Browsing by Author "Wu, Guosong"

Now showing 1 - 3 of 3
  • Thumbnail Image
    ItemOpen Access
    Development of machine learning models for the detection of surgical site infections following total hip and knee arthroplasty: a multicenter cohort study
    (2023-09-02) Wu, Guosong; Cheligeer, Cheligeer; Southern, Danielle A.; Martin, Elliot A.; Xu, Yuan; Leal, Jenine; Ellison, Jennifer; Bush, Kathryn; Williamson, Tyler; Quan, Hude; Eastwood, Cathy A.
    Abstract Background Population based surveillance of surgical site infections (SSIs) requires precise case-finding strategies. We sought to develop and validate machine learning models to automate the process of complex (deep incisional/organ space) SSIs case detection. Methods This retrospective cohort study included adult patients (age ≥ 18 years) admitted to Calgary, Canada acute care hospitals who underwent primary total elective hip (THA) or knee (TKA) arthroplasty between Jan 1st, 2013 and Aug 31st, 2020. True SSI conditions were judged by the Alberta Health Services Infection Prevention and Control (IPC) program staff. Using the IPC cases as labels, we developed and validated nine XGBoost models to identify deep incisional SSIs, organ space SSIs and complex SSIs using administrative data, electronic medical records (EMR) free text data, and both. The performance of machine learning models was assessed by sensitivity, specificity, positive predictive value, negative predictive value, F1 score, the area under the receiver operating characteristic curve (ROC AUC) and the area under the precision–recall curve (PR AUC). In addition, a bootstrap 95% confidence interval (95% CI) was calculated. Results There were 22,059 unique patients with 27,360 hospital admissions resulting in 88,351 days of hospital stay. This included 16,561 (60.5%) TKA and 10,799 (39.5%) THA procedures. There were 235 ascertained SSIs. Of them, 77 (32.8%) were superficial incisional SSIs, 57 (24.3%) were deep incisional SSIs, and 101 (42.9%) were organ space SSIs. The incidence rates were 0.37 for superficial incisional SSIs, 0.21 for deep incisional SSIs, 0.37 for organ space and 0.58 for complex SSIs per 100 surgical procedures, respectively. The optimal XGBoost models using administrative data and text data combined achieved a ROC AUC of 0.906 (95% CI 0.835–0.978), PR AUC of 0.637 (95% CI 0.528–0.746), and F1 score of 0.79 (0.67–0.90). Conclusions Our findings suggest machine learning models derived from administrative data and EMR text data achieved high performance and can be used to automate the detection of complex SSIs.
  • Thumbnail Image
    ItemOpen Access
    The Development of Patient Safety Indicators for Critically Ill Patients Admitted to the Intensive Care Unit
    (2020-07-09) Wu, Guosong; Stelfox, Henry Thomas; Wu, Qunhong; Quan, Hude; Ronksley, Paul Everett; Holroyd-Leduc, Jayna M.
    Patient safety is a top priority in critical care. Patient safety indicators (PSIs) have been developed to measure critical care safety. Despite several approaches have been used to develop PSIs (deductive, inductive, and joint approach), PSIs are widely applied to measure the care performance in Intensive Care Units (ICUs). The overall objective of this thesis was to use a joint approach to develop a PSI that can be used to evaluate the safety of critical care and to assess the reliability and validity of this PSI. Three studies were conducted to achieve this objective. In the first study, we systematically reviewed the literature of PSIs proposed for patients admitted to an ICU. Heterogeneity of PSI definitions were observed. Among 44 unique PSIs identified from 21 studies, reliability and validity of four PSIs was only reported in one study. In the second study, a panel of ICU experts evaluated the 44 unique PSIs identified in our systematic review. The expert panel proposed an additional 20 PSIs and, during a second round, evaluated and prioritized the 64 PSIs using a modified RAND Appropriateness Method. The top ten ranked PSIs included clinically important concepts that had data available in a local electronic database. This included patient falls, which we investigated in our final study. In the third study, we constructed a fall risk predictive model leveraging provincial electronic medical record (EMR) and examined its reliability and validity. Overall, fall incidence rate was 1.55 (95% CI 1.36-1.76) per 1,000 patient days. The model displayed an excellent discrimination (AUC 0.82) and calibration. We propose the definition of Risk-adjusted Standardized Fall Rate (RSFR) to compared critical care performance across sites and overtime in order to test effectiveness of quality improvement strategies. RSFR demonstrates good test re-test reliability (r=0.662) and construct validity (r=0.533). The study proposed PSI for patient falls was developed from a comprehensive systematic review of the literature, structured expert panel review of candidate PSIs, and validated using a local EMR database. The main findings of this thesis support RSFR could be measured and monitored provincially or nationally to benchmark critical care safety performance.
  • Thumbnail Image
    ItemOpen Access
    Validation of large language models for detecting pathologic complete response in breast cancer using population-based pathology reports
    (2024-10-03) Cheligeer, Ken; Wu, Guosong; Laws, Alison; Quan, May L.; Li, Andrea; Brisson, Anne-Marie; Xie, Jason; Xu, Yuan
    Abstract Aims The primary goal of this study is to evaluate the capabilities of Large Language Models (LLMs) in understanding and processing complex medical documentation. We chose to focus on the identification of pathologic complete response (pCR) in narrative pathology reports. This approach aims to contribute to the advancement of comprehensive reporting, health research, and public health surveillance, thereby enhancing patient care and breast cancer management strategies. Methods The study utilized two analytical pipelines, developed with open-source LLMs within the healthcare system’s computing environment. First, we extracted embeddings from pathology reports using 15 different transformer-based models and then employed logistic regression on these embeddings to classify the presence or absence of pCR. Secondly, we fine-tuned the Generative Pre-trained Transformer-2 (GPT-2) model by attaching a simple feed-forward neural network (FFNN) layer to improve the detection performance of pCR from pathology reports. Results In a cohort of 351 female breast cancer patients who underwent neoadjuvant chemotherapy (NAC) and subsequent surgery between 2010 and 2017 in Calgary, the optimized method displayed a sensitivity of 95.3% (95%CI: 84.0–100.0%), a positive predictive value of 90.9% (95%CI: 76.5–100.0%), and an F1 score of 93.0% (95%CI: 83.7–100.0%). The results, achieved through diverse LLM integration, surpassed traditional machine learning models, underscoring the potential of LLMs in clinical pathology information extraction. Conclusions The study successfully demonstrates the efficacy of LLMs in interpreting and processing digital pathology data, particularly for determining pCR in breast cancer patients post-NAC. The superior performance of LLM-based pipelines over traditional models highlights their significant potential in extracting and analyzing key clinical data from narrative reports. While promising, these findings highlight the need for future external validation to confirm the reliability and broader applicability of these methods.