Identification of muscle-invasion status in bladder cancer patients using natural language processing and machine learning.

Understanding treatment patterns and effectiveness for patients with metastatic prostate cancer (mPCa) is dependent on accurate assessment of metastatic status. The objective was to develop a natural language processing (NLP) model for identifying patients with mPCa and evaluate the model's performance against chart-reviewed data and an International Classification of Diseases (ICD) 9/10 code-based method. In total, 139,057 radiology reports on 6,211 unique patients from the Department of Veterans Affairs were used. The gold standard was metastases by detailed chart review of radiology reports. NLP performance was assessed by sensitivity, specificity, positive predictive value, negative predictive value, and date of metastases detection. Receiver operating characteristic curves was used to assess model performance. When compared with chart review, the NLP model had high sensitivity and specificity (85% and 96%, respectively). The NLP model was able to predict patient-level metastasis status with a sensitivity of 91% and specificity of 81%, whereas sensitivity and specificity using ICD9/10 billing codes were 73% and 86%, respectively. For the NLP model, date of metastases detection was exactly concordant and within < 1 week in 55% and 58% of patients, compared with 8% and 17%, respectively, using the ICD9/10 billing codes method. The area under the curve for the NLP model was 0.911. A limitation is the NLP model was developed on the basis of a subset of patients with mPCa and may not be generalizable to all patients with mPCa. This population-level NLP model for identifying patients with mPCa was more accurate than using ICD9/10 billing codes when compared with chart-reviewed data. Upon further validation, this model may allow for efficient population-level identification of patients with mPCa.

Natural language processing with deep learning for medical adverse event detection from free-text medical narratives: A case study of detecting total hip replacement dislocation

Microbiome Profiling in Bladder Cancer Patients Using the First-morning Urine Sample

To develop and evaluate an internally validated natural language processing (NLP) model to determine guideline adherence of antibiotic decisions in the emergency department (ED) for skin and soft tissue infections (SSTIs). This cross-sectional pilot study developed and applied a random forest (RF)/NLP model to classify clinical narratives of patients with skin infections as requiring either methicillin-resistant Staphylococcus aureus (MRSA) or non-MRSA antibiotic coverage. Conducted at a quaternary care children's hospital with an annual volume of 50,000 ED visits, our study included patients aged 1 to 18 years presenting for SSTIs to the ED from July 1, 2018, through June 30, 2023. Main outcomes included the NLP model's sensitivity, specificity, and receiver operator characteristic (ROC) curve compared to gold standard manual physician review of the medical records. A total of 1588 patients were part of the training data set, with an additional 423 patients utilized for validation. The RF model achieved an area under the curve (AUC) of 0.99, 97.0% sensitivity, and 94.9% specificity. In the validation data set, the model achieved 96.6% sensitivity and 90.1% specificity. Performance remained strong despite absent/missing patient history in certain narratives. The NLP model demonstrated that automated analysis of clinical narratives for determining guideline adherence is feasible. Despite missing data in narratives, the model's high performance suggests potential for broader application. Given the rise in antibiotic-resistant infections and the role of judicious antibiotic use, developing automated systems through NLP can significantly contribute to health care delivery and patient safety. The methodology of this study provides a feasible, sustainable path for similar applications in the emergency setting.

