Concerns surrounding application of artificial intelligence in hip and knee arthroplasty

Abstract

Literature surrounding artificial intelligence (AI)-related applications for hip and knee arthroplasty has proliferated. However, meaningful advances that fundamentally transform the practice and delivery of joint arthroplasty are yet to be realized, despite the broad range of applications as we continue to search for meaningful and appropriate use of AI. AI literature in hip and knee arthroplasty between 2018 and 2021 regarding image-based analyses, value-based care, remote patient monitoring, and augmented reality was reviewed. Concerns surrounding meaningful use and appropriate methodological approaches of AI in joint arthroplasty research are summarized. Of the 233 AI-related orthopaedics articles published, 178 (76%) constituted original research, while the rest consisted of editorials or reviews. A total of 52% of original AI-related research concerns hip and knee arthroplasty (n = 92), and a narrative review is described. Three studies were externally validated. Pitfalls surrounding present-day research include conflating vernacular (“AI/machine learning”), repackaging limited registry data, prematurely releasing internally validated prediction models, appraising model architecture instead of inputted data, withholding code, and evaluating studies using antiquated regression-based guidelines. While AI has been applied to a variety of hip and knee arthroplasty applications with limited clinical impact, the future remains promising if the question is meaningful, the methodology is rigorous and transparent, the data are rich, and the model is externally validated. Simple checkpoints for meaningful AI adoption include ensuring applications focus on: administrative support over clinical evaluation and management; necessity of the advanced model; and the novelty of the question being answered. Cite this article: Bone Joint J 2022;104-B(12):1292–1303.

Is ChatGPT a trusted source of information for total hip and knee arthroplasty patients?

“…Additional potential applications of LLMs like ChatGPT include clinical decision-making support, clinical documentation assistance, administrative work, and medical education. 40,41 To conclude, we found that ChatGPT answered common patient questions about THA and TKA with accuracy comparable to the average orthopaedic patient informational website, with adequate comprehensiveness, and with readability that can be improved upon request. However, improvements are needed before ChatGPT can be trusted to give safe, fully comprehensive information to patients.…”

“…Additional potential applications of LLMs like ChatGPT include clinical decision-making support, clinical documentation assistance, administrative work, and medical education. 40 , 41 …”

Artificial intelligence (AI) and large data registries: Understanding the advantages and limitations of contemporary data sets for use in AI research

“…Administrative registries represent heterogeneous combinations of data, resulting in prediction models that are more generalizable when compared to a model developed on data from a single surgeon (Table 2). Formulating a methodologically appropriate AI investigation is partially contingent on using a data set of high quality with sufficient features to answer the proposed scientific question [16,17]. The profound sample sizes of many large registries are optimal to allow for creation of appropriately sized training and testing cohorts necessary during AI analyses, as smaller sample sizes may predispose these methods to fatal flaws that preclude them from being clinically useful, such as model overfitting.…”

Predictive Tool Use and Willingness for Surgery in Patients With Knee Osteoarthritis

“…While this difference was not statistically significant, it necessitated adjustments in the analysis to account for this baseline variation. Two main justifications guided these adjustments 33 : (1) the magnitude of the baseline willingness difference was substantial, with an 8.8% advantage favoring the tool, and (2) the primary outcome of willingness for surgery at 6 months is directly linked to baseline willingness for surgery. Given that the purpose of a randomized study is to balance covariates (both measured and unmeasured), the adjustments were deemed appropriate considering the evident imbalance in the data.…”

“…Rapid advancements in artificial intelligence technologies have led to the development of numerous clinical predictive tools, including those for patients with knee osteoarthritis (OA) considering total knee arthroplasty (TKA). 1 , 2 , 3 , 4 , 5 While the integration of predictive tools into clinical practice is progressing quickly, few of these tools have undergone rigorous evaluation through randomized clinical trials (RCTs). 6 , 7 Consequently, the effectiveness of these tools in surgical decision-making for both patients and clinicians remains uncertain.…”