1
|
Venäläinen MS, Panula VJ, Eskelinen AP, Fenstad AM, Furnes O, Hallan G, Rolfson O, Kärrholm J, Hailer NP, Pedersen AB, Overgaard S, Mäkelä KT, Elo LL. Prediction of Early Adverse Events After THA: A Comparison of Different Machine-Learning Strategies Based on 262,356 Observations From the Nordic Arthroplasty Register Association (NARA) Dataset. ACR Open Rheumatol 2024. [PMID: 39040016 DOI: 10.1002/acr2.11709] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 12/10/2023] [Revised: 04/20/2024] [Accepted: 06/08/2024] [Indexed: 07/24/2024] Open
Abstract
OBJECTIVE Preoperative risk prediction models can support shared decision-making before total hip arthroplasties (THAs). Here, we compare different machine-learning (ML) approaches to predict the six-month risk of adverse events following primary THA to obtain accurate yet simple-to-use risk prediction models. METHODS We extracted data on primary THAs (N = 262,356) between 2010 and 2018 from the Nordic Arthroplasty Register Association dataset. We benchmarked a variety of ML algorithms in terms of the area under the receiver operating characteristic curve (AUROC) for predicting the risk of revision caused by periprosthetic joint infection (PJI), dislocation or periprosthetic fracture (PPF), and death. All models were internally validated against a randomly selected test cohort (one-third of the data) that was not used for training the models. RESULTS The incidences of revisions because of PJI, dislocation, and PPF were 0.8%, 0.4%, and 0.3%, respectively, and the incidence of death was 1.2%. Overall, Lasso regression with stable iterative variable selection (SIVS) produced models using only four to five input variables but with AUROC comparable to more complex models using all 32 variables available. The SIVS-based Lasso models based on age, sex, preoperative diagnosis, bearing couple, fixation, and surgical approach predicted the risk of revisions caused by PJI, dislocations, and PPF, as well as death, with AUROCs of 0.61, 0.67, 0.76, and 0.86, respectively. CONCLUSION Our study demonstrates that satisfactory predictive potential for adverse events following THA can be reached with parsimonious modeling strategies. The SIVS-based Lasso models may serve as simple-to-use tools for clinical risk assessment in the future.
Collapse
Affiliation(s)
- Mikko S Venäläinen
- Turku University Hospital, University of Turku and Åbo Akademi University, Turku, Finland
| | | | - Antti P Eskelinen
- Coxa Hospital for Joint Replacement and University of Tampere, Tampere, Finland, and the Finnish Arthroplasty Register, Finnish Institute for Health and Welfare, Helsinki, Finland
| | | | - Ove Furnes
- Haukeland University Hospital and University of Bergen, Bergen, Norway
| | - Geir Hallan
- Haukeland University Hospital and University of Bergen, Bergen, Norway
| | - Ola Rolfson
- University of Gothenburg, Gothenburg, Sweden
| | | | | | - Alma B Pedersen
- Aarhus University Hospital and Aarhus University, Aarhus, Denmark
| | - Søren Overgaard
- Copenhagen University Hospital and University of Copenhagen, Copenhagen, Denmark
| | - Keijo T Mäkelä
- Turku University Hospital and University of Turku, Turku, Finland, and the Finnish Arthroplasty Register, Finnish Institute for Health and Welfare, Helsinki, Finland
| | - Laura L Elo
- University of Turku and Åbo Akademi University, Turku, Finland
| |
Collapse
|
2
|
Perni S, Prokopovich P. Risk equations for prosthetic joint infections (PJIs) in UK: a retrospective study using the Clinical Practice Research Datalink (CPRD) AURUM and GOLD databases. BMJ Open 2024; 14:e082501. [PMID: 38719289 PMCID: PMC11086542 DOI: 10.1136/bmjopen-2023-082501] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Received: 11/28/2023] [Accepted: 04/26/2024] [Indexed: 05/12/2024] Open
Abstract
BACKGROUND Prosthetic joint infections (PJIs) are a serious negative outcome of arthroplasty with incidence of about 1%. Risk of PJI could depend on local treatment policies and guidelines; no UK-specific risk scoring is currently available. OBJECTIVE To determine a risk quantification model for the development of PJI using electronic health records. DESIGN Records in Clinical Practice Research Datalink (CPRD) GOLD and AURUM of patients undergoing hip or knee arthroplasty between January 2007 and December 2014, with linkage to Hospital Episode Statistics and Office of National Statistics, were obtained. Cohorts' characteristics and risk equations through parametric models were developed and compared between the two databases. Pooled cohort risk equations were determined for the UK population and simplified through stepwise selection. RESULTS After applying the inclusion/exclusion criteria, 174 905 joints (1021 developed PJI) were identified in CPRD AURUM and 48 419 joints (228 developed PJI) in CPRD GOLD. Patients undergoing hip or knee arthroplasty in both databases exhibited different sociodemographic characteristics and medical/drug history. However, the quantification of the impact of such covariates (coefficients of parametric models fitted to the survival curves) on the risk of PJI between the two cohorts was not statistically significant. The log-normal model fitted to the pooled cohorts after stepwise selection had a C-statistic >0.7. CONCLUSIONS The risk prediction tool developed here could help prevent PJI through identifying modifiable risk factors pre-surgery and identifying the patients most likely to benefit from close monitoring/preventive actions. As derived from the UK population, such tool will help the National Health Service reduce the impact of PJI on its resources and patient lives.
