Advertisement for orthosearch.org.uk
Results 1 - 20 of 437
Results per page:
The Bone & Joint Journal
Vol. 106-B, Issue 12 | Pages 1408 - 1415
1 Dec 2024
Wall L Bunzli S Nelson E Hawke LJ Genie M Hinwood M Lang D Dowsey MM Clarke P Choong PF Balogh ZJ Lohmander LS Paolucci F

Aims

Surgeon and patient reluctance to participate are potential significant barriers to conducting placebo-controlled trials of orthopaedic surgery. Understanding the preferences of orthopaedic surgeons and patients regarding the design of randomized placebo-controlled trials (RCT-Ps) of knee procedures can help to identify what RCT-P features will lead to the greatest participation. This information could inform future trial designs and feasibility assessments.

Methods

This study used two discrete choice experiments (DCEs) to determine which features of RCT-Ps of knee procedures influence surgeon and patient participation. A mixed-methods approach informed the DCE development. The DCEs were analyzed with a baseline category multinomial logit model.


The Bone & Joint Journal
Vol. 106-B, Issue 12 | Pages 1426 - 1430
1 Dec 2024
Warne CN Ryan S Yu E Osmon DR Berry DJ Abdel MP

Aims

Cutibacterium acnes (C. acnes; previously known as Propionibacterium acnes or P. acnes) periprosthetic hip and knee infections are under-reported. While culture contamination with C. acnes occurs, true infections are important to recognize and treat. We sought to describe the demographics and treatment outcomes of patients with C. acnes periprosthetic joint infections (PJIs) of the hip and knee.

Methods

Patients with C. acnes PJI between January 2005 and December 2018 were retrospectively reviewed utilizing the institutional total joint registry. Patients with monomicrobial PJI and two or more positive cultures were considered to have true C. acnes PJI. Patients with polymicrobial infection or with only one positive culture were excluded. This resulted in 35 PJIs (21 hips and 14 knees); the patients’ mean age was 63 years (35 to 84) and 15 (43%) were female. Mean follow-up was five years (1 to 14).


The Bone & Joint Journal
Vol. 106-B, Issue 12 | Pages 1377 - 1384
1 Dec 2024
Fontalis A Yasen AT Giebaly DE Luo TD Magan A Haddad FS

Periprosthetic joint infection (PJI) represents a complex challenge in orthopaedic surgery associated with substantial morbidity and healthcare expenditures. The debridement, antibiotics, and implant retention (DAIR) protocol is a viable treatment, offering several advantages over exchange arthroplasty. With the evolution of treatment strategies, considerable efforts have been directed towards enhancing the efficacy of DAIR, including the development of a phased debridement protocol for acute PJI management. This article provides an in-depth analysis of DAIR, presenting the outcomes of single-stage, two-stage, and repeated DAIR procedures. It delves into the challenges faced, including patient heterogeneity, pathogen identification, variability in surgical techniques, and antibiotics selection. Moreover, critical factors that influence the decision-making process between single- and two-stage DAIR protocols are addressed, including team composition, timing of the intervention, antibiotic regimens, and both anatomical and implant-related considerations. By providing a comprehensive overview of DAIR protocols and their clinical implications, this annotation aims to elucidate the advancements, challenges, and potential future directions in the application of DAIR for PJI management. It is intended to equip clinicians with the insights required to effectively navigate the complexities of implementing DAIR strategies, thereby facilitating informed decision-making for optimizing patient outcomes.

Cite this article: Bone Joint J 2024;106-B(12):1377–1384.


The Bone & Joint Journal
Vol. 106-B, Issue 12 | Pages 1363 - 1368
1 Dec 2024
Chen DB Wood JA Griffiths-Jones W Bellemans J Haddad FS MacDessi SJ

As advancements in total knee arthroplasty progress at an exciting pace, two areas are of special interest, as they directly impact implant design and surgical decision making. Knee morphometry considers the three-dimensional shape of the articulating surfaces within the knee joint, and knee phenotyping provides the ability to categorize alignment into practical groupings that can be used in both clinical and research settings. This annotation discusses the details of these concepts, and the ways in which they are helping us better understand the individual subtleties of each patient’s knee.

Cite this article: Bone Joint J 2024;106-B(12):1363–1368.


