Possibility of predicting recurrence in the analysis of a database of patients with periprosthetic hip joint infection

Автор: Bozhokin M.S., Bozhkova S.A., Kochish A.A., Korneva Yu.S., Nikonorova M.L., Daloul F., Artyukh V.A.

Журнал: Вестник медицинского института "РЕАВИЗ": реабилитация, врач и здоровье @vestnik-reaviz

Рубрика: Информационно-вычислительные технологии в медицине

Статья в выпуске: 5 т.15, 2025 года.

Бесплатный доступ

Introduction. Musculoskeletal diseases and injuries are a pressing problem for millions of people. Damage to the hyaline cartilage leads to further degradation of the articular surface. Conservative treatment methods are ineffective, and at the final stage, a highly traumatic, costly procedure of joint arthroplasty is required. After it, in 3% of cases, complications occur in the form of periprosthetic infection, the treatment of which requires significant additional economic costs, reducing the quality of life and increasing the risk of disability of patients. Of particular interest is the prediction of recurrence of periprosthetic infection associated with the chronicity of the process and a significant increase in the duration of treatment and its costs. The aim of the study. Creation of a structured database for subsequent analysis of factors influencing the risk of recurrence of periprosthetic infection of the hip joint using the Python programming language based on archival data of patients with hip joint arthroplasty. Materials and methods. The work was carried out using information about patients with periprosthetic infection of the hip joint, who were treated at the Center in the period from 2010 to 2022. Discussion. The obtained data allow predicting the risk of recurrent PJI, as well as analyzing the causes leading to it, which will allow adjusting the further treatment regimen for such patients in order to avoid or minimize the development of recurrent periprosthetic infection. Results. A structured database of cleaned and prepared data for further analysis was obtained with 1611 unique patients, each of whom was described by 101 unique attributes. The possibility of predicting the risk of developing periprosthetic infection based on the information obtained using an automatic algorithm was shown. Conclusion: further analysis of the data bank will allow us to deepen our understanding of the causes of recurrent periprosthetic infection and consolidate the experience of traumatologists and orthopedists regarding the management of this cohort of patients.

Еще

Prosthesis-related infections [D016459], arthroplasty, replacement, hip [D019644], recurrence [D012008], forecasting [D011379], databases [D030541], machine learning [D000069550], risk factors [D012307], Python/software [D012984], reoperation [D012086], data mining [D057225]

Еще

Короткий адрес: https://sciup.org/143185350

IDR: 143185350   |   УДК: 616.728.2-089.844-06-022.7-037   |   DOI: 10.20340/vmi-rvz.2025.5.ITM.1