Thoracolumbar injury classification systems are not used or researched extensively in paediatric population yet. This systematic review aims to explore the validity and reliability of the two main thoracolumbar injury classification systems in the paediatric population (age ≤ 18). It also aims to explore the transferability of adult classification systems to paediatrics. The Thoracolumbar Injury Classification System (TLICS) published in 2005 and the AO Spine published in 2013 were assessed in this paper because they both provide guidance for the assessment of the severity of an injury and recommend management strategies. A literature search was conducted on the following databases: Medline, EMBASE, Ovid during the period November 2020 to December 2020 for studies looking at the reliability and validity of the TLICS and AO Spine classification systems in paediatric population. Data on validity (to what extent TLICS/ AO Spine recommended treatment matched the actual treatment) and reliability (inter-rater and intra-rater reliability) was extracted. There is an “almost perfect validity” for TLICS. There is a “strong association” between the validity of TLICS and AO Spine. The intra-rater reliability is “moderate” for TLICS and “substantial” for AO Spine. The intra-rater reliability is “substantial” for TLICS and “almost perfect” for AO Spine. The six studies show a good overall validity and reliability for the application of TLICS and AO Spine in pediatric thoracolumbar fractures. However, implication of treatment and anatomical differences of the growing spine should be explored in detail. Therefore, AO Spine can be used in absence of any other classification system for paediatrics

Purpose of study and background. Clinical researchers use Pfirrmann classification for grading intervertebral disc degeneration radiologically. Basic researchers have access to morphology and instead use the Thompson score. The aim of this study was to assess the inter-observer reliability of both classifications, along with their correlation. Methods and Results. We obtained T2-weighted MR images of 80 human lumbar intervertebral discs with various stages of degeneration to assess the Pfirrmann-score. Then the discs were dissected midsagittally to obtain the Thompson-score. The observers were typical users of both grading systems: a spine surgeon, radiology resident, orthopaedic resident, and a basic scientist, all experts on intervertebral disc degeneration. Cohen's kappa (CK) was used to determine inter-observer reliability, and intra-class correlation (ICC) as a measure for the variation between the outcomes. For the Thompson score, the average CK was 0.366 and ICC score 0.873. The average inter-observer reliability for the Pfirrmann score was 0.214 (CK) and 0.790 (ICC). Comparing the grading systems, the intra-observer agreement was 0.240 (CK) and 0.685 (ICC). Conclusion. With substantial variation between observers, the inter-observer agreements for the Pfirrmann and Thompson grading systems were moderate. This may explain the poor relationship between radiological and clinical observations in patients and raises questions about the validity of the Pfirrmann score. The mediocre intra-observer agreement between the Pfirrmann and Thompson score shows that there is no clear definition of intervertebral disc degeneration. The field is in need for a new, objective and quantitative classification system to better define and evaluate disc degeneration. There are no conflicts of interest. Funded in part by Annafonds Netherlands and Dutch Spine Society