PurposeHigher levels of serum 25-hydroxyvitamin D 25(OHD) are associated with better prognosis in breast and colorectal cancer. However, the evidence is still inconclusive for bladder cancer (BC). Herein, we investigated the diagnosis and prognosis roles of serum levels of 25(OHD) in suspected BC patients presented by hematuria.MethodsThis prospective cohort study involved suspected patients of BC presented with hematuria. Patients were evaluated by CT urogram, office cystoscopy and urine cytology with subsequent inpatient biopsy for positive findings. Baseline blood samples were collected for measurement of 25(OHD) by electrochemiluminescence binding assay at the time of diagnosis. Patients with non-muscle-invasive BC (NMIBC) underwent transurethral resection of bladder tumor (TURBT) and adjuvant intravesical chemotherapy or BCG instillation. Patients were followed up for their recurrence status during 10 to 24 months. Recurrence was defined as the first time of NMIBC pathological relapse during the follow up period.ResultsA total of 115 patients were included in the final analysis. Patients had proven pathological BC (64 with NMIBC, and 20 with muscle invasive) and 31 patients were considered as control group. Controls were those patients with BC-free workup (including cytology, cystoscopy, and upper tract imaging). BC group showed a lower level of 25(OHD) than control group 16.47±5.88 versus 28.99±3.19 ng/mL (p<0.001). In addition, muscle invasive group also showed a lower level than NMIBC group 13.17±4.5 versus 17.49±5.04 ng/mL (P = 0.003). During the follow-up period of, tumor recurrence occurred in 16 (25%) of NMIBC patients. The baseline 25(OHD) were decreased in patients who experienced early recurrence; without being statistically significant (15.99 ± 5.17 vs. 18.38 ± 5.14 ng/mL; p = 0.08). 25(OHD) deficiency/insufficiency occurred in 5 (16.1%) and 64 (76.2%) in control and BC patients, respectively, (odds-ratios (OR): 2.13; 95% confidence intervals (CI), 1.52–2.99; P < 0.0001).ConclusionSerum 25(OHD) is significantly decreased in BC patients especially those with tumor muscle invasive group. However, the baseline serum 25(OHD) does not predict the recurrence in the NMIBC patients.

AimsHeart failure (HF) is a clinical syndrome with no definitive diagnostic tests. HF registries are often based on manual reviews of medical records of hospitalized HF patients identified using International Classification of Diseases (ICD) codes. However, most HF patients are not hospitalized, and manual review of big electronic health record (EHR) data is not practical. The US Department of Veterans Affairs (VA) has the largest integrated healthcare system in the nation, and an estimated 1.5 million patients have ICD codes for HF (HF ICD‐code universe) in their VA EHR. The objective of our study was to develop artificial intelligence (AI) models to phenotype HF in these patients.Methods and resultsThe model development cohort (n = 20 000: training, 16 000; validation 2000; testing, 2000) included 10 000 patients with HF and 10 000 without HF who were matched by age, sex, race, inpatient/outpatient status, hospital, and encounter date (within 60 days). HF status was ascertained by manual chart reviews in VA's External Peer Review Program for HF (EPRP‐HF) and non‐HF status was ascertained by the absence of ICD codes for HF in VA EHR. Two clinicians annotated 1000 random snippets with HF‐related keywords and labelled 436 as HF, which was then used to train and test a natural language processing (NLP) model to classify HF (positive predictive value or PPV, 0.81; sensitivity, 0.77). A machine learning (ML) model using linear support vector machine architecture was trained and tested to classify HF using EPRP‐HF as cases (PPV, 0.86; sensitivity, 0.86). From the ‘HF ICD‐code universe’, we randomly selected 200 patients (gold standard cohort) and two clinicians manually adjudicated HF (gold standard HF) in 145 of those patients by chart reviews. We calculated NLP, ML, and NLP + ML scores and used weighted F scores to derive their optimal threshold values for HF classification, which resulted in PPVs of 0.83, 0.77, and 0.85 and sensitivities of 0.86, 0.88, and 0.83, respectively. HF patients classified by the NLP + ML model were characteristically and prognostically similar to those with gold standard HF. All three models performed better than ICD code approaches: one principal hospital discharge diagnosis code for HF (PPV, 0.97; sensitivity, 0.21) or two primary outpatient encounter diagnosis codes for HF (PPV, 0.88; sensitivity, 0.54).ConclusionsThese findings suggest that NLP and ML models are efficient AI tools to phenotype HF in big EHR data to create contemporary HF registries for clinical studies of effectiveness, quality improvement, and hypothesis generation.

