Development of a Conservative Care and Bracing Registry (CCBR)

Idiopathic scoliosis (IS) is the most common form of scoliosis, defined as a lateral curvature of the spine of at least 10°. When left untreated in growing children, curves can progress leading to back pain, cardiopulmonary dysfunction, and substantial deformity affecting both physical and psychosocial functioning. IS is common, affecting approximately 3% of the population; nearly 10% of those patients require treatment and 0.1% meet surgical indications. Treatment options, including observation, bracing, physiotherapy, and surgery, are patient specific depending on risk factors for progression including skeletal maturity, curve magnitude, and family history. Surgery is typically recommended for patients whose curves reach 50° or greater. The goal of surgical correction is to permanently improve the alignment and balance of the trunk. The most common procedure is a posterior spinal fusion where rigid rods are used to successfully achieve these outcomes, but at the cost of loss of flexibility in the operated segment. Unfortunately, surgery exposes patients to risk through complications. Studies have reported a surgical complication rate of 0%-15.4%, with reoperation rates reported between 12.9% and 47.5%. Complications include infections, wound healing problems, neurologic injury, skin breakdown, implant failure and poor cosmesis creating physical and mental health burdens for patients and their caregivers. Besides the negative impact of complications due to surgery, surgical correction of IS is quite expensive. Vigneswaran et al reported that hospital charges for IS surgery more than tripled from approximately $55,000 in 1997 to over $177,000 in 2012. These numbers do not include the consequences of complications and re-operations which add to these costs and increase the burdens on patients, their families, and the healthcare system as a whole. The goal of non-operative therapy for IS, bracing and physiotherapy, is to prevent or limit curve progression and obviate the need for surgery, thus avoiding these costs and potential complications. This is the critical goal of managing IS, and bracing, the mainstay of conservative management in IS, has been shown in many studies to reduce the risk of curve progression and need for surgery over natural history. The 2013 BrAIST study, a randomized clinical trial, has provided the most compelling evidence in favor of bracing to date, and this landmark study dramatically increased the interest and utilization of bracing globally. Yet, over a decade later, strong, evidence-based guidelines for brace treatment are still missing. Yes, the current literature consistently demonstrates that improved brace outcomes are seen in older patients with smaller curves and better brace compliance, but when one tries to distill specific criteria from these general statements (e.g. how small a curve to brace, or how many hours in the brace is enough), the literature is tremendously variable; details are not well understood nor well agreed upon. Variability is seen throughout the brace literature with little consistency in indications for brace treatment, brace types, brace prescriptions, use of monitors, timing of radiographs, goals of bracing, and evaluation of skeletal maturity. A recent best practice guideline on bracing in AIS, published in 2020 by Roye et al., highlighted this challenge by demonstrating wide variability in such basic factors as initiating brace care, brace prescriptions, brace types, use of compliance monitoring, and how and when to discontinue treatment. These findings were replicated in a recent SRS Member survey that demonstrated variability in all aspects of brace management from initiation to discontinuation, evaluation of brace quality and objective brace monitoring. The many unknowns and the many gaps in the current literature demonstrate a clear and urgent need for a large bracing registry to evaluate bracing regimens and ultimately optimize outcomes. With regards to previous work in this area,there is any multi-center registries studying brace treatment of IS. A 2015 systematic review exploring the impact of spine registry studies found no registries collecting data on conservative care (all were surgical). The authors concluded there was a lack of evidence for the role of existing registries in improving quality of care, and recommended building more specific registries based on well-defined research questions. Real world data coming from the registry design for the proposed brace registry will be a source of data providing prospective evidence and scientific support to guidelines in brace management. In summary, IS is a common condition affecting a significant portion of the population. Surgical IS is costly and places patients at risk of significant adverse outcomes, including infection and spinal cord injury. Bracing is an established and effective treatment that has been shown to clearly reduce the risk of curve progression and the risk of surgery, but significant questions remain about optimal indications, treatment regimens, and brace cessation. Despite the many thousands of patients undergoing brace treatment for IS today, there is tremendous variability in bracing regimens, and there is no large, multi-institutional registry monitoring this population to help answer these critical questions. There is a clear and urgent need for support for such a registry, and the framework for developing and initiating such a registry for brace treatment in IS is proposed below. A brace registry is the ideal way to answer