The Bone & Joint Journal
Vol. 106-B, Issue 11 | Pages 1240 - 1248
1 Nov 2024
Smolle MA Keintzel M Staats K Böhler C Windhager R Koutp A Leithner A Donner S Reiner T Renkawitz T Sava M Hirschmann MT Sadoghi P

Aims. This multicentre retrospective observational study’s aims were to investigate whether there are differences in the occurrence of radiolucent lines (RLLs) following total knee arthroplasty (TKA) between the conventional Attune baseplate and its successor, the novel Attune S+, independent from other potentially influencing factors; and whether tibial baseplate design and presence of RLLs are associated with differing risk of revision. Methods. A total of 780 patients (39% male; median age 70.7 years (IQR 62.0 to 77.2)) underwent cemented TKA using the Attune Knee System) at five centres, and with the latest radiograph available for the evaluation of RLL at between six and 36 months from surgery. Univariate and multivariate logistic regression models were performed to assess associations between patient and implant-associated factors on the presence of tibial and femoral RLLs. Differences in revision risk depending on RLLs and tibial baseplate design were investigated with the log-rank test. Results. The conventional and novel Attune baseplates were used in 349 (45%) and 431 (55%) patients, respectively. At a median follow-up of 14 months (IQR 11 to 25), RLLs were present in 29% (n = 228/777) and 15% (n = 116/776) of the tibial and femoral components, respectively, and were more common in the conventional compared to the novel baseplate. The novel baseplate was independently associated with a lower incidence of tibial and femoral RLLs (both regardless of age, sex, BMI, and time to radiograph). One- and three-year revision risk was 1% (95% CI 0.4% to 1.9%)and 6% (95% CI 2.6% to 13.2%), respectively. There was no difference between baseplate design and the presence of RLLs on the the risk of revision at short-term follow-up. Conclusion. The overall incidence of RLLs, as well as the incidence of tibial and femoral RLLs, was lower with the novel compared to the conventional tibial Attune baseplate design, but higher than in the predecessor design and other commonly used TKA systems. Cite this article: Bone Joint J 2024;106-B(11):1240–1248


The Bone & Joint Journal
Vol. 106-B, Issue 11 | Pages 1231 - 1239
1 Nov 2024
Tzanetis P Fluit R de Souza K Robertson S Koopman B Verdonschot N

Aims

The surgical target for optimal implant positioning in robotic-assisted total knee arthroplasty remains the subject of ongoing discussion. One of the proposed targets is to recreate the knee’s functional behaviour as per its pre-diseased state. The aim of this study was to optimize implant positioning, starting from mechanical alignment (MA), toward restoring the pre-diseased status, including ligament strain and kinematic patterns, in a patient population.

Methods

We used an active appearance model-based approach to segment the preoperative CT of 21 osteoarthritic patients, which identified the osteophyte-free surfaces and estimated cartilage from the segmented bones; these geometries were used to construct patient-specific musculoskeletal models of the pre-diseased knee. Subsequently, implantations were simulated using the MA method, and a previously developed optimization technique was employed to find the optimal implant position that minimized the root mean square deviation between pre-diseased and postoperative ligament strains and kinematics.


The Bone & Joint Journal
Vol. 106-B, Issue 10 | Pages 1039 - 1043
1 Oct 2024
Luo TD Kayani B Magan A Haddad FS

The subject of noise in the operating theatre was recognized as early as 1972 and has been compared to noise levels on a busy highway. While noise-induced hearing loss in orthopaedic surgery specifically has been recognized as early as the 1990s, it remains poorly studied. As a result, there has been renewed focus in this occupational hazard. Noise level is typically measured in decibels (dB), whereas noise adjusted for human perception uses A-weighted sound levels and is expressed in dBA. Mean operating theatre noise levels range between 51 and 75 dBA, with peak levels between 80 and 119 dBA. The greatest sources of noise emanate from powered surgical instruments, which can exceed levels as high as 140 dBA. Newer technology, such as robotic-assisted systems, contribute a potential new source of noise. This article is a narrative review of the deleterious effects of prolonged noise exposure, including noise-induced hearing loss in the operating theatre team and the patient, intraoperative miscommunication, and increased cognitive load and stress, all of which impact the surgical team’s overall performance. Interventions to mitigate the effects of noise exposure include the use of quieter surgical equipment, the implementation of sound-absorbing personal protective equipment, or changes in communication protocols. Future research endeavours should use advanced research methods and embrace technological innovations to proactively mitigate the effects of operating theatre noise.

Cite this article: Bone Joint J 2024;106-B(10):1039–1043.