Medical educators pay increasing attention to the potential utility of narrative data for assessment, but lack of efficient and standardized ways of interpreting the data have limited its use. Natural language processing (NLP) algorithms could provide new insights for using narrative data within assessment especially identifying at-risk or low-performing students. Assessment data were reviewed from 16 cohorts of medical students from the University of Cincinnati College of Medicine (graduating classes of 2006-2022). A T-score Average (TSA) was calculated for each student based on clerkship assessment data. Narrative data from core clerkship evaluations in responses to the prompt "any opportunities for improvement" were utilized for analysis. The narrative data and calculated TSA were then used to train and test 4 NLP models with the goal of utilizing NLP to identify at-risk students as defined by the bottom 10% average TSA score. Based on typical NLP methods, the developed NLP models all performed adequately in identifying student performance with an overall accuracy of 0.8 across all 4 models. However, none of the NLP models was able to identify students within the bottom 10% of performance. During this process, we uncovered the presence of "copy/paste" behavior, a previously undocumented phenomenon within narrative data where preceptors duplicated comments from 1 student to the next. Training NLP models including "copy/paste" comments improved NLP ability to identify students within the bottom and top 10% of performance. NLP models were unable to accurately identify at-risk students. Model accuracy increased with the inclusion of "copy/paste" comments, indicating there may be some discriminatory functionality within aspects of narrative data beyond keywords such as lexical diversity and word quantity. Future work will explore how these other aspects of narrative data associate with performance and utilize more advanced large language models of analysis.Based on typical natural language processing (NLP) methods, the tested NLP models performed adequately in identifying student performance with an overall accuracy of 0.8 across all models. However, none of the NLP models was able to identify students within the bottom 10% of performance, revealing widespread "copy/paste" behavior in evaluations that may limit narrative data's discriminatory utility.

108 Background: Classification of disease response is an essential task in cancer research and needs to be done at scale. Automating this process can improve efficiency in the generation of real-world evidence, potentially leading to better patient outcomes. We aim to develop and evaluate Natural Language Processing (NLP) models for this task. Methods: Using 6203 computed tomography (CT) and 1358 magnetic resonance imaging (MRI) reports from 587 patients with lung cancer of all stages seen at the National Cancer Centre Singapore (NCCS), we trained four NLP models (BioBERT, RadBERT-RoBERTA, BioClinicalBERT, GatorTron) to classify the reports into one of four categories: no evidence of disease, stable disease, partial response or disease progression. Model output was compared against human-curated ground truth and performance was evaluated by accuracy. Results: Of the 4 models, GatorTron performed the best (accuracy = 97.1%), followed by RadBERT-RoBERTA (accuracy = 96.2%), BioBERT (accuracy = 94.2%), with BioClinicalBERT being last (accuracy = 90.4%). NLP Model runtimes for the dataset were relatively short, with BioBERT and BioClinicalBERT taking 3 minutes per epoch, RadBERT-RoBERTA taking 6 minutes per epoch, and GatorTron taking 10 minutes per epoch on a single central processing unit (CPU). Conclusions: We have demonstrated the effectiveness of NLP models for classifying disease responses in radiology reports of lung cancer patients. This has the potential to help derive progression-free survival for real-world evidence generation.

How Well Can You Actually Predict Which Non–Muscle-Invasive Bladder Cancer Patients Will Progress?

Treatment Patterns and Overall Survival Outcomes of Octogenarians with Muscle Invasive Cancer of the Bladder: An Analysis of the National Cancer Database

521 Background: Bladder cancer is a common cancer in the US. Approximately seventy-five percent of new cases present as non-muscle invasive bladder cancer (NMIBC) with a high recurrence rate. Our group has demonstrated in well-characterized human NMIBC cell lines that ethacrynic acid (EA) suppresses growth by inhibiting proliferation, clonogenicity, spheroid formation and inducing apoptosis. Additionally, EA affects markers of stemness and may work through the NOTCH signaling pathway. We present a clinical trial characterizing the safety and urinary tract exposure to EA and metabolites following oral administration in patients undergoing transurethral resection of bladder tumor (TURBT). Methods: Institutional review board approval (#3674) was obtained for a Pilot trial in patients with presumed non-muscle invasive bladder cancer undergoing TURBT. All participants were given a single, 50 mg oral dose of EA in the pre-operative bay. Urine was collected at baseline, upon insertion of cystoscope (30-60 minutes post dose), and post-operatively. Urine concentrations of EA and metabolites were determined using a fully-validated UPLC-MS/MS-based analytical method. Participants were monitored for adverse events. Results: Twelve participants participated in the trial between August 2016 and February 2017. Eleven male and one female patient (median age 70.5 years) were enrolled. Final pathology demonstrated urothelial carcinoma in 5/12 (42%) subjects, three with high grade T1 and two with low-grade Ta tumors. Urine concentrations of EA and conjugates observed 30-60 minutes post-dose were as follows: EA 0.1 ug/mL, EA-glutathione 0.2 ug/mL, EA-cysteine 3 ug/mL, and EA-mercapturate 2 ug/mL. No drug-related adverse events were observed following oral EA administration. Conclusions: Despite EA receiving FDA approval over 40 years ago, our pilot trial describes for the first time, urinary tract exposure of EA and its active metabolites in bladder cancer patients after a single oral dose. These data are guiding ongoing preclinical proof of principle studies evaluating EA alone and in combination with standard of care agents in bladder cancer models. These studies warrant further studies of EA in patients with non-muscle invasive bladder cancer. Clinical trial information: NCT02852564.

