Over 8000 total hip arthroplasties (THA) in the UK were revised in 2019, half for aseptic loosening. It is believed that Artificial Intelligence (AI) could identify or predict failing THA and result in early recognition of poorly performing implants and reduce patient suffering.

The aim of this study is to investigate whether Artificial Intelligence based machine learning (ML) / Deep Learning (DL) techniques can train an algorithm to identify and/or predict failing uncemented THA.

Consent was sought from patients followed up in a single design, uncemented THA implant surveillance study (2010–2021). Oxford hip scores and radiographs were collected at yearly intervals. Radiographs were analysed by 3 observers for presence of markers of implant loosening/failure: periprosthetic lucency, cortical hypertrophy, and pedestal formation.

DL using the RGB ResNet 18 model, with images entered chronologically, was trained according to revision status and radiographic features. Data augmentation and cross validation were used to increase the available training data, reduce bias, and improve verification of results.

184 patients consented to inclusion. 6 (3.2%) patients were revised for aseptic loosening. 2097 radiographs were analysed: 21 (11.4%) patients had three radiographic features of failure.

166 patients were used for ML algorithm testing of 3 scenarios to detect those who were revised. 1) The use of revision as an end point was associated with increased variability in accuracy. The area under the curve (AUC) was 23–97%. 2) Using 2/3 radiographic features associated with failure was associated with improved results, AUC: 75–100%. 3) Using 3/3 radiographic features, had less variability, reduced AUC of 73%, but 5/6 patients who had been revised were identified (total 66 identified).

The best algorithm identified the greatest number of revised hips (5/6), predicting failure 2–8 years before revision, before all radiographic features were visible and before a significant fall in the Oxford Hip score. True-Positive: 0.77, False Positive: 0.29.

ML algorithms can identify failing THA before visible features on radiographs or before PROM scores deteriorate. This is an important finding that could identify failing THA early.

Aims

Total hip arthroplasty (THA) and total knee arthroplasty (TKA) are common orthopaedic procedures requiring postoperative radiographs to confirm implant positioning and identify complications. Artificial intelligence (AI)-based image analysis has the potential to automate this postoperative surveillance. The aim of this study was to prepare a scoping review to investigate how AI is being used in the analysis of radiographs following THA and TKA, and how accurate these tools are.

Abstract

Aims

Knee arthroplasty surgery is a highly effective treatment for arthritis and disorders of the knee. There are a wide variety of implant brands and types of knee arthroplasty available to surgeons. As a result of a number of highly publicized failures, arthroplasty surgery is highly regulated in the UK and many other countries through national registries, introduced to monitor implant performance, surgeons, and hospitals. With time, the options available within many brand portfolios have grown, with alternative tibial or femoral components, tibial insert materials, or shapes and patella resurfacings. In this study we have investigated the effect of the expansion of implant brand portfolios and where there may be a lack of transparency around a brand name. We also aimed to establish the potential numbers of compatible implant construct combinations.

This study utilized data from the NJR dataset on all Corail/Pinnacle total hip replacements (THR) to determine (a) the level of unit variation of the Corail/Pinnacle 36mm Metal On Metal THR within England and Wales; (b) patient, implant and surgeon factors that may be associated with higher revision rates; (c) Account for the influence of the MHRA announcement in 2010.

The national Revision Rate (RR) for the Corail / Pinnacle MOM THR was 10.77% (OR:1.46; CI:1.17–1.81). This was significantly greater than other articulation combinations (MOP 1.72%, COP 1.36%, COC 2.19%).

The 2010 MHRA announcement did not increase rate of revision (X²=1649.63, df=13, p<.001).

Patient factors associated with significantly increased revision rates included, female gender (OR 1.38 (CI 1.17–1.63, p<.001) and younger age OR 0.99 (CI 0.98–0.99), p<.001). Implant factor analysis demonstrated an inverse relationship between cup size and revision. As head length increased RR increased – highest risk of revision +12.5 (OR 1.69 (CI 1.12–2.55), p=0.13). Coxa vara, high offset stems had a higher risk of revision compared to standard offset stems (OR:1.41 (CI 1.15–1.74; p<.001). As stem size increased risk of revision decreased (OR 0.89 (CI 0.85–0.93); p<.001). Surgeon grade did not influence RR.

There was significant variation in RR between hospitals with 7 units (7/61 excluding low volume centres, <50 implants) identified as having significant higher rates of revision. However, for each of these units there was a greater proportion of higher risk patients (female, cup size 50–54, stem type).