The Bone & Joint Journal
Vol. 106-B, Issue 10 | Pages 1059 - 1066
1 Oct 2024
Konishi T Hamai S Tsushima H Kawahara S Akasaki Y Yamate S Ayukawa S Nakashima Y

Aims. The Coronal Plane Alignment of the Knee (CPAK) classification has been developed to predict individual variations in inherent knee alignment. The impact of preoperative and postoperative CPAK classification phenotype on the postoperative clinical outcomes of total knee arthroplasty (TKA) remains elusive. This study aimed to examine the effect of postoperative CPAK classification phenotypes (I to IX), and their pre- to postoperative changes on patient-reported outcome measures (PROMs). Methods. A questionnaire was administered to 340 patients (422 knees) who underwent primary TKA for osteoarthritis (OA) between September 2013 and June 2019. A total of 231 patients (284 knees) responded. The ­Knee Society Score 2011 (KSS 2011), Knee injury and Osteoarthritis Outcome Score-12 (KOOS-12), and Forgotten Joint Score-12 (FJS-12) were used to assess clinical outcomes. Using preoperative and postoperative anteroposterior full-leg radiographs, the arithmetic hip-knee-ankle angle (aHKA) and joint line obliquity (JLO) were calculated and classified based on the CPAK classification. To investigate the impact on PROMs, multivariable regression analyses using stepwise selection were conducted, considering factors such as age at surgery, time since surgery, BMI, sex, implant use, postoperative aHKA classification, JLO classification, and changes in aHKA and JLO classifications from preoperative to postoperative. Results. The preoperative and postoperative CPAK classifications were predominantly phenotype I (155 knees; 55%) and phenotype V (73 knees; 26%), respectively. The change in the preoperative to postoperative aHKA classification was a significant negative predictive factor for KOOS-12 and FJS-12, while postoperative apex proximal JLO was a significant negative predictive factor for KSS 2011 and KOOS-12. Conclusion. In primary TKA for OA, preoperative and postoperative CPAK phenotypes were associated with PROMs. Alteration in varus/valgus alignment from preoperative to postoperative was recognized as a negative predictive factor for both KOOS-12 and FJS-12. Moreover, the postoperative apex proximal JLO was identified as a negative factor for KSS 2011 and KOOS-12. Determining the target alignment for each preoperative phenotype with reproducibility could improve PROMs. Cite this article: Bone Joint J 2024;106-B(10):1059–1066


The Bone & Joint Journal
Vol. 106-B, Issue 9 | Pages 916 - 923
1 Sep 2024
Fricka KB Wilson EJ Strait AV Ho H Hopper, Jr RH Hamilton WG Sershon RA

Aims

The optimal bearing surface design for medial unicompartmental knee arthroplasty (UKA) remains controversial. The aim of this study was to compare outcomes of fixed-bearing (FB) and mobile-bearing (MB) UKAs from a single high-volume institution.

Methods

Prospectively collected data were reviewed for all primary cemented medial UKAs performed by seven surgeons from January 2006 to December 2022. A total of 2,999 UKAs were identified, including 2,315 FB and 684 MB cases. The primary outcome measure was implant survival. Secondary outcomes included 90-day and cumulative complications, reoperations, component revisions, conversion arthroplasties, range of motion, and patient-reported outcome measures. Overall mean age at surgery was 65.7 years (32.9 to 94.3), 53.1% (1,593/2,999) of UKAs were implanted in female patients, and demographics between groups were similar (p > 0.05). The mean follow-up for all UKAs was 3.7 years (0.0 to 15.6).


The Bone & Joint Journal
Vol. 106-B, Issue 9 | Pages 935 - 941
1 Sep 2024
Ailaney N Guirguis PG Ginnetti JG Balkissoon R Myers TG Ramirez G Thirukumaran CP Ricciardi BF

Aims

The purpose of this study was to determine the association between prior sleeve gastrectomy in patients undergoing primary total hip and knee arthroplasty, and 90-day complications, incidence of revision arthroplasty, and patient-reported outcome scores at final follow-up.

Methods

This is a retrospective, single-centre analysis. Patients undergoing primary hip or knee arthroplasty with a prior sleeve gastrectomy were eligible for inclusion (n = 80 patients). A morbidly obese control group was established from the same institutional registry using a 1:2 match, for cases:controls with arthroplasty based on propensity score using age, sex, pre-sleeve gastrectomy BMI, Current Procedural Terminology code to identify anatomical location, and presurgical haemoglobin A1C. Outcomes included 90-day complications, incidence of revision arthroplasty, and patient-reported outcome scores at final follow-up. Multivariable logistic regressions evaluated associations of underlying preoperative demographic and treatment characteristics with outcomes.