Purpose of the study. We propose a simple classification for adolescent idiopathic scoliosis (AIS) based on two components which include the curve type and shoulder level and suggest a treatment algorithm for AIS. Introduction. Few Classification systems for adolescent idiopathic scoliosis (AIS) have helped in communicating, understanding and selecting a treatment for this condition; however, most of these classifications are complex and include many subtypes, making it difficult for the orthopaedic surgeon to use them in clinical practice. The variable reliability and reproducibility of these studies make recommendations and comparisons between various operative treatments a difficult task. Furthermore, none of these classifications has taken the shoulder imbalance into account, despite its importance as a clinical parameter and outcome measure. Methods. We developed a classification system with two components: curve type (I through III) and shoulder level (A or B). The curve types are divided into type I: Primary lumbar-thoracolumbar +/− secondary dorsal; type II: Primary dorsal secondary lumbar and type III: Dorsal. Each curve pattern is subdivided into type A or B depending on the shoulder level. In type A, the lower shoulder is ipsilateral to the concavity of the primary curve. In type B, the shoulders are level or the lower shoulder is on the convexity of the primary curve. This classification was tested for interobserver reliability and intraobserver reproducibility by six surgeons using radiographs of 28 patients. We performed a retrospective analysis of the radiographs of 232 consecutive AIS cases to assess the prevalence of curve types and tested the surgical treatment against the proposed treatment algorithm. Results. Three major types and six subtypes were identified, of which type I accounted for 30%, type II 28% and type III 42%. The kappa coefficient for interobserver reliability was 0.943, while the kappa value for intraobserver reproducibility was 0.964. There was a complete concordance with the shoulder level component. Of the 232 cases reviewed, with a minimum two-year follow-up, only three patients developed a decompensation distal to the instrumentation requiring fusion extension. Conclusion. This classification is the first of its kind to specifically address shoulder imbalance in the surgical decision-making process. The high interobserver reliability and intraobserver reproducibility is due in part to the simplicity of this classification, which makes it an invaluable tool to describe scoliosis curves and offers a potential treatment algorithm in correcting scoliosis

Introduction. Morphological parameters are used to describe curve characters in AIS like curve location, curve magnitude, stiffness etc. Like all other morphological metrics the accuracy is more when digital imaging, archiving and extraction of features is used rather than manual measurements. The content Based Image Retrieval system is anew software that allows rapid, accurate documentation of AIS images and their retrieval by visual content. Classification systems and their shortcomings. Traditional classifications only looked at curve location (Ponsetti/Friedman); this was enhanced to add curve flexibility (to include or exclude secondary curves in fusion) (PUMC, King/Moe etc). Newer classifications like the Lenke have added sagittal profile into the decision making equation. From 5 basic curve types the subtypes have increased to 42 potential curve patterns by the addition of one parameter!! In future as we understand the 3-D geometry of these curves better we may want to add more measureable items (like degree of rotation) and by adding one term the subtypes would be 128!!! This suggests that we need to have a simple easy to remember way of classifying or eliminate classifications altogether. Experimental evidence. Several experiments were conducted with the new CBIR software which showed that similar images of scoliosis cases could be retrieved without resorting to a classification scheme. Even surgical planning can be made by downloading all similar cases operated before. The variability can be set to any level of precision desired. Significance. In future we may eliminate classifications to decide on curve types and for surgical planning and recall from a large multicentre database similar curves and their surgical plan

Background. Leg pain frequently accompanies low back pain and is associated with increased levels of disability and higher health costs than simple low back pain. Distinguishing between different types of low back- related leg pain (LBLP) is important for clinical management and research applications. The aim of this systematic review was to identify, describe and appraise papers that classify or subgroup populations with LBLP. Methods. The search strategy involved nine electronic databases including Medline and Embase, reference lists of eligible studies and relevant reviews. Selected papers were quality appraised independently by two reviewers using a standardised scoring tool. Results. Of 13,337 potential eligible citations, 49 relevant papers were identified that reported on 20 classification systems. Papers were grouped according to purpose and criteria of the classification systems. Five themes emerged: (i) pathoanatomical sources of pain (ii) clinical features, (iii) pain mechanisms, (iv) treatment based approach and (v) screening tools and prediction rules. Four of the twenty systems focused specifically on LBLP populations. Pain mechanisms and treatment based approach systems scored highest following quality appraisal as authors generally included statistical methods to develop their classifications and supporting work had been published on the systems' validity, reliability and generalizability. Conclusion. Numerous classification systems exist that include patients with leg pain, a minority of them focus specifically on distinguishing between different presentations of leg pain. Further work is needed to identify clinically meaningful subgroups of LBLP patients, ideally based on large primary care cohort populations and using stringent methods for classification system development. This abstract has not been previously published in whole or substantial part nor has it been presented previously at a national meeting. Conflicts of interest: No conflicts of interest. S Stynes is supported by an NIHR/HEE Clinical Doctoral Research Fellowship. Dr Konstantinou is supported by an HEFCE/NIHR Senior Clinical Lectureship. Professor Dunn is supported by the Wellcome Trust (083572)

