Abstract
Introduction
Periprosthetic joint infection (PJI) is a serious complication after total joint arthroplasty. It is important to accurately identify PJI and monitor postoperative blood biochemical marker changes for the appropriate treatment strategy. In this study, we aimed to monitor the postoperative blood biochemical characteristics of PJI by contrasting with non-PJI joint replacement cases to understand how the characteristics change postoperatively.
Materials and methods
A total of 144 cases (52 of PJI and 92 of non-PJI) were reviewed retrospectively and split into development and validation cohorts. After exclusion of 11 cases, a total of 133 (PJI: 50, non-PJI: 83) cases were enrolled finally. An RF classifier was developed to discriminate between PJI and non-PJI cases based on 18 preoperative blood biochemical tests. We evaluated the similarity/dissimilarity between cases based on the RF model and embedded the cases in a two-dimensional space by Uniform Manifold Approximation and Projection (UMAP). The RF model developed based on preoperative data was also applied to the same 18 blood biochemical tests at 3, 6, and 12 months after surgery to analyze postoperative pathological changes in PJI and non-PJI. A Markov chain model was applied to calculate the transition probabilities between the two clusters after surgery.
Results
PJI and non-PJI were discriminated with the RF classifier with the area under the receiver operating characteristic curve of 0.778. C-reactive protein, total protein, and blood urea nitrogen were identified as the important factors that discriminates between PJI and non-PJI patients. Two clusters corresponding to the high- and low-risk populations of PJI were identified in the UMAP embedding. The high-risk cluster, which included a high proportion of PJI patients, was characterized by higher CRP and lower hemoglobin. The frequency of postoperative recurrence to the high-risk cluster was higher in PJI than in non-PJI.
Conclusions
Although there was overlap between PJI and non-PJI, we were able to identify subgroups of PJI in the UMAP embedding. The machine-learning-based analytical approach is promising in consecutive monitoring of diseases such as PJI with a low incidence and long-term course.
Similar content being viewed by others
Code availability
All R and Python codes used in this study are included the GitHub repository (https://github.com/eiryo-kawakami/PJI_2022_code).
Data availability
The data that support the findings of this study are available from the corresponding author, [N.K.], upon reasonable request.
References
Kapadia BH, Berg RA, Daley JA, Fritz J, Bhave A, Mont MA (2016) Periprosthetic joint infection. Lancet 387:386–394. https://doi.org/10.1016/S0140-6736(14)61798-0
Dale H, Hallan G, Hallan G, Espehaug B, Havelin LI, Engesaeter LB (2009) Increasing risk of revision due to deep infection after hip arthroplasty. Acta Orthop 80:639–645. https://doi.org/10.3109/17453670903506658
Parvizi J, Pawasarat IM, Azzam KA, Joshi A, Hansen EN, Bozic KJ (2010) Periprosthetic joint infection: the economic impact of methicillin-resistant infections. J Arthroplasty 25:103–107. https://doi.org/10.1016/j.arth.2010.04.011
Everhart JS, Altneu E, Calhoun JH (2013) Medical comorbidities are independent preoperative risk factors for surgical infection after total joint arthroplasty. Clin Orthop Relat Res 471:3112–3119. https://doi.org/10.1007/s11999-013-2923-9
Berbari EF, Osmon DR, Lahr B, Eckel-Passow JE, Tsaras G, Hanssen AD et al (2012) The Mayo prosthetic joint infection risk score: implication for surgical site infection reporting and risk stratification. Infect Control Hosp Epidemiol 33:774–781. https://doi.org/10.1086/666641
Parvizi J, Tan TL, Goswami K, Higuera C, Della Valle C, Chen AF et al (2018) The 2018 definition of periprosthetic hip and knee infection: an evidence-based and validated criteria. J Arthroplasty 33:1309–14.e2. https://doi.org/10.1016/j.arth.2018.02.078