The Bone & Joint Journal
Vol. 106-B, Issue 9 | Pages 892 - 897
1 Sep 2024
Mancino F Fontalis A Kayani B Magan A Plastow R Haddad FS

Advanced 3D imaging and CT-based navigation have emerged as valuable tools to use in total knee arthroplasty (TKA), for both preoperative planning and the intraoperative execution of different philosophies of alignment. Preoperative planning using CT-based 3D imaging enables more accurate prediction of the size of components, enhancing surgical workflow and optimizing the precision of the positioning of components. Surgeons can assess alignment, osteophytes, and arthritic changes better. These scans provide improved insights into the patellofemoral joint and facilitate tibial sizing and the evaluation of implant-bone contact area in cementless TKA. Preoperative CT imaging is also required for the development of patient-specific instrumentation cutting guides, aiming to reduce intraoperative blood loss and improve the surgical technique in complex cases. Intraoperative CT-based navigation and haptic guidance facilitates precise execution of the preoperative plan, aiming for optimal positioning of the components and accurate alignment, as determined by the surgeon’s philosophy. It also helps reduce iatrogenic injury to the periarticular soft-tissue structures with subsequent reduction in the local and systemic inflammatory response, enhancing early outcomes. Despite the increased costs and radiation exposure associated with CT-based navigation, these many benefits have facilitated the adoption of imaged based robotic surgery into routine practice. Further research on ultra-low-dose CT scans and exploration of the possible translation of the use of 3D imaging into improved clinical outcomes are required to justify its broader implementation. Cite this article: Bone Joint J 2024;106-B(9):892–897


Aims. The aim of this study was to compare the migration of the femoral component, five years postoperatively, between patients with a highly cross-linked polyethylene (HXLPE) insert and those with a conventional polyethylene (PE) insert in an uncemented Triathlon fixed insert cruciate-retaining total knee arthroplasty (TKA). Secondary aims included clinical outcomes and patient-reported outcome measures (PROMs). We have previously reported the migration and outcome of the tibial components in these patients. Methods. A double-blinded randomized controlled trial was conducted including 96 TKAs. The migration of the femoral component was measured with radiostereometry (RSA) at three and six months and one, two, and five years postoperatively. PROMs were collected preoperatively and at all periods of follow-up. Results. There was no clinically relevant difference in terms of migration of the femoral component or PROMs between the HXLPE and PE groups. The mean difference in migration (maximum total point motion), five years postopeatively, was 0.04 mm (95% CI -0.06 to 0.16) in favour of the PE group. Conclusion. There was no clinically relevant difference in migration of the femoral component, for up to five years between the two groups. These findings will help to establish a benchmark for future studies on migration of femoral components in TKA. Cite this article: Bone Joint J 2024;106-B(8):826–833


The Bone & Joint Journal
Vol. 106-B, Issue 8 | Pages 817 - 825
1 Aug 2024
Borukhov I Ismailidis P Esposito CI LiArno S Lyon J McEwen PJ

Aims. This study aimed to evaluate if total knee arthroplasty (TKA) femoral components aligned in either mechanical alignment (MA) or kinematic alignment (KA) are more biomimetic concerning trochlear sulcus orientation and restoration of trochlear height. Methods. Bone surfaces from 1,012 CT scans of non-arthritic femora were segmented using a modelling and analytics system. TKA femoral components (Triathlon; Stryker) were virtually implanted in both MA and KA. Trochlear sulcus orientation was assessed by measuring the distal trochlear sulcus angle (DTSA) in native femora and in KA and MA prosthetic femoral components. Trochlear anatomy restoration was evaluated by measuring the differences in medial, lateral, and sulcus trochlear height between native femora and KA and MA prosthetic femoral components. Results. Femoral components in both MA and KA alignments exhibited a more valgus DTSA compared to native femora. However, DTSA deviation from native was significantly less in KA than in MA (4.8° (SD 2.2°) vs 8.8° (SD 1.8°); p < 0.001). DTSA deviation from native orientation correlated positively with the mechanical lateral distal femoral angle (mLDFA) in KA and negatively in MA (r = 0.53, p < 0.001; r = -0.18, p < 0.001). Medial trochlear height was not restored with either MA or KA, with MA resulting in lower medial trochlear height than KA in the proximal 20% of the trochlea. Lateral and sulcus trochlear height was not restored with either alignment in the proximal 80% of the trochlea. At the terminal arc point, KA replicated sulcus and lateral trochlear height, while MA led to over-restoration. Conclusion. Femoral components aligned in KA demonstrated greater biomimetic qualities than those in MA regarding trochlear sulcus orientation and trochlear height restoration, particularly in valgus femora. Variability across knees was observed, warranting further research to evaluate the clinical implications of these findings. Cite this article: Bone Joint J 2024;106-B(8):817–825