Collapse
|
3
|
Fenske F, Kujat B, Krause L, Meyer S, Sander AK, Repmann J, Neuhaus M, Haak R, Roth A, Lethaus B, Ziebolz D, Schmalz G. Preoperative dental screening can reduce periprosthetic infections of hip and knee endoprostheses in the first month after surgery: results of a cohort study. Infection 2024; 52:535-543. [PMID: 38060067 PMCID: PMC10954939 DOI: 10.1007/s15010-023-02128-2] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 04/18/2023] [Accepted: 10/27/2023] [Indexed: 12/08/2023]
Abstract
PURPOSE The oral cavity and, in particular, potential oral foci might pose a risk of periprosthetic joint infection (PJI). The aim of this cohort study was to determine whether practical preoperative dental screening would reduce the prevalence of early PJI in the first month after surgery. METHODS Patients attending a specialized endoprosthesis implantation clinic between 2018 and 2022 were recruited. Two groups were examined. The test group consisted of patients attending the clinic between 2020 and 2022 and who were referred to their family dentist using a standardized form. The comparison group consisted of patients who were treated in the clinic between 2018 and 2020. They were not referred to their family dentist. The two groups were compared for the prevalence of PJI. Univariate analysis followed by multiple logistic regression was performed to confirm risk factors for PJI in this cohort. RESULTS 2560 individuals (test group: 1227, comparison group: 1333) were included. The prevalence of PJI was significantly lower in the test group (0.8% vs. 1.8%, p = 0.04). Multiple logistic regression with PJI as the dependent variable showed that a dental referral was a strong predictor of a lower prevalence of PJI (OR: 0.43, CI95 0.205-0.917, p = 0.03). Male gender was also strongly associated with a higher frequency of PJI (OR: 2.68, CI95 1.32-5.42, p = 0.01). Age (OR: 1.06, CI95 1.01-1.10, p = 0.01) and BMI (OR: 1.11, CI95 1.05-1.17, p < 0.01) had little effect on the risk of PJI. CONCLUSION Dental referral using a standardized form can reduce the prevalence of early PJI. Accordingly, orthopedists and dentists should collaborate in this practical way.
Collapse
Affiliation(s)
- Fabian Fenske
- Department of Oral and Maxillofacial Surgery, University of Leipzig, Liebigstr. 12, 04103, Leipzig, Germany.
