incremental cost effectiveness ratio was noticeably below willingness to pay threshold. [33] Need for government continuous quality improvement procedure Appraisal of consumer's experience for the provision of BAP by government organisation is yet to be developed. A series of standardised surveys could assess delivery of particular prosthetic care and/or experience with prosthetic components (e.g., SERVQAL, OPUS, QUEST). [34][35][36][37][38][39][40][41] However, their relevance to provision of specific BAP care by government organisations is limited.
QALS has carried out a SSP-focused continuous quality improvement (CQI) procedure detailed below that has emerged through regular revisions over the last decade in response to best practice and legal obligations imposed by government as well as expectations from consumer advisory groups. Whilst some aspects of this CQI procedure are relevant to the provision of BAP, there is a need to further investigate developments of CQI procedure capable of appraising consumer's experience for the provision of BAP by government organisation.

Objectives
The aim of this work was to improve government health service delivery of prosthetic care specific to individuals fitted with BAP.
The purpose of this process re-design descriptive study was to share the experience gained by a government organisation while developing a specific BAP-inclusive CQI.
The primary objective was to present the methods of model re-design with particular emphasis on outcomes of key steps required to plan (e.g., case-mix profiling, stakeholder analysis) and create (e.g., diagnosis, pros-cons analysis of technical options, cost) this specific CQI procedure. The secondary objective was to highlight key barriers and facilitators of the transition from a pre-existing SSP-focused to the proposed BAP-inclusive CQI procedure.

Setting
This study was undertaken by QALS in the jurisdiction of the Queensland State Government Minister of Health, Australia. With a yearly budget of $5.4 million, QALS provides equitable funding for prosthetic services to 3,600 active consumers annually through a network of up to 10 individual prosthetists referred to as Prosthetic Service Provider (PSP). [29,30,33]

Participants
The development of the CQI procedure was led by a QALS steering committee including QALS management team, two researchers in health services, three PSPs and the five first consumers representing 8% of the QALS's population fitted with BAP. [42,43]

Study design
The descriptive study started in July 2015, shortly after the whole QALS's procedure for provision of BAP was completed. [29] As detailed in Table 1, the development of the specific CQI procedure was achieved using the following two-step re-design process:  Step 1 to plan the procedure, including: o Step 1A identifying problems to solve using root cause analysis that involved case-mix profiling achieved by looking at typical demographics, amputation, as well as access to prosthetic care and funders data extracted from QALS client information system. o Step 1B identifying deliverables of the CQI procedure that involved stakeholder analysis using typical matrix ranking selected organisations in relation to their power and interest in CQI depending on capacity to influence allocation of resources and to provide prosthetic and medical care, respectively. Stakeholders were classified as controllers (high power, low interest), promoters (high power, high interest), providers (low power, high interest) or advocates (low power, low interests). [44]  Step 2 to create procedure including: o Step 2A diagnosing quantitatively the suitability of the current SSP-focused CQI procedure by counting the number of sections, questions and possible answers in each evaluation form and by categorising the focus of each question as administration (e.g., consumer's identification, processing status, quality control), service (e.g., labour associated with provision of prosthetic services and/or components), prosthesis (e.g., provision of repair, fitting, replacement of prosthetic components and/or cosmetic cover), socket (e.g., light, definitive) or BAP (e.g., provision of all interventions to fit a BAP). Questions focusing on administration, service and BAP were considered relevant while those focusing on prosthesis and socket were deemed partially suitable and irrelevant to appraisal of provision of BAP, respectively. o Step 2B exploring options for new specific CQI procedure that relied on pros-cons analysis of pathways and products investigated for administration and analysis of surveys and content management system (CMS). o Step 2C adapting the existing CQI procedure to each phasis of BAP treatment, adjusting forms and estimating participants' typical time commitment. [29] Cost for PSPs' contribution and internal labour (e.g., QALS staff time) allocated to CQI procedure (e.g., data collection, entry and reporting) was set at hourly fee of $160 and $30, respectively. All costs are reported in Australian dollars (1 Australian dollar ≈ 0.63 Euro ≈ 0.56 British pound ≈ 0.74 US dollar) according to 2017-18 prices. In all steps, the steering committee considered critical qualitative and quantitative information and applied a typical standards for interactive inquiry process and datadriven collaboration leading to consensus. [45] * Table 1 *

Definition of specific procedure Case mix profile
The characterisation of case-mix presented in Table 2 involved the 65 QALS consumers fitted with lower limb BAP since 2011, representing 16% and 7% of existing BAP population estimated at 400 in Australia and 950 worldwide, respectively.