This study has provided insight into unit variation, risk factors and the long term outcome of the Corail/Pinnacle 36mm MOMTHR. Future aims are to use these results to develop a risk stratified algorithm for the long term follow of these patients to minimize patient inconvenience and excess use of limited NHS resources.

Aims

Periprosthetic fracture (PF) after primary total hip arthroplasty (THA) is an uncommon but potentially devastating complication. This study aims to investigate the influence of cemented stem designs on the risk of needing a revision for a PF.

Introduction

Mix and Match (M&M) describes the use of components from more than one manufacturer in a total hip replacement (THR) The NJR has records of over 90,000 instances where this practice, which is contrary to the advice of most manufacturers and regulators, has been followed.

Background

Current analysis of unicondylar knee replacements (UKR) by national registries is based on the pooled results of medial and lateral implants. Using data from the National Joint Registry for England and Wales (NJR) we aimed to determine the proportion of lateral UKR implanted, their survival and reason for failure in comparison to medial UKR.

Current analysis of unicondylar knee replacements (UKRs) by national registries is based on the pooled results of medial and lateral implants. Consequently, little is known about the differential performance of medial and lateral replacements and the influence of each implant type within these pooled analyses. Using data from the National Joint Registry for England and Wales (NJR) we aimed to determine the proportion of UKRs implanted on the lateral side of the knee, and their survival and reason for failure compared with medial UKRs. By combining information on the side of operation with component details held on the NJR, we were able to determine implant laterality (medial versus lateral) for 32 847 of the 35 624 unicondylar replacements (92%) registered before December 2010. Of these, 2052 (6%) were inserted on the lateral side of the knee. The rates of survival at five years were 93.1% (95% confidence interval (CI) 92.7 to 93.5) for medial and 93.0% (95% CI 91.1 to 94.9) for lateral UKRs (p = 0.49). The rates of failure remained equivalent after adjusting for patient age, gender, American Society of Anesthesiologists (ASA) grade, indication for surgery and implant design using Cox’s proportional hazards method (hazard ratio for lateral relative to medial replacement = 0.88 (95% CI 0.69 to 1.13); p = 0.32). Aseptic loosening/lysis and unexplained pain were the main reasons for revision in both groups, although the reasons did vary depending on whether a mobile- or a fixed-bearing design was used. At a maximum of eight years the mid-term survival rates of medial and lateral UKRs are similar.

Summary

We report the first use of synchrotron xray spectroscopy to characterize and compare the chemical form and distribution of metals found in tissues surrounding patients with metal-on-metal hip replacements that failed with (Ultima hips) or without (current generation, large diameter hips) corrosion.

Objectives: We aimed to assess the incidence of requirement for shoe raises for a leg length discrepancy (LLD) after total hip replacement (THR). We also assessed the patient satisfaction with, and continued use of shoe raises for symptomatic LLD after THR.

Methods: We searched the orthotics records at our institution to identify all patients who had required a shoe raise for symptomatic LLD after primary unilateral total hip replacement between January 2003 and October 2008. 75 patients were identified. 72 were still alive. In the same period 4270 primary hip replacements were carried out at the institution. A questionnaire was sent out to all living patients. Patient details (including satisfaction) and operative details were recorded. Pre-operative and post-operative radiological measurements of leg length discrepancy (LLD) were performed.

Results: The incidence of requirement of a shoe raise for LLD after THR at our institution was 1.8%. 68% were women. 84% of questionnaires were returned. 31% had stopped using their shoe raise completely. Two-thirds of patients found the raise improved their symptoms of a LLD. Symptoms causing dissatisfaction with the shoe raise included new or worsening back pain, limp, uneven walking, self awareness, need to adjust trouser length, pain in other hip, discomfort while walking, and difficulty buying shoes. Patient overall satisfaction with their THR was poor in the patients who were dissatisfied with the shoe raise, but was good in those who found the raise useful.

Conclusion: About 2% of patients may require a shoe raise for symptomatic LLD after THR. Of these 65% will find the shoe raise helpful. Patient with a LLD after a THR that is not helped by a shoe raise are very dissatisfied. It is important that surgeons should take great care to avoid causing a LLD after THR as it can be a cause of very low patient satisfaction.

Our objective was to investigate how patients, general practitioners (GPs) and orthopaedic trainees, feel about the proposed governmental changes to reduce orthopaedic outpatient clinics by having GPs and specialist nurses follow-up post operative orthopaedics patients in the community. This was a cross-sectional questionnaire study of Teaching hospital and general practitioners in the Norfolk primary care trust.