Objective. The purpose of this study is to describe and validate a CT based classification of lumbosacral segment abnormalities. Method. 400 CT scans were retrospectively reviewed, a classification devised and incidence of abnormalities recorded. 5 types of abnormality were identified. Type 0 is normal; Type 1 describes an asymmetrical shortening of the iliolumbar ligament; Type 2's have the transverse process of L5 within 2 mm of the sacrum but not forming a joint; Type 3's have formed a diarthrodial joint, with 3A's showing no evidence of degeneration and 3B's displaying degenerative changes; In type 4's the transverse process and sacrum have fused; Type5's have involvement of L4. In order to validate the classification, 40 scans were selected with a full cross section of types. 4 independent observers classified each scan in 2 separate sessions, 2 weeks apart. Results. In the study population there was an abnormality in 54.5% of individuals. In order to validate the classification the intra-observer and inter-observer ratings were analysed. The kappa values for the intra-observer results were between 0.69 and 0.88, indicating substantial agreement (using the Landis and Koch kappa interpretation). The results for inter-observer ratings also gave a combined score of over 0.7 for both sessions, again indicating substantial agreement. Conclusion. A CT classification of lumbosacral segment abnormalities, which is both straight forward to use and repeatable, has been produced. The incidence of these abnormalities is higher in our population of CT scans compared to previous published series using plain radiographs

Aim:. The Lenke Classification for adolescent idiopathic scoliosis (AIS) classifies curves as nonstructural if they reduce to less than 25° on bending radiographs. We aimed to establish whether there is a significant difference in curves assessed as structural/ nonstructural when comparing bending radiographs to forced traction radiographs. Methods:. We undertook a retrospective database review of 100 consecutive AIS patients having undergone surgical correction by the 2 senior authors, together with radiographic review. Curves were classified according to the Lenke system including modifiers. Magnitude of the minor curves were compared on plain PA standing radiographs, bending radiographs and forced traction radiographs. Results:. There were 35 Lenke 1 curves. 24 Of the minor curves measured greater than 25° on plain PA radiographs. None of these remained greater than 25° on either bending or forced traction radiographs. There were 13 Lenke 2 curves, 11 of which had minor curves greater than 25°. None of these remained greater than 25° on either bending or forced traction radiographs. 14 Lenke 3 curves, 7 of which had minor proximal thoracic curves greater than 25°. 2 of these remained greater than 25° on bending whilst all reduced to less than 25° with traction. 8 Lenke 5 pattern curves, 5 of which had minor curves >25°. One curve remained >25° on both bending and traction radiographs. 8 Lenke 6 pattern curves none of which had a minor curve exceeding 25°. Conclusions:. Our results have shown no significant differences between curves classified as structural/ non-structural based on results obtained for bending radiographs versus forced traction radiographs for any of the Lenke curve patterns. Conflict Of Interest Statement: No conflict of interest

Aim:. To simplify sagittal plane spinal assessment by describing a single novel angle in the lumbar spine equivalent to the difference between pelvic incidence (PI) and lumbar lordosis (LL) and evaluate its reliability. Methods:. New sagittal modifiers in the classification of adult degenerative spinal deformity have been shown to be valid and reliable with the greatest variability being for pelvic incidence minus lumbar lordosis (PI-LL). This measurement can be simplified to a new angle (alpha) without the need to determine either PI or LL. This angle is between a line intersecting the bicoxofemoral centre and perpendicular to the L1 endplate (alpha line) and a line from the bicoxofemoral centre to the centre of the sacral endplate. Two readers graded 40 non-premarked cases twice each, approximately 1 week apart. Inter- and intra-rater variability and agreement were determined for PI-LL and alpha angle separately. Fleiss' kappa was used for reliability measures. Results:. Inter-rater kappa for PI-LL showed substantial reliability and alpha angle showed almost perfect agreement. Intra-rater kappa showed substantial reliability for PI-LL but almost perfect agreement for alpha angle across both readers. Discussion:. Alpha angle can be used as a surrogate for PI-LL and is easier and more reliable to measure. When PI=LL, alpha angle is 0, so the alpha line bisects the sacral plate. Conflict Of Interest Statement: No conflict of interest

Aims

There is a paucity of information on the pre-operative coronal imbalance in patients with degenerative lumbar scoliosis (DLS) and its influence on surgical outcomes.