Stakeholders analysis
Twenty key stakeholders were identified with half operating at state or national levels as presented in Figure  1C and further detailed in Figure 1 of Supplement. No stakeholder was identified as controllers. As expected, the six (30%) promoters involved the decisional entities around QALS including consumer advisory group, in particular, as well as national government funding agencies. The seven (35%) providers involved all health professionals in the clinical teams responsible for osseointegration treatments in state and interstate, including essentially prosthetists. The seven (35%) advocates included mainly consumer support groups and professional associations as well as other artificial limbs services across Australia.

Creation of specific procedure Diagnosis of initial CQI procedure
The appraisal of QALS' provision of prosthetic services involved a series of evaluations supported by three paperbased forms including a total of 73 items allowing 240 possible answers. As detailed in Table 3, the review this CQI showed that:  Evaluation A, involving seven steps, relied on Prosthetic Issue Document (PID) to acknowledge PSP's service that triggers QALS' reimbursement. Circulated by mail, the PID included 18 (25%) of all the questions that were completed by PSPs and consumers after each service. A total of 61% of the questions focused on the whole prosthesis.  Evaluation B, involving four steps, relied on a Validation of Services (VOS) form designed to assess a consumer's satisfaction with quality of the prosthetic service delivered by PSP. The VOS included 25 (34%) of all the questions that were completed by QALS staff while talking to consumers over the phone after each service.  Evaluation C, involving four steps, relied on Prosthetic Service Evaluation (PSE) form designed to assess overall consumer's experience with service provided by QALS. The PSE included 30 (41%) of all the questions that were completed by consumers yearly. Overall, 51%, 40% and 9% of the questions were relevant, partially suitable and irrelevant to the development of CQI for BAP, respectively.
The content of the paper version of each form was manually tabulated by QALS staff into a purposely-designed CMS, easily adjustable in-house in response to stakeholders' regular changes in reporting expectations, including a series of spreadsheets organising entries, analysis and reporting of consumer experience information.

Explore options for specific CQI
Opportunity for redesigning a specific CQI procedure relying on new pathways, forms and cloud-based technological platform was initially investigated. [37,46,47] Quote from external professional provider with relevant programming skills indicated that such project will require approximately 200 hours of labour at the cost of $33,000.
Alternatively, keeping the current delivery pathway and adjusting forms and CMS was considered. We made the assumption that these adjustments could be achieved in approximately 120 hours for in-house knowledgeable staff labour at an internal cost of $3,600.
The latter option was deemed the most sensible and cost-effective.

Creation of BAP-inclusive CQI procedure
A dynamic overview of the proposed BAP-inclusive CQI procedure in Figure 1 detailed the intersections between phases of the treatment and each of the three sequential evaluations with emphasis on contribution of participants, documents and forms used, the tasks achieved, CMS used to collect, analysis and report consumer experiences outcomes. Evaluations A and B were required to be completed after each service that occurred typically at least four times during the first year of the BAP treatment between pre-operative consultation and fitting of definitive prosthesis. Evaluation C occurs usually at least once a year after delivery of definitive prosthesis.
The adaptation and reorganisation of all forms was achieved by implementing basic principles of so-called computerized adaptive testing. [48] The first part of all forms involving administration items, mainly focusing on identification of consumer and processing information, remained unchanged. However, a two-answer routing question was added at the end of administration section asking consumers what type of attachment they use. The SSP users were directed to the second part including essential questions in current forms related to socket and prosthesis cleaned of any BAP related items. Those using BAP were directed to the third part including newly developed questions. Practically, this third part in PID and VOS forms required consumers to indicate at which of the five stages of the treatment they were at. Developments of the third section in the PSE form was more involved and lead to design of custom-made survey including 32 questions as outlined in Table 4 to assess clinical outcomes in six domains. [49] Benefits were assessed into two domains including the health-related quality of life and mobility outcome using the standardized self-report Short Form 12V2 Health Survey (SF12) and Prosthetic Limb Users Survey of Mobility (PLUS-M™) 12-item short forms, respectively. [39,48] Safety was assessed by self-reporting selected adverse events into four domains including fixation stability, fixation integrity, injuries as infections. [5] * Table 4 * As presented in Table 5, resources provided to PSP's efforts toward CQI procedure represented 1.3 hrs out of 22 hrs and $213 out of $3,300 or 6% of the whole procedure for the provision of the BAP detailed previously. [29] The time spend by PSP, consumer and QALS staff represented approximately, 24%, 24% and 53% of the time of the whole procedure, respectively. The cost for reimbursement of PSP's and QALS staff labour represented 70% and 30% of the total costs of the CQI administration per consumer, respectively. Altogether, the typical total cost per consumer for the first year of treatment with BAP was approximately $416. * Table 5 *