Of 73 orthopaedic post-operative patients who attended outpatients over a one week period in July 2007, all responded. Of 250 General Practitioners, 239 responded. Of 38 orthopaedic trainees at the level of SHO (post MRCS) and SpR’s (Eastern deanery rotation and Pott rotation) 30 responded. Main outcome measures: opinions of GPs, Surgical trainees and Patients regarding proposed changes to outpatient orthopaedic clinics.

Of the patients 56 of 73 (77%) felt the surgeon was best suited to manage them post-operatively. Of these 47 felt that it was very important that the surgical team saw them post-operatively. Also 53 felt that their GP did not have sufficient knowledge and experience to adequately deal with their current orthopaedic problem. General Practitioners: Only 12 of 239 (5%) felt very confident assessing post-operative patients. Inadequate resources available to diagnose and treat post-operative complications was noted by 74% as the reason for not performing follow-up in primary care, and only 18% felt they should follow up post-operative patients. Trainees: All felt that following up their own post-operative patients was important to their training.

Most patients, GPs, and orthopaedic trainees had serious doubts about proposed governmental changes to reduce orthopaedic outpatient clinics by having GPs and specialist nurses follow-up post operative orthopaedics patients in the community.

Introduction: Leg length discrepancy (LLD) following total hip replacement (THR) is not uncommon. Some patients are symptomatic, with problems such as gait imbalance or back pain. LLD is a potential cause of litigation following THR.

We have observed that some patients perceive their LLD to be much greater than the true LLD. A large LLD is sometimes reported by therapists, despite only a small true LLD.

We have found that abduction tightness is a potent cause of apparent LLD, and report our investigations into this phenomenon.

Method: We have identified a series of patients with abductor tightness and a significant apparent LLD. The LLD becomes apparent when the operated leg is adducted to the midline (or when the patient stands with their ankles together). This causes the contralateral pelvis to elevate and the un-operated leg to “shorten”.

Clinical photographs and videos have been produced to demonstrate this phenomenon.

A 2-dimensional model has been made to demonstrate how the degree of abduction, offset and over-lengthening affect this phenomenon.

A computer model has been used to quantify these effects.

Results: An abduction contracture after THR will cause the un-operated leg to be apparently and functionally short, even in the absence of a true discrepancy.

Even with only minor abductor tightness, increasing the true length will disproportionately increase the apparent LLD.

Conclusion: An abduction contracture is a potent cause of apparent LLD. Even a small degree of true over lengthening will be greatly magnified by this phenomenon. We recommend careful clinical assessment for abductor tightness when examining patients complaining of a LLD after THR.

Introduction: Tendon lengthening is an important cause of morbidity after Achilles tendon rupture. However, direct measurement of the tendon length is difficult. Ankle dorsiflexion has therefore been used as a surrogate measure, on the assumption that it is the Achilles tendon that limits this movement. The aim of this investigation was to assess the relationship between Achilles tendon length and ankle dorsiflexion. The primary research question was whether or not the Achilles tendon is the structure that limits ankle dorsiflexion. The secondary purpose was to quantify the relationship between Achilles tendon lengthening and dorsiflexion at the ankle joint.

Methods: Five cadaveric specimens were dissected to expose the tendons and capsular tissue of the leg and hindfoot. Fixed bony reference points were used as markers for the measurements. In the first specimen, the Achilles tendon was intact and the other structures that may limit ankle dorsiflexion were sequentially divided. In the other specimens the Achilles tendon was lengthened by 1cm intervals and the effect upon ankle dorsiflexion movement was recorded.

Results: Division of the other tendons and the capsular tissue around the ankle joint did not affect the range of ankle dorsiflexion. When the Achilles was divided the foot could be dorsiflexed until the talar neck impinged upon the anterior aspect of the distal tibia. There was a mean increase of 12 degrees of dorsiflexion for each centimetre increase in tendon length.

Conclusion: The Achilles tendon is the anatomical structure that limits ankle dorsiflexion, even when the tendon is lengthened. There was a linear relationship between the length of the Achilles tendon and the range of ankle dorsiflexion in this cadaveric model. Ankle dorsiflexion would appear to be a clinically useful indicator of tendon length.

A computer system has been developed that allows rapid collection of data about joint replacements in the operating room which can then be transmitted to a server via the internet for later analysis.

The programme comprises three main fields, a demographic field, an implant field and a field to allow specific questions.

The specific features include:-

An unique programme linking the manufacturer’s barcodes in a readable form.

The system will be demonstrated in the course of the presentation. The perceived advantages over a paper system , validation, “windows anywhere” and confidentiality will be discussed.

The adaptability of the system for use in outpatients and for the production of operation notes will be outlined.

The cost, about £700, per unit plus software charges will be discussed.