| | - Benjamin Kujat
- Specialized Clinic for Orthopedics, Mediclin Waldkrankenhaus Bad Düben, 04848, Bad Düben, Germany
| | - Leah Krause
- Department of Cariology, Endodontology and Periodontology, University of Leipzig, 04103, Leipzig, Germany
| | - Stephan Meyer
- Department of Oral and Maxillofacial Surgery, University of Leipzig, Liebigstr. 12, 04103, Leipzig, Germany
| | - Anna-Katharina Sander
- Department of Oral and Maxillofacial Surgery, University of Leipzig, Liebigstr. 12, 04103, Leipzig, Germany
| | - Jaqueline Repmann
- Specialized Clinic for Orthopedics, Mediclin Waldkrankenhaus Bad Düben, 04848, Bad Düben, Germany
| | - Michael Neuhaus
- Department of Oral and Maxillofacial Surgery, University of Leipzig, Liebigstr. 12, 04103, Leipzig, Germany
| | - Rainer Haak
- Department of Cariology, Endodontology and Periodontology, University of Leipzig, 04103, Leipzig, Germany
| | - Andreas Roth
- Department of Orthopaedics, Trauma and Plastic Surgery, University Hospital Leipzig, 04103, Leipzig, Germany
| | - Bernd Lethaus
- Department of Oral and Maxillofacial Surgery, University of Leipzig, Liebigstr. 12, 04103, Leipzig, Germany
| | - Dirk Ziebolz
- Department of Cariology, Endodontology and Periodontology, University of Leipzig, 04103, Leipzig, Germany
| | - Gerhard Schmalz
- Department of Cariology, Endodontology and Periodontology, University of Leipzig, 04103, Leipzig, Germany
| |
Collapse
|
4
|
Zeng ZJ, Yao FM, He W, Wei QS, He MC. Incidence of periprosthetic joint infection after primary total hip arthroplasty is underestimated: a synthesis of meta-analysis and bibliometric analysis. J Orthop Surg Res 2023; 18:610. [PMID: 37605235 PMCID: PMC10440885 DOI: 10.1186/s13018-023-04060-5] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Received: 05/21/2023] [Accepted: 07/30/2023] [Indexed: 08/23/2023] Open
Abstract
The American Musculoskeletal Society updated the diagnostic criteria for periprosthetic joint infection (PJI) in 2011 and 2018. However, the overall incidence of PJI since the introduction of these new standards has not been assessed. In order to fill this knowledge gap, a single-group meta-analysis was conducted using articles obtained from several databases, focusing on the incidence of PJI after primary total hip arthroplasty (THA). Our study revealed a significant difference in the incidence of PJI reported by different national or regional databases. Moreover, most cases of PJI were found to be underestimated. This highlights the crucial need for standardized diagnostic criteria and monitoring methods to accurately identify and track cases of PJI. Furthermore, a bibliometric analysis was conducted to provide a comprehensive overview of the current state of research on PJI after THA. This analysis explored the most productive countries, organizations, journals, and individuals in this research area. Additionally, it identified the research trends and hotspots of the last decade, highlighting the advancements and areas of focus in this field. By conducting these analyses, the study aims to contribute to the understanding of PJI after THA and provide valuable insights for clinicians, researchers, and policymakers involved in the management of this condition.
Collapse
Affiliation(s)
- Zi-Jun Zeng
- The Third School of Clinical Medicine, Guangzhou University of Chinese Medicine, Guangzhou, China
| | - Fang-Ming Yao
- The Third School of Clinical Medicine, Guangzhou University of Chinese Medicine, Guangzhou, China
| | - Wei He
- The Third School of Clinical Medicine, Guangzhou University of Chinese Medicine, Guangzhou, China.
- Guangdong Research Institute for Orthopedics and Traumatology of Chinese Medicine, Guangzhou, China.
- The Third Affiliated Hospital of Guangzhou University of Chinese Medicine, Guangzhou, China.
| | - Qiu-Shi Wei
- The Third School of Clinical Medicine, Guangzhou University of Chinese Medicine, Guangzhou, China.
- Guangdong Research Institute for Orthopedics and Traumatology of Chinese Medicine, Guangzhou, China.
- The Third Affiliated Hospital of Guangzhou University of Chinese Medicine, Guangzhou, China.
| | - Min-Cong He
- The Third School of Clinical Medicine, Guangzhou University of Chinese Medicine, Guangzhou, China.
- Guangdong Research Institute for Orthopedics and Traumatology of Chinese Medicine, Guangzhou, China.
- The Third Affiliated Hospital of Guangzhou University of Chinese Medicine, Guangzhou, China.