The Bone & Joint Journal
Vol. 106-B, Issue 8 | Pages 760 - 763
1 Aug 2024
Mancino F Fontalis A Haddad FS


The Bone & Joint Journal
Vol. 106-B, Issue 8 | Pages 802 - 807
1 Aug 2024
Kennedy JW Sinnerton R Jeyakumar G Kane N Young D Meek RMD

Aims

The number of revision arthroplasties being performed in the elderly is expected to rise, including revision for infection. The primary aim of this study was to measure the treatment success rate for octogenarians undergoing revision total hip arthroplasty (THA) for periprosthetic joint infection (PJI) compared to a younger cohort. Secondary outcomes were complications and mortality.

Methods

Patients undergoing one- or two-stage revision of a primary THA for PJI between January 2008 and January 2021 were identified. Age, sex, BMI, American Society of Anesthesiologists grade, Charlson Comorbidity Index (CCI), McPherson systemic host grade, and causative organism were collated for all patients. PJI was classified as ‘confirmed’, ‘likely’, or ‘unlikely’ according to the 2021 European Bone and Joint Infection Society criteria. Primary outcomes were complications, reoperation, re-revision, and successful treatment of PJI. A total of 37 patients aged 80 years or older and 120 patients aged under 80 years were identified. The octogenarian group had a significantly lower BMI and significantly higher CCI and McPherson systemic host grades compared to the younger cohort.


The Bone & Joint Journal
Vol. 106-B, Issue 8 | Pages 808 - 816
1 Aug 2024
Hall AJ Cullinan R Alozie G Chopra S Greig L Clarke J Riches PE Walmsley P Ohly NE Holloway N

Aims. Total knee arthroplasty (TKA) with a highly congruent condylar-stabilized (CS) articulation may be advantageous due to increased stability versus cruciate-retaining (CR) designs, while mitigating the limitations of a posterior-stabilized construct. The aim was to assess ten-year implant survival and functional outcomes of a cemented single-radius TKA with a CS insert, performed without posterior cruciate ligament sacrifice. Methods. This retrospective cohort study included consecutive patients undergoing TKA at a specialist centre in the UK between November 2010 and December 2012. Data were collected using a bespoke electronic database and cross-referenced with national arthroplasty audit data, with variables including: preoperative characteristics, intraoperative factors, complications, and mortality status. Patient-reported outcome measures (PROMs) were collected by a specialist research team at ten years post-surgery. There were 536 TKAs, of which 308/536 (57.5%) were in female patients. The mean age was 69.0 years (95% CI 45.0 to 88.0), the mean BMI was 32.2 kg/m. 2. (95% CI 18.9 to 50.2), and 387/536 (72.2%) survived to ten years. There were four revisions (0.7%): two deep infections (requiring debridement and implant retention), one aseptic loosening, and one haemosiderosis. Results. Kaplan-Meier analysis demonstrated no difference in implant survival according to sex, age, or obesity status. Ten-year PROMs were available for 196/387 (50.6%) surviving patients and were excellent: mean Oxford Knee Score 34.4 (95% CI 32.7 to 36.1); mean Forgotten Joint Score (FJS) 51.2 (95% CI 16.1 to 86.3); mean EuroQol five-dimension five-level questionnaire score 69.9 (95% CI 46.8 to 93.0); 141/196 (71.9%) achieved the 22-point FJS patient-acceptable symptom state (PASS); and 156/196 (79.6%) were “very satisfied or satisfied”. Conclusion. This is the only large study reporting ten-year implant survival and functional outcomes of TKA using a cemented single-radius design and with a CS tibial bearing construct. The findings of excellent implant survival, safety, and functional outcomes indicate that this combination is a safe and effective option in routine TKA. Further investigation of this single-radius design TKA with CS tibial bearings with well-matched patient study groups will allow further insight into the performance of these implants. Cite this article: Bone Joint J 2024;106-B(8):808–816