DISCUSSION Outcomes
This study revealed that a government organisation, such as QALS, can redesign a CQI procedure for comprehensive appraisal of the provision of prosthesis that could be inclusive of BAP while been minimally timeconsuming for PSPs and affordable.
The transition from a SSP-focused to the proposed minimally disruptive BAP-inclusive CQI procedure was facilitated by the following redesign inputs:  Capitalising on prior knowledge. Initial understanding of specific rehabilitation program following BAP treatment was gained during development of the QALS overall procedure to support provision of BAP. This elucidated involvements of PSP in the delivery of services and components during treatment that was essential to determine workload and cost. [29]  Identification of the stakeholders. Early selection, organisation of key stakeholders helped to ascertain common and separate expectations and subsequently prioritise reporting requirements.  Adapting current CQI procedure. Redesigning a BAPspecific CQI might lead to increased delivery efficiency and, more importantly, suitability of tailored forms providing distinctive results for this group of consumers. However, such parallel CQI procedure has several shortcomings including, but not limited to, confusion of consumers used to initial CQI, significant cost required to build dedicated CMS, lack of consistency in reporting limiting benchmarking with other consumers. [49] Alternatively, keeping the initial procedure relying on three evaluations but adapting the forms and CMS was deemed the most sensible and cost-effective option.
The main obstacles to the development of BAPinclusive CQI procedure were associated with adjustments of PSE form, particularly the selection of relevant outcomes to consider. Only confounders of customer's experience responsible for potential cross-correlation with provision of particular components and clinical outcomes were selected from classification of benefits and safety of BAP treatment presented by Frossard (2015) further detailed in Figure 2 of Supplement. [49] Also challenging was to balance the selection of outcomes (inclusion vs exclusion), the choice of instrument to measure selected outcomes (e.g., preference on validated over in-house design self-reported surveys) and the overall length of the survey (e.g., preference on short over long forms of surveys). Generic health-related quality of life measures were achieved using validated and widely used SF12 because the outcome could be readily benchmarked and converted in quality-adjusted life-year required for subsequent cost-utility analyses. [31,33] Mobility outcomes associated with actual usage of the prosthesis using T-score from PLUS-M™ could be supplemented by physical tasks such as Time Up and Go and 6-Minute Walk accessible from consumer's passport completed by PSP after fitting of definitive components as described in , if needed. [29] For the safety outcomes, a decision was made to discard issues of soft tissues management, skin at stoma interface and phantom pain as they have no established links with prosthetic components. Adverse events associated with fixation stability and integrity as well as injuries and infections were applicable since they might be inherent with the load generated by prosthetic components and, therefore, might have potential legal bearings. Another hurdle to overcome was the lack of validated instruments to report advert events and complications. Alternatively, a custommade survey involving a short series of eight questions was collectively elaborated, pilot tested with selected consumers and implemented.