research questions related to the effect of bracing on the natural history of scoliosis. There are many factors that influence the effect of bracing on patients with scoliosis. For example, brace compliance has been shown to be critically important in successful brace outcomes, and compliance itself is impacted by wide ranging factors across multiple domains including: psychosocial factors (e.g. personality and family support), patient related factors (e.g. age), brace related factors (e.g. brace design, brace material, brace prescription), and curve related factors (e.g. curve magnitude and stiffness). The longitudinal collection of data, from patients managed by different medical doctors will provide guidance to clinical decision making through evidence-based treatment and diagnostic protocols. The goal of this research is to retrospectively gather comprehensive, longitudinal data on braced idiopathic scoliosis patients to evaluate outcomes and their determinants, inform clinical decision-making, increase standardization of treatment, and ultimately improve care. This goal will be met through the following process: Database design and construction: data variables to be collected will be defined and organized in a hierarchical structure accounting for longitudinal data. A multidisciplinary team including expert clinicians, skilled researchers and expert orthotist will collaborate to design the database registry. The database will be set up for both retrospective and prospective data entry. Data collection time points: baseline; brace follow up visits until brace discharge at skeletal maturity, at minimum annually. The follow up visits will be distinguished between brace check and adjustment and brace renewal and delivery. Post bracing follow-up at 12 and 24 months after brace discharge. Retrospective data input: all patient data collected in pdf files, will be checked and input in the database, creating a retrospective cohort for the preliminary analysis. To optimize the process, data input will start from the most recently discharged patients and move backward in time. This will allow the registry to include patients with the most recent clinical management (e.g. more advanced braces, standard use of sensors to measure compliance, and modern radiographic protocols). Retrospective data analysis: description and missing data check will guide variable refinement and variables will be classified according to their importance. Logistic and linear regression models will be utilized for treatment outcomes analysis. There will be comparisons across patients with a growth modulation with those with stabilization of the curves and those with progression over time until surgery. Different radiographic protocols used will be compared to check the differences in the Cobb angle variation over time. The registry under development is going to answer multiple research questions. Data collected in the retrospective phase of the study will be used to explore a variety of factors impacting brace outcomes. For example, there is growing evidence that brace treatment can lead to growth modulation and improvement in curve size over time. The study is aimed to identify factors associated with growth modulation including curve magnitude and location, skeletal maturity, brace type, brace exposure (time in brace), and orthotist experience (years in practice). Other research priorities will include creating protocols to optimize timing/frequency of radiographs, lending to best practice guideline development for orthotist's patient outcomes assessments. Additionally, brace discontinuation is a completely unexplored topic; there is no agreement on what criteria to use when deciding to stop bracing, how, or even if, to wean the brace over time, and how to monitor patients after bracing has been discontinued. The ultimate goal is to create standardized protocols to maximize successful radiographic outcomes while limiting costs and preserving quality of life outcomes for our patients and their families. Material and method The current project is describing the retrospective phase, which is functioning as a pilot of the prospective one. The retrospective phase will also include an ambispective cohort, meaning the patients included in the retrospective data collection effort will be consented for prospective longitudinal follow-up visits. Database development Feasibility is one of the major goals to be achieved. Real world data has become more and more relevant, but greater attention to data quality is an essential requirement. Quality comes from a careful selection of the variables to be collected by balancing the amount of information needed to provide research evidence filling a gap of knowledge, and ensuring complete data capture and standardized data collection methodology. To minimize missing data, it is important to have clinicians involved in data input, avoid overloading data entry with not useful information and provide an easy-to-use registry platform. The current team has extensive expertise in this field and will implement a stepwise procedure for this registry development. This step-by-step development procedure will allow a gradual and progressive optimization of the final version of the platform informed by the retrospective data from a unique, high-volume center and subsequently from the prospective collection of patient data from select, pilot clinics following homogeneous protocols and procedures of treatment. Finally, the major criteria for centers to participate in a multicenter longitudinal registry study will be defined. Retrospective phase This