Background

Low back-related leg pain (LBLP) is clinically diagnosed as referred leg pain or sciatica. Within the spectrum of LBLP there may be unrecognised subgroups of patients. This study aimed to identify and describe clusters of LBLP patients using latent class analysis (LCA).

Introduction. Children with early-onset scoliosis (EOS) with rib hump chest-wall distortion or fused/absent ribs have thoracic insufficiency syndrome (TIS). Commonly, respiration is adversely affected by loss of lung volume from chest-wall constriction and clinical loss of active rib cage expansion. The dynamic thoracic components of diaphragm or rib cage lung expansion during respiration is poorly characterised by radiograph or CT scan. Pulmonary function tests indicate only hemithorax performance. Dynamic lung MRI, however, can visualise both chest-wall and diaphragm motion, allowing assessment of each individual hemithorax performance, so that a dynamic classification system of the thoracic function can be developed. Methods. Ten patients with TIS underwent dynamic lung MRI testing as part of the routine clinical preoperative work-up. Each hemithorax was graded: 1=intact motion of both chest wall and diaphragm; 2=primarily loss of chest-wall motion with minimal diaphragm abnormality; 3=substantial loss of diaphragm excursion with minimal loss or compensatory hyperkinesis of chest wall; and 4=substantial loss of both diaphragm and chest-wall motion. The grades for each hemithorax were added and averaged to form the thoracic function score. Ranges of scores were grouped into levels of clinical thoracic performance: level I (score 1–1·5); level II (>1·5–2·5); level III (>2·5–3·5); and level IV(>3·5–4·0). Results. Of nine patients with EOS, two were level I, three were level II, and four were level III. In four patients there was marked posterior obstruction of diaphragmatic excursion by soft-tissue organs. One patient with hypoplastic thorax without scoliosis was level II. Conclusions. Thoracic function index is a new thoracic performance approach based on dynamic lung MRI that has potential to identify biomechanical abnormalities of the thorax in EOS that cause restrictive lung disease. This index could provide insight into how to reverse the abnormality with new types of surgeries. Posterior obstructive blockade of the diaphragm is identified as a new cause for restrictive lung disease in EOS

Introduction. Lumbosacral transitional vertebrae (LSTV) are increasingly recognized as a common anatomical variant and is the most common congenital anomaly of the lumbosacral spine. Patients can have symptomatic LSTV, known as Bertolotti's Syndrome, where transitional anatomy can cause back, L5 distribution leg, hip, and groin pain. We propose an outline for diagnosis and treatment of Bertolotti's Syndrome. Methods. We retrospectively reviewed over 500 patients presenting to the primary author with low back, buttock, hip, groin and/or leg pain from April 2009 through April 2024. Patients with radiographic findings of an LSTV and clinical presentation underwent diagnostic injections to confirm diagnosis of Bertolotti's syndrome. Treatment was determined based on patient's LSTV classification. 157 patients with confirmed Bertolotti's syndrome underwent surgical treatment. Results. Over 500 patients presented with an appropriate clinical presentation and radiographic findings of an LSTV. Diagnostic injections were targeted into the transitional anatomy confirming the LSTV as the primary pain generator to make the diagnosis of Bertolotti's syndrome. The decision in the type of surgical intervention, resection or fusion, was made based on patient's LSTV anatomy. 157 patients with confirmed Bertolotti's Syndrome underwent surgical treatment (121 fusions (77%), 36 resections (23%)). The classification system and surgical outcomes, in part, have been previously published in World Neurosurgery. Conclusion. We have outlined the best practice of diagnosis and treatment selection for Bertolotti's syndrome. We have shown significant improvement in outcomes based on this method. We hope to aid in both patient education and provide an outline on how clinicians can become knowledgeable on Bertolotti's syndrome. Conflicts of interest. No conflicts of interest. Sources of funding. No funding obtained