| |
Collapse
|
5
|
Affiliation(s)
- Robin Patel
- From the Division of Clinical Microbiology, Department of Laboratory Medicine and Pathology, and the Division of Public Health, Infectious Diseases, and Occupational Medicine, Department of Medicine, Mayo Clinic, Rochester, MN
| |
Collapse
|
6
|
Cheng X, Liu Y, Wang W, Yan J, Lei X, Wu H, Zhang Y, Zhu Y. Preoperative Risk Factor Analysis and Dynamic Online Nomogram Development for Early Infections Following Primary Hip Arthroplasty in Geriatric Patients with Hip Fracture. Clin Interv Aging 2022; 17:1873-1883. [PMID: 36575659 PMCID: PMC9790145 DOI: 10.2147/cia.s392393] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 10/05/2022] [Accepted: 12/13/2022] [Indexed: 12/24/2022] Open
Abstract
Background Hip arthroplasty is in increasing demand with the aging of the world population, and early infections, such as pneumonia, surgical site infection (SSI), and urinary tract infection (UTI), are uncommon but fatal complications following hip arthroplasty. This study aimed to identify preoperative risk factors independently associated with early infections following primary arthroplasty in geriatric hip fracture patients, and to develop a prediction nomogram. Methods Univariate and multivariate logistical analyses were performed to identify the independent risk factors for early infections, which were combined and transformed into a nomogram model. The prediction model was evaluated by using the area under the receiver operating characteristic curve (AUC), Hosmer-Lemeshow test, concordance index (C-index), 1000 bootstrap replications, decision curve analysis (DCA), and calibration curve. Results One thousand eighty-four eligible patients got included and 7 preoperative variables were identified to be independently associated with early infections, including heart disease (odds ratio (OR): 2.17; P: 0.026), cerebrovascular disease (OR: 2.25; P: 0.019), liver disease (OR: 8.99; P: <0.001), time to surgery (OR: 1.10; P: 0.012), hematocrit (<lower limit; OR: 3.72; P: 0.015), the platelet-to-mean platelet volume ratio (PMR; >44.52; OR: 2.73; P: 0.047), and high-sensitivity C-reactive protein (HCRP; >78.64mg/L; OR: 3.71; P: <0.001). For the nomogram model, AUC was 0.807 (95% confidence interval (CI): 0.742-0.873), the Hosmer-Lemeshow test demonstrated no overfitting (P = 0.522), and C-index was 0.807 (95% CI: 0.742-0.872) with corrected value of 0.784 after 1000 bootstrapping validations. Moreover, the calibration curve and DCA exhibited the tools' good prediction consistency and clinical practicability. Conclusion Heart disease, cerebrovascular disease, liver disease, time to surgery, hematocrit, PMR, and HCRP were significant preoperative predictors for early infections following primary arthroplasty in elderly hip fracture patients, and the converted nomogram model had strong discriminatory ability and translatability to clinical application.
Collapse
Affiliation(s)
- Xinqun Cheng
- Department of Orthopedic Surgery, the 3rd Hospital of Hebei Medical University, Shijiazhuang, Hebei, 050051, People’s Republic of China,Hebei Orthopedic Research Institute, Key Laboratory of Biomechanics of Hebei Province, Shijiazhuang, Hebei, 050051, People’s Republic of China
| | - Yan Liu
- Cheeloo College of Medicine, Qilu Hospital of Shandong University, Jinan, Shandong, 250012, People’s Republic of China
| | - Weitong Wang
- Department of Marketing and International Business, Lingnan University, Hong Kong, People’s Republic of China
| | - Jincheng Yan
- Department of Orthopedic Surgery, the 3rd Hospital of Hebei Medical University, Shijiazhuang, Hebei, 050051, People’s Republic of China
| | - Xiang Lei
- Department of Orthopedic Surgery, the 3rd Hospital of Hebei Medical University, Shijiazhuang, Hebei, 050051, People’s Republic of China
| | - Haifeng Wu
- Department of Orthopedic Surgery, the 3rd Hospital of Hebei Medical University, Shijiazhuang, Hebei, 050051, People’s Republic of China
| | - Yingze Zhang
- Department of Orthopedic Surgery, the 3rd Hospital of Hebei Medical University, Shijiazhuang, Hebei, 050051, People’s Republic of China,Hebei Orthopedic Research Institute, Key Laboratory of Biomechanics of Hebei Province, Shijiazhuang, Hebei, 050051, People’s Republic of China,Correspondence: Yingze Zhang; Yanbin Zhu, Hebei Bone Research Institute, Key Laboratory of Biomechanics of Hebei Province, Shijiazhuang, Hebei, 050051, People’s Republic of China, Email ;
| | - Yanbin Zhu
- Department of Orthopedic Surgery, the 3rd Hospital of Hebei Medical University, Shijiazhuang, Hebei, 050051, People’s Republic of China,Hebei Orthopedic Research Institute, Key Laboratory of Biomechanics of Hebei Province, Shijiazhuang, Hebei, 050051, People’s Republic of China
| |
Collapse
|