ObjectiveTo determine patient experience and perception following a diagnosis of non‐muscle‐invasive bladder cancer (NMIBC).Patient and methodsPatients were part of a prospective multicentre observational study recruiting patients with NMIBC for a urine biomarker study (DETECT II; ClinicalTrials.gov: NCT02781428). A mixed‐methods approach comprising: (i) the Brief Illness Perception Questionnaire (Brief‐IPQ) and (ii) semi‐structured interviews to explore patients’ experience of having haematuria, and initial and subsequent experience with a NMIBC diagnosis. Both assessments were completed at 6 months after NMIBC diagnosis.ResultsA total of 213 patients completed the Brief‐IPQ. Patients felt that they had minimal symptoms (median [interquartile range, IQR] score 2 [0–5]) and were not particularly affected emotionally (median [IQR] score 3 [1–6]) with a minimal effect to their daily life (median [IQR] score 2 [0–5]). However, they remained concerned about their cancer diagnosis (median [IQR] score 5 [3–8]) and felt that they had no personal control over the cancer (median [IQR] score 2 [2–5]) and believed that their illness would affect them for some time (median [IQR] score 6 [3–10]). A significant association with a lower personal control of the disease (P < 0.05) and a poorer understanding of the management of NMIBC (P < 0.05) was seen in patients aged >70 years. Many patients were uncertain about the cause of bladder cancer. Qualitative analysis found that at initial presentation of haematuria, most patients were not aware of the risk of bladder cancer. Patients were most anxious and psychologically affected between the interval of cystoscopy diagnosis and transurethral resection of bladder tumour (TURBT). Following TURBT, most patients were positive about their cancer prognosis.ConclusionPatients with NMIBC have a poor perception of disease control and believe that their disease will continue over a prolonged period of time. This is particularly more pertinent in the elderly. Patients are most psychologically affected during the interval between cancer diagnosis following cystoscopy and TURBT. Health awareness about bladder cancer remained poor with a significant number of patients unaware of the causes of bladder cancer. Psychological support and prompt TURBT following bladder cancer diagnosis would help improve the mental health of patients with NMIBC.

Patient and Provider Perception of Transurethral Resection of Bladder Tumor vs Chemoablation for Nonmuscle-invasive Bladder Cancer Treatment.

MAINTENANCE BACILLUS CALMETTE-GUERIN IMMUNOTHERAPY FOR RECURRENT TA, T1 AND CARCINOMA IN SITU TRANSITIONAL CELL CARCINOMA OF THE BLADDER: A RANDOMIZED SOUTHWEST ONCOLOGY GROUP STUDY

Transurethral resection of bladder tumour (TURBT) is the gold standard in diagnosing and treating non-muscle-invasive bladder cancer (NMIBC). The first resectoscope has been introduced more than a century ago, and TURBT remains as a cornerstone in the management of bladder cancer. Although it is a minimally-invasive surgery that has gained favour over the years, the recurrence rate of NMIBC is in general unsatisfactory ranging from 15% to 61% at 1 year and 31% to 78% at 5 years. Tremendous efforts have been made to improve the treatment outcomes of NMIBC, and the current treatment algorithm has been shaped over the past decades. This book chapter shall first discuss about the history of TURBT and the current standard of the initial management of NMIBC, followed by newer concepts that have been proposed including enhanced imaging during TURBT and en bloc resection of bladder tumour. We hope to provide our readers the backbone of the TURBT procedure and new insights that might be helpful in optimizing the management of NMIBC.