registry will be developed within an established registry platform with over 25 years of experience in multicenter research and database development. This registry platform collaboration will decrease costs and will streamline feasibility and implementation. This HIPPA compliant platform will support ethical and legal considerations, and data quality assurance strategies. The retrospective data is composed of clinical and radiographic data to be gathered from patients cared for by a single orthotics group over the past decade. Selection of data variables Data variables to be collected will be defined and organized in a hierarchical structure accounting for longitudinal data. The choice of variables is first based on data availability. The list of data variables to be included in the retrospective phase has been discussed in a multidisciplinary team including expert clinicians, skilled researchers and expert orthotist. This team is collaborating to design the database registry. Data collection timepoints to include: baseline; brace follow up visits (annually, at a minimum) until brace discharge at skeletal maturity. The follow up visits will be distinguished between brace check and adjustment and brace renewal and delivery. Post bracing follow-up at 12 and 24 months after brace discharge. The minimum follow-up duration will be three months. Retrospective data input: all patient data collected in pdf files, will be QA checked (for illogical and erroneous data) and input in the database, creating a retrospective cohort for the preliminary analysis. To optimize the process, data input will start from the most recently discharged patients and move backward in time. This will allow the registry to include patients with the most recent clinical management (e.g. more advanced braces, standard use of sensors to measure compliance, and modern radiographic protocols). The selection of patients to be included will be based on the above pre-specified inclusion criteria. Exceptions and changes in the criteria could be made. Retrospective data analysis: descriptive statistics will be performed as interim analyses after data entry of blocks of 50 patients. At the end of retrospective data entry, outcome variables will be classified according to their importance. The team will discuss the missing information and will refine the list of essential data variables compared to "nice to have" information. This vital process will aid in the development of the prospective registry; ensuring the practical feasibility of data entry done by clinicians and balancing the minimum time for data entry with the minimum and ideal amount of information needed to produce relevant research. Retrospective cohort sample size: A cohort of 1445 AIS patients who have completed 2 years of post-bracing follow up will comprise the retrospective registry. This retrospective cohort is established from multiple surgeon referrals to one orthotist practice. Participants Inclusion CRITERIA:all subjects affected by scoliosis, treated with brace and followed up until 2 years after brace discharge. From the Retrospective cohort, there are 1900 patients who are still in treatment, who will be included (and consented into the ambispective cohort) to be followed until treatment completion. Patients will be included in the retrospective cohort according to the following inclusion criteria Inclusion criteria ● Scoliosis diagnosis, idiopathic and secondary forms ● Skeletally immature subjects at baseline(Risser \<=3 and Sanders \<=6) ● Brace with compliance monitor ● Signed informed consent (for the ambispective follow-up data collection, if patients have not met the 2 year follow-up timepoint) Exclusion criteria ● Patients skeletally mature at baseline (Risser \>3 and Sanders\>6) ● Patients with other spine deformities like hyperkyphosis, Scheuermann deformity or spondylolisthesis unless associated with scoliosis. ● Those not consenting to participate (for the ambispective follow-up data collection cohort) Design The registry data will allow answers to different questions with different study designs: cohort studies and case-control studies mainly. The longitudinal structure of data will allow studies controlling for differences between and within subjects. Outcomes The registry database will provide multiple outcome measures to be considered depending on the research questions to be answered. The following table shows the outcome measures framework. The selection of patient reported outcomes will be done after a literature review to define the ideal questionnaires to be included, to allow comparisons with previous published studies, to explore new aspects and to minimize missing data and optimize data collection. Data variables (Figure 1 shows data variables over time and measurement occasions) Historical data: ● Age ● Gender ● Age at menarche ● Previous treatment ● Referred brace wearing time Clinical data: ● Asymmetries at the shoulder and waist level ● Trunk unbalance ● Hump measure with the angle of trunk rotation ● Height and weight Radiographic data ● in or out of brace x-ray, time from brace removal will be collected ● curve types ● Rigo classification ● Cobb angle ● Curve apex and side of the curves ● Pelvic measures ● Thoracic Kyphosis ● Lumbar Lordosis ● Bone maturity score: Risser and Sanders Treatment protocols will be compared to check if this is influencing the primary outcome: ● Part time prescription vs full time prescription ● Night time prescription vs full time prescription ● Night time prescription vs part time prescription ● Orthotist and medical doctors' names ● compliance