Aims. Vertebral body tethering (VBT) is a non-fusion technique to correct scoliosis. It allows correction of scoliosis through growth modulation (GM) by tethering the convex side to allow concave unrestricted growth similar to the hemiepiphysiodesis concept. The other modality is anterior scoliosis correction (ASC) where the tether is able to perform most of the correction immediately where limited growth is expected. Methods. We conducted a retrospective analysis of clinical and radiological data of 20 patients aged between 9 and 17 years old, (with a 19 female: 1 male ratio) between January 2014 to December 2016 with a mean five-year follow-up (4 to 7). Results. There were ten patients in each group with a total of 23 curves operated on. VBT-GM mean age was 12.5 years (9 to 14) with a mean Risser classification of 0.63 (0 to 2) and VBT-ASC was 14.9 years (13 to 17) with a mean Risser classification of 3.66 (3 to 5). Mean preoperative VBT-GM Cobb was 47.4° (40° to 58°) with a Fulcrum unbend of 17.4 (1° to 41°), compared to VBT-ASC 56.5° (40° to 79°) with 30.6 (2° to 69°)unbend. Postoperative VBT-GM was 20.3° and VBT-ASC Cobb angle was 11.2°. The early postoperative correction rate was 54.3% versus 81% whereas Fulcrum Bending Correction Index (FBCI) was 93.1% vs 146.6%. The last Cobb angle on radiograph at mean five years’ follow-up was 19.4° (VBT-GM) and 16.5° (VBT-ASC). Patients with open triradiate cartilage (TRC) had three over-corrections. Overall, 5% of patients required fusion. This one patient alone had a over-correction, a second-stage tether release, and final conversion to fusion. Conclusion. We show a high success rate (95%) in helping children avoid fusion at five years post-surgery. VBT is a safe technique for correction of scoliosis in the skeletally immature patient. This is the first report at five years that shows two methods of VBT can be employed depending on the skeletal maturity of the patient: GM and ASC. Cite this article: Bone Jt Open 2022;3(2):123–129

Aims. To determine the major risk factors for unplanned reoperations (UROs) following corrective surgery for adult spinal deformity (ASD) and their interactions, using machine learning-based prediction algorithms and game theory. Methods. Patients who underwent surgery for ASD, with a minimum of two-year follow-up, were retrospectively reviewed. In total, 210 patients were included and randomly allocated into training (70% of the sample size) and test (the remaining 30%) sets to develop the machine learning algorithm. Risk factors were included in the analysis, along with clinical characteristics and parameters acquired through diagnostic radiology. Results. Overall, 152 patients without and 58 with a history of surgical revision following surgery for ASD were observed; the mean age was 68.9 years (SD 8.7) and 66.9 years (SD 6.6), respectively. On implementing a random forest model, the classification of URO events resulted in a balanced accuracy of 86.8%. Among machine learning-extracted risk factors, URO, proximal junction failure (PJF), and postoperative distance from the posterosuperior corner of C7 and the vertical axis from the centroid of C2 (SVA) were significant upon Kaplan-Meier survival analysis. Conclusion. The major risk factors for URO following surgery for ASD, i.e. postoperative SVA and PJF, and their interactions were identified using a machine learning algorithm and game theory. Clinical benefits will depend on patient risk profiles. Cite this article: Bone Joint Res 2023;12(4):245–255