Limitations
The PSE form might be deemed onerous by some consumers because of redundancy of SF12 and PLUS-M with regular follow ups conducted by treating clinical teams. Purposely designed survey lacked typical statistical validation. Other limitations derived from typical intrinsic shortcomings of prospective study presenting the initial steps of action research cycle. Beyond the scope of this study, the lack of actual long-term consumer's experience data limited the validation of this proposed CQI procedure.
The generalization of the outcomes must be considered carefully. The proposed CQI was purposely designed to fulfil specific needs for an Australian State organisation providing funding for prosthetic care only. However, stakeholders and treatment pathways for provision of BAP could differ between jurisdictions, particularly in European and North American countries. [33] Indeed, the scalability of this CQI procedure within and between jurisdictions is yet to be established, particularly its capacity to integrate requests from broader stakeholders, the geographical spread of consumers extending to rural areas with limited access to a PSP, the increasing number of treatment sites in Australia and abroad as surgeries are more routinely performed. Nonetheless, a series of valuable insights provided could be readily integrated by other organisations while customizing their own BAP-inclusive CQI procedure, including the importance of understanding rehabilitation programs, identification and organisation of the stakeholders (e.g., local, regional, national), benefits and ways to adapt existing procedures (e.g., pathways, forms and CMS), methods to determine involvement of participants (e.g., consumers, PSPs, funder) as well as consideration for confounders of customer's experience with provision of BAP (e.g., clinical benefits and safety).

Future studies
Future developments of the proposed CQI procedure will be facilitated by additional longitudinal studies providing experience outcomes for a large cohort of BAP users over an extended period of time that could be benchmarked against other BAP or SSP users.

CONCLUSIONS
The early development of a CQI procedure, including the management of barriers and transferable facilitators, to appraise the provision of BAP by a governmental organization was shared for the first time. This work was an initial effort toward the assessment of fair and equitable governmental financial assistance programs for individuals choosing BAP. Altogether, this study should be considered as a stepping-stone providing a working approach for BAP-inclusive CQI to other organizations worldwide.

ACKNOWLEDGEMENT
The authors wish to express their gratitude to John Vasil for his contribution to the development of this project as well as Fiona Barnett, Stephan Laux, Barry Leech and Luke Lorenzin for their insight into prosthetic care.

DECLARATION OF CONFLICTING INTERESTS
The authors have no conflict of interest. AUTHOR CONTRIBUTION  Laurent Frossard has developed the study design including collection, analysis, presentation the data as well as writing this manuscript.  Luciann Ferrada has contributed to the collection and analysis of the data as well as the writing of the manuscript.  Tanya Quincey has contributed to the collection and analysis of the data as well as the writing of the manuscript.  Brendan Burkett has contributed to the writing and reviewing of the manuscript.  Debra Berg has lead the whole project including the study design, collection, analysis, presentation the data as well as writing this manuscript. Table 1. Timeline and actions of the two-step process taken by the Queensland Artificial Limb Services (QALS) to develop specific Continuous Quality Improvement (CQI) procedure to appraise consumer's experience for provision bone-anchored prostheses (BAP   QALS  38  58  ----DVA  8  12  ----NDIS  12 18 ----

Mean SD Range Participants
Assessing the provision of bone-anchored limb prosthesis Canadian Prosthetics & Orthotics Journal. 2018. 1 (2) Page 10 of 15 Table 3. Overview of structure with number of sections, questions and possible answers and percentage of questions focusing on administration, service, prosthesis, socket or bone-anchored prosthesis (BAP) for each initial form of the continuous quality improvement (CQI) procedure used by Queensland Artificial Limb Services (QALS  Table 4. Overview of 32 questions asked in third part of the Prosthetic Service Evaluation (PSE) form extracted from two standards surveys and eight specifically-designed questions to assess six evaluation domains related to benefits and safety experienced by consumers fitted with bone-anchored prosthesis provided by QALS.  Table 5. Breakdown of typical time commitment and costs associated with administration of the QALS's Continuous Quality Improvement (CQI) procedure for Evaluations A and B that repeated four times during the course of BAP treatment and Evaluation C that is conducted once and repeated yearly (Prosthetic Service Provider (PSP) labour = $160 per hour accordingly to the schedule of allowable fixed expenses in QALS' procedure, QALS' staff time = $30 per hour).