monitor data and referred time of brace wear when unavailable monitors. ● association with other treatment, including different brace design. ● Level of physical activity ● Scoliosis specific exercises, type and modalities Radiographic protocol will be compared to check if they are influencing the primary outcome: number and timing of in-brace and out of brace x-rays. Procedures and intervention Prospective Recruitment and retention Successful registries depend on a sustainable workflow model that can be integrated into the day-to-day clinical practice of active health care professionals, with minimal disruption. In the development of this registry, having the orthotist facility as principal recruiters of the patients' data can offer a minimization of time disruption for MDs and a maximization in the motivation in providing data, thanks to the biggest interest in exploring brace effects. Furthermore, the retrospective phase is working as a pilot test, providing good strategies to avoid time disruption and increasing motivation of inputters and patients' retention and minimizing missing data. Another advantage of recruitment at the orthotist facility is the collection of homogenous data on brace characteristics, fabrication and orthotists expertise. On the other hand, some heterogeneity in treatment and diagnostic protocols is expected to provide evidence and scientific support to guidelines in brace management and radiographic protocols. This is an innovative strategy compared to what has been published in the field. The retrospective analysis will also provide information on missing data to develop strategies to minimize missing data, by prioritizing between 'nice to have' information and must have type of variables. Data obtaining and quality assurance In the retrospective phase data will be entered from the case report paper forms completed in the past. In the prospective phase data will be input as an electronic case report form developed specifically for the purposes of this registry, in a longitudinal format, and accessible through a website platform. Data can be entered directly during everyday practice by the health care professional or by previously educated administration staff. Quality and time-effectiveness will be evaluated at the end of this project to help decision making of the multicenter phase. Data cleaning will be manual and automatic. Periodic generation of query reports related to the quality of the data received will be done by the registry coordinating center. Data tracking will also guide quality assurance procedures to be implemented in the future development of the registry. This registry will be developed within a well-established registry platform with over 25 years of experience in multicenter research and database development. This custom multicenter database build will use SQL Server and will be hosted by Amazon (AWS), which is backed up nightly using native SQL Server Command/schedule jobs. The backup is retained to AWS S3 for 30 days. Management services of this database platform are contracted through Cloud Advisors and includes: Amazon Web Services Design/Environmental Support, Monitoring \& Alerts and Security. This registry platform collaboration will obviate many issues, like related costs, feasibility and implementation issues, but also ethical and legal consideration, and last but not least quality assurance strategies. Intervention and description of the recruitment center The retrospective data comes from prospectively collected chart and radiographic data from patients cared for by a single orthotics group over the past decade. This group, run by one of the co-investigators (LS) with over 25 years' experience, is widely recognized as a leading producer of spinal orthoses with outstanding quality and a history of academic/ research involvement. This is evidenced by their significant footprint; they have a large regional presence that receives referrals from many spine surgeons not just from around the region, but from around the country and beyond. For the subsequent prospective component of this registry, there will be multiple recruiting centers. The orthotist group mentioned above will be a major recruitment center and they are well situated to do this given their centralized leadership and history of being involved in brace research. Statistical analysis: Benchmarking and auditing to check data quality will be performed regularly. After demographic description, any planned statistical analysis will be refined according to the specific research questions asked of the registry. TIMELINE Year 1: Construction of retrospective database, obtain IRB approval (retrospective \& ambispective study), enter all retrospective data, analyze data, define prospective database requirements, submit manuscript for publication. Year 2: Refine prospective database \& QI Registry construction, refine prospective, multicenter protocol and case report forms, obtain IRB approval (prospective study), engage participating research sites, and launch prospective, multicenter study and QI Registry. Figure 2 is showing the timeline for the development of the registry and its refinement. (NOTE: the current project is referring to year 1. Next steps This project involves a retrospective pilot registry which will serve to implement an optimized multicenter registry for prospective collection of brace treatment data. The findings and the expertise coming from the professionals involved in this project can advise future collaboration with other established registries, to enlarge the body of conservative treatment data.