Aims. The aim of this study was to investigate whether including the stages of ulnar physeal closure in Sanders stage 7 aids in a more accurate assessment for brace weaning in patients with adolescent idiopathic scoliosis (AIS). Methods. This was a retrospective analysis of patients who were weaned from their brace and reviewed between June 2016 and December 2018. Patients who weaned from their brace at Risser stage ≥ 4, had static standing height and arm span for at least six months, and were ≥ two years post-menarche were included. Skeletal maturity at weaning was assessed using Sanders staging with stage 7 subclassified into 7a, in which all phalangeal physes are fused and only the distal radial physis is open, with narrowing of the medial physeal plate of the distal ulna, and 7b, in which fusion of > 50% of the medial growth plate of distal ulna exists, as well as the distal radius and ulna (DRU) classification, an established skeletal maturity index which assesses skeletal maturation using finer stages of the distal radial and ulnar physes, from open to complete fusion. The grade of maturity at the time of weaning and any progression of the curve were analyzed using Fisher’s exact test, with Cramer’s V, and Goodman and Kruskal’s tau. Results. We studied a total of 179 patients with AIS, of whom 149 (83.2%) were female. Their mean age was 14.8 years (SD 1.1) and the mean Cobb angle was 34.6° (SD 7.7°) at the time of weaning. The mean follow-up was 3.4 years (SD 1.8). At six months after weaning, the rates of progression of the curve for patients weaning at Sanders stage 7a and 7b were 11.4% and 0%, respectively for those with curves of < 40°. Similarly, the rates of progression of the curve for those being weaned at ulnar grade 7 and 8 using the DRU classification were 13.5% and 0%, respectively. The use of Sanders stages 6, 7a, 7b, and 8 for the assessment of maturity at the time of weaning were strongly and significantly associated (Cramer’s V 0.326; p = 0.016) with whether the curve progressed at six months after weaning. Weaning at Sanders stage 7 with subclassification allowed 10.6% reduction of error in predicting the progression of the curve. Conclusion. The use of Sanders stages 7a and 7b allows the accurate assessment of skeletal maturity for guiding brace weaning in patients with AIS. Weaning at Sanders stage 7b, or at ulnar grade 8 with the DRU classification, is more appropriate as the curve did not progress in any patient with a curve of < 40° immediately post-weaning. Thus, reaching full fusion in both distal radial and ulnar physes (as at Sanders stage 8) is not necessary and this allows weaning from a brace to be initiated about nine months earlier. Cite this article: Bone Joint J 2021;103-B(1):141–147

Introduction. Multiple studies have identified Cutibacterium acnes (C.acnes) and other microbes in intervertebral disc tissue using 16S DNA Sequencing and microbial cultures. However, it remains unclear whether these bacteria are native to the discs or result from perioperative contamination. Our study aimed to detect Gram-positive bacteria in non-herniated human disc samples and explore correlations with Toll-like receptors (TLR) 2, TLR4, NLRP3, and Gasdermin D. Methods. Immunohistochemical staining was conducted on 75 human IVD samples for Gram-positive bacteria, S. aureus, C.acnes, TLR2, TLR4, NLRP3, and Gasdermin D. Cell detection and classification were performed using QuPath. NP cells were treated with Lipopolysaccharide (LPS) and Peptidoglycan (PGN) in monolayer and alginate beads for up to 72 hours, followed by secretome analysis using Luminex. Statistical analysis included Kruskal-Wallis, Dunn's multiple comparison test, and Pearson correlation. Results. Immunohistochemical staining revealed Gram-positive bacteria exclusively within cells, with C. acnes positivity ranging from 5–99% and correlating with patient age (r=0.41, p= 0.007). TLR2 positivity ranged from 5–99% and TLR4 from 3–72%, showing a strong correlation (r= 0.62, p= 1.5e-006). Females with mid-degenerative grades exhibited significantly decreased TLR2 expression compared to those without degeneration signs. Treatment with LPS and PGN increased catabolic cyto- and chemokines associated with IVD degeneration. Conclusion. In conclusion, this study confirms Gram-positive bacteria presence in non-herniated human disc samples and highlights their role in triggering a catabolic response in disc cells. No conflicts of interest.  . Sources of funding. This project is part of the Disc4All Training network to advance integrated computational simulations in translational medicine, applies to intervertebral disc degeneration and funded by Horizon 2020 (H2020-MSCA-ITN-ETN-2020 GA: 955735)