The Bone & Joint Journal
Vol. 106-B, Issue 7 | Pages 669 - 679
1 Jul 2024
Schnetz M Maluki R Ewald L Klug A Hoffmann R Gramlich Y

Aims

In cases of severe periprosthetic joint infection (PJI) of the knee, salvage procedures such as knee arthrodesis (KA) or above-knee amputation (AKA) must be considered. As both treatments result in limitations in quality of life (QoL), we aimed to compare outcomes and factors influencing complication rates, mortality, and mobility.

Methods

Patients with PJI of the knee and subsequent KA or AKA between June 2011 and May 2021 were included. Demographic data, comorbidities, and patient history were analyzed. Functional outcomes and QoL were prospectively assessed in both groups with additional treatment-specific scores after AKA. Outcomes, complications, and mortality were evaluated.


The Bone & Joint Journal
Vol. 106-B, Issue 7 | Pages 680 - 687
1 Jul 2024
Mancino F Fontalis A Grandhi TSP Magan A Plastow R Kayani B Haddad FS

Aims. Robotic arm-assisted surgery offers accurate and reproducible guidance in component positioning and assessment of soft-tissue tensioning during knee arthroplasty, but the feasibility and early outcomes when using this technology for revision surgery remain unknown. The objective of this study was to compare the outcomes of robotic arm-assisted revision of unicompartmental knee arthroplasty (UKA) to total knee arthroplasty (TKA) versus primary robotic arm-assisted TKA at short-term follow-up. Methods. This prospective study included 16 patients undergoing robotic arm-assisted revision of UKA to TKA versus 35 matched patients receiving robotic arm-assisted primary TKA. In all study patients, the following data were recorded: operating time, polyethylene liner size, change in haemoglobin concentration (g/dl), length of inpatient stay, postoperative complications, and hip-knee-ankle (HKA) alignment. All procedures were performed using the principles of functional alignment. At most recent follow-up, range of motion (ROM), Forgotten Joint Score (FJS), and Oxford Knee Score (OKS) were collected. Mean follow-up time was 21 months (6 to 36). Results. There were no differences between the two treatment groups with regard to mean change in haemoglobin concentration (p = 0.477), length of stay (LOS, p = 0.172), mean polyethylene thickness (p = 0.065), or postoperative complication rates (p = 0.295). At the most recent follow-up, the primary robotic arm-assisted TKA group had a statistically significantly improved OKS compared with the revision UKA to TKA group (44.6 (SD 2.7) vs 42.3 (SD 2.5); p = 0.004) but there was no difference in the overall ROM (p = 0.056) or FJS between the two treatment groups (86.1 (SD 9.6) vs 84.1 (4.9); p = 0.439). Conclusion. Robotic arm-assisted revision of UKA to TKA was associated with comparable intraoperative blood loss, early postoperative rehabilitation, functional outcomes, and complications to primary robotic TKA at short-term follow-up. Robotic arm-assisted surgery offers a safe and reproducible technique for revising failed UKA to TKA. Cite this article: Bone Joint J 2024;106-B(7):680–687


The Bone & Joint Journal
Vol. 106-B, Issue 7 | Pages 642 - 645
1 Jul 2024
Harris IA Sidhu VS MacDessi SJ Solomon M Haddad FS


The Bone & Joint Journal
Vol. 106-B, Issue 7 | Pages 688 - 695
1 Jul 2024
Farrow L Zhong M Anderson L

Aims

To examine whether natural language processing (NLP) using a clinically based large language model (LLM) could be used to predict patient selection for total hip or total knee arthroplasty (THA/TKA) from routinely available free-text radiology reports.

Methods

Data pre-processing and analyses were conducted according to the Artificial intelligence to Revolutionize the patient Care pathway in Hip and knEe aRthroplastY (ARCHERY) project protocol. This included use of de-identified Scottish regional clinical data of patients referred for consideration of THA/TKA, held in a secure data environment designed for artificial intelligence (AI) inference. Only preoperative radiology reports were included. NLP algorithms were based on the freely available GatorTron model, a LLM trained on over 82 billion words of de-identified clinical text. Two inference tasks were performed: assessment after model-fine tuning (50 Epochs and three cycles of k-fold cross validation), and external validation.