Kunutsor SK, Whitehouse MR, Lenguerrand E, Blom AW, Beswick AD, INFORM Team (2016) Re-infection outcomes following one- and two-stage surgical revision of infected knee prosthesis: a systematic review and meta-analysis. PLoS ONE 11:0151537. https://doi.org/10.1371/journal.pone.0151537
EPRD (2021) The German Arthroplasty Registry (EPRD) Annual Report 2021, EPRD Deutsche Endoprothesenregister gGmbH. EPRD. https://doi.org/10.36186/reporteprd052022
Esteva A, Kuprel B, Novoa RA, Ko J, Swetter SM, Blau HM et al (2017) Dermatologist-level classification of skin cancer with deep neural networks. Nature 542:115–118. https://doi.org/10.1038/nature21056
Coudray N, Ocampo PS, Sakellaropoulos T, Narula N, Snuderl M, Fenyö D et al (2018) Classification and mutation prediction from non-small cell lung cancer histopathology images using deep learning. Nat Med 24:1559–1567. https://doi.org/10.1038/s41591-018-0177-5
Khalilia M, Chakraborty S, Popescu M (2011) Predicting disease risks from highly imbalanced data using random forest. BMC Med Inform Decis Mak 11:51. https://doi.org/10.1186/1472-6947-11-51
Kawakami E, Tabata J, Yanaihara N, Ishikawa T, Koseki K, Iida Y et al (2019) Application of artificial intelligence for preoperative diagnostic and prognostic prediction in epithelial ovarian cancer based on blood biomarkers. Clin Cancer Res 25:3006–3015. https://doi.org/10.1158/1078-0432.CCR-18-3378
Kuo F-C, Hu W-H, Hu Y-J (2022) Periprosthetic joint infection prediction via machine learning: comprehensible personalized decision support for diagnosis. J Arthroplasty 37:132–141. https://doi.org/10.1016/j.arth.2021.09.005
Parvizi J, Gehrke T (2013) Proceedings of the international consensus meeting on periprosthetic joint infection. Work group 7, diagnosis of periprosthetic joint infection. J Arthroplasty. https://doi.org/10.1016/j.arth.2014.03.009
Breiman L (2001) Random forests. Mach Learn 45:5–32. https://doi.org/10.1023/A:1010933404324
McInnes L, Healy J, Melville J (2018) UMAP uniform manifold approximation and projection for dimension reduction. J Open Sour Softw. https://doi.org/10.21105/joss.00861
Beck JR, Robert BJ (1988) Markov models of natural history. J Clin Epidemiol 41:619–621. https://doi.org/10.1016/0895-4356(88)90113-8
Sonnenberg FA, Beck JR (1993) Markov models in medical decision making: a practical guide. Med Decis Making 13:322–338. https://doi.org/10.1177/0272989X9301300409
Tada T, Kumada T, Toyoda H, Ohisa M, Akita T, Tanaka J (2018) Long-term natural history of liver disease in patients with chronic hepatitis B virus infection: an analysis using the Markov chain model. J Gastroenterol 53:1196–1205. https://doi.org/10.1007/s00535-018-1467-x
Lundberg SM, Erion G, Chen H, DeGrave A, Prutkin JM, Nair B et al (2020) From local explanations to global understanding with explainable AI for trees. Nat Mach Intell 2:56–67. https://doi.org/10.1038/s42256-019-0138-9
Jupiter JB, Karchmer AW, Lowell JD, Harris WH (1981) Total hip arthroplasty in the treatment of adult hips with current or quiescent sepsis. J Bone Joint Surg Am 63:194–200
Peersman G, Laskin R, Davis J, Peterson M (2001) Infection in total knee replacement: a retrospective review of 6489 total knee replacements. Clin Orthop Relat Res 392:15–23
Marchant MH Jr, Viens NA, Cook C, Vail TP, Bolognesi MP (2009) The impact of glycemic control and diabetes mellitus on perioperative outcomes after total joint arthroplasty. J Bone Joint Surg Am 91:1621–1629. https://doi.org/10.2106/JBJS.H.00116
Jaberi FM, Parvizi J, Haytmanek CT, Joshi A, Purtill J (2008) Procrastination of wound drainage and malnutrition affect the outcome of joint arthroplasty. Clin Orthop Relat Res 466:1368–1371. https://doi.org/10.1007/s11999-008-0214-7
Mills E, Eyawo O, Lockhart I, Kelly S, Wu P, Ebbert JO (2011) Smoking cessation reduces postoperative complications: a systematic review and meta-analysis. Am J Med 124:144–54.e8. https://doi.org/10.1016/j.amjmed.2010.09.013