Aims. With resumption of elective spine surgery services in the UK following the first wave of the COVID-19 pandemic, we conducted a multicentre British Association of Spine Surgeons (BASS) collaborative study to examine the complications and deaths due to COVID-19 at the recovery phase of the pandemic. The aim was to analyze the safety of elective spinal surgery during the pandemic. Methods. A prospective observational study was conducted from eight spinal centres for the first month of operating following restoration of elective spine surgery in each individual unit. Primary outcome measure was the 30-day postoperative COVID-19 infection rate. Secondary outcomes analyzed were the 30-day mortality rate, surgical adverse events, medical complications, and length of inpatient stay. Results. In all, 257 patients (128 males) with a median age of 54 years (2 to 88) formed the study cohort. The mean number of procedures performed from each unit was 32 (16 to 101), with 118 procedures (46%) done as category three prioritization level. The majority of patients (87%) were low-medium “risk stratification” category and the mean length of hospital stay was 5.2 days. None of the patients were diagnosed with COVID-19 infection, nor was there any mortality related to COVID-19 during the 30-day follow-up period, with 25 patients (10%) having been tested for symptoms. Overall, 32 patients (12%) developed a total of 34 complications, with the majority (19/34) being grade 1 to 2 Clavien-Dindo classification of surgical complications. No patient required postoperative care in an intensive care setting for any unexpected complication. Conclusion. This study shows that safe and effective planned spinal surgical services can be restored avoiding viral transmission, with diligent adherence to national guidelines and COVID-19-secure pathways tailored according to the resources of the individual spinal units. Cite this article: Bone Jt Open 2021;2(12):1096–1101

Aims

Patients with cauda equina syndrome (CES) require emergency imaging and surgical decompression. The severity and type of symptoms may influence the timing of imaging and surgery, and help predict the patient’s prognosis. Categories of CES attempt to group patients for management and prognostication purposes. We aimed in this study to assess the inter-rater reliability of dividing patients with CES into categories to assess whether they can be reliably applied in clinical practice and in research.

Aims

To determine whether side-bending films in scoliosis are assessed for adequacy in clinical practice; and to introduce a novel method for doing so.

Purposes and Background. Musculoskeletal disorders including as back and neck pain are leading causes of work disability. Effective interventions exist (i.e. functional restoration, multidisciplinary biopsychosocial rehabilitation, workplace-based interventions, etc.), but it is difficult to select the optimal intervention for specific patients. The Work Assessment Triage Tool (WATT) is a clinical decision support tool developed using machine learning to help select interventions. The WATT algorithm categorizes patients based on individual, occupational, and clinical characteristics according to likelihood of successful return-to-work following rehabilitation. Internal validation showed acceptable classification accuracy, but WATT has not been tested beyond the original development sample. Our purpose was to externally validate the WATT. Methods and Results. A population-based cohort design was used, with administrative and clinical data extracted from a Canadian provincial compensation database. Data were available on workers being considered for rehabilitation between January 2013 and December 2016. Data was obtained on patient characteristics (ie. age, sex, education level), clinical factors (ie. diagnosis, part of body affected, pain and disability ratings), occupational factors (ie. occupation, employment status, modified work availability), type of rehabilitation program undertaken, and return-to-work outcomes (receipt of wage replacement benefits 30 days after assessment). Analysis included classification accuracy statistics of WATT recommendations for selecting interventions that lead to successful RTW outcomes. The sample included 5296 workers of which 33% had spinal conditions. Sensitivity of the WATT was 0.35 while specificity was 0.83. Overall accuracy was 73%. Conclusion. Accuracy of the WATT for selecting successful rehabilitation programs was modest. Algorithm revision and further validation is needed. No conflicts of interest. Sources of funding: Funding was provided by the Workers' Compensation Board of Alberta