Lieberman JR, Fuchs MD, Haas SB, Garvin KL, Goldstock L, Gupta R et al (1995) Hip arthroplasty in patients with chronic renal failure. J Arthroplasty 10:191–195. https://doi.org/10.1016/s0883-5403(05)80126-3
Pour AE, Matar WY, Mehdi Jafari S, Purtill JJ, Austin MS, Parvizi J (2011) Total joint arthroplasty in patients with hepatitis C. J Bone Joint Surg 93:1448–1454. https://doi.org/10.2106/jbjs.j.00219
Shohat N, Goswami K, Tan TL, Yayac M, Soriano A, Sousa R et al (2020) 2020 Frank Stinchfield award: identifying who will fail following irrigation and debridement for prosthetic joint infection. Bone Joint J. 102-B:11–19. https://doi.org/10.1302/0301-620X.102B7.BJJ-2019-1628.R1
Rouzrokh P, Ramazanian T, Wyles CC, Philbrick KA, Cai JC, Taunton MJ et al (2021) Deep learning artificial intelligence model for assessment of hip dislocation risk following primary total hip arthroplasty from postoperative radiographs. J Arthroplasty 36:2197–203.e3. https://doi.org/10.1016/j.arth.2021.02.028
Ye Y, Chen W, Gu M, Xian G, Pan B, Zheng L et al (2020) Serum globulin and albumin to globulin ratio as potential diagnostic biomarkers for periprosthetic joint infection: a retrospective review. J Orthop Surg Res 15:459. https://doi.org/10.1186/s13018-020-01959-1
Greenky M, Gandhi K, Pulido L, Restrepo C, Parvizi J (2012) Preoperative anemia in total joint arthroplasty: is it associated with periprosthetic joint infection? Clin Orthop Relat Res 470:2695–2701. https://doi.org/10.1007/s11999-012-2435-z
Sodhi N, Anis HK, Vakharia RM, Acuña AJ, Gold PA, Garbarino LJ et al (2020) What are risk factors for infection after primary or revision total joint arthroplasty in patients older than 80 years? Clin Orthop Relat Res 478:1741–1751. https://doi.org/10.1097/corr.0000000000001389
Qin L, Li F, Gong X, Wang J, Huang W, Hu N (2020) Combined measurement of D-Dimer and C-reactive protein levels: highly accurate for diagnosing chronic periprosthetic joint infection. J Arthroplasty 35:229–234. https://doi.org/10.1016/j.arth.2019.08.012
Li C, Ojeda Thies C, Xu C, Trampuz A (2020) Is combining serum interleukin-6 and C-reactive protein a reliable diagnostic tool in periprosthetic joint infections? J Orthop Surg Res 15:450. https://doi.org/10.1186/s13018-020-01864-7
Maier SP, Klemt C, Tirumala V, Oganesyan R, van den Kieboom J, Kwon Y-M (2020) Elevated ESR/CRP ratio is associated with reinfection after debridement, antibiotics, and implant retention in chronic periprosthetic joint infections. J Arthroplasty 35:3254–3260. https://doi.org/10.1016/j.arth.2020.06.007
Acknowledgements
We thank Shiho Saito for invaluable assistance with data management.
Funding
We thank Life Innovation Platform YOKOHAMA (LIP. Yokohama) for financial support. This work was, in part, supported by SECOM Science and Technology Foundation (to E.K.), Japan Science and Technology Agency (JST) Moonshot R&D Grants (JPMJMS2025, to E.K.), a JST CREST Grant (JPMJCR20H4, to E.K.), and Japan Agency for Medical Research and Development (AMED) Grants (JP21wm0325007, JP20fk0108412, JP20fk0108413, JP21gm5010003, all to E.K.).
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of interest
The authors declare that they have no conflicts of interest.
Ethical approval
This retrospective observational study was approved by our institutional research ethics committee (UMIN ID: UMIN000030180).
Informed consent
Informed consent was obtained by opt-out form.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Kawakami, E., Kobayashi, N., Ichihara, Y. et al. Monitoring of blood biochemical markers for periprosthetic joint infection using ensemble machine learning and UMAP embedding. Arch Orthop Trauma Surg 143, 6057–6067 (2023). https://doi.org/10.1007/s00402-023-04898-8
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00402-023-04898-8