Machine learning predicts cancer-associated deep vein thrombosis using clinically available variables.

PURPOSE: To develop and validate machine learning (ML) models for cancer-associated deep vein thrombosis (DVT) and to compare the performance of these models with the Khorana score (KS). METHODS: We randomly extracted data of 2100 patients with cancer between Jan. 1, 2017, and Oct. 31, 2019, and 1035 patients who underwent Doppler ultrasonography were enrolled. Univariate analysis and Lasso regression were applied to select important predictors. Model training and hyperparameter tuning were implemented on 70% of the data using a ten-fold cross-validation method. The remaining 30% of the data were used to compare the performance with seven indicators (area under the receiver operating characteristic curve [AUC], sensitivity, specificity, accuracy, balanced accuracy, Brier score, and calibration curve), among all five ML models (linear discriminant analysis [LDA], logistic regression [LR], classification tree [CT], random forest [RF], and support vector machine [SVM]), and the KS. RESULTS: The incidence of cancer-associated DVT was 22.3%. The top five predictors were D-dimer level, age, Charlson Comorbidity Index (CCI), length of stay (LOS), and previous VTE (venous thromboembolism) history according to RF. Only LDA (AUC = 0.773) and LR (AUC = 0.772) outperformed KS (AUC = 0.642), and combination with D-dimer showed improved performance in all models. A nomogram and web calculator https://webcalculatorofcancerassociateddvt.shinyapps.io/dynnomapp/ were used to visualize the best recommended LR model. CONCLUSION: This study developed and validated cancer-associated DVT predictive models using five ML algorithms and visualized the best recommended model using a nomogram and web calculator. The nomogram and web calculator developed in this study may assist doctors and nurses in evaluating individualized cancer-associated DVT risk and making decisions. However, other prospective cohort studies should be conducted to externally validate the recommended model.

Bowel Dysfunction and Self-management for Bowel Symptoms After Sphincter-Preserving Surgery: A Cross-sectional Survey of Chinese Rectal Cancer Patients.

BACKGROUND: After sphincter-preserving surgery (SPS), patients with rectal cancer present multiple bowel symptoms, which significantly affect their daily lives. However, to date, few studies have been conducted in China to identify bowel symptoms and self-management behaviors and to explore the relationship between these behaviors and bowel symptoms. OBJECTIVE: To describe bowel symptoms among Chinese patients with rectal cancer after SPS, the self-management behaviors used, and the relationship between the symptoms and behaviors. METHODS: A convenience sample of 175 rectal cancer patients who underwent SPS was recruited from 2 tertiary hospitals in Beijing, China, between July and December 2014. The participants completed a general information structural questionnaire, the Chinese version of the Memorial Sloan-Kettering Cancer Center Bowel Function Instrument, and the Bowel Symptoms Self-Management Behaviors Questionnaire. RESULTS: The frequencies of 18 bowel symptoms ranged between 16.0% and 96.0%. Diet self-management was the most common behavior and therapy self-management was the least frequent behavior. Although therapy self-management had best effects relatively, other effective behaviors included self-management of the perianal skin and self-management of social activities. The frequency of use of self-management behaviors was significantly positively associated with bowel symptoms (r=-0.232 to -0.580, P < .01). CONCLUSIONS: Several bowel symptoms after SPS are highly prevalent. Patients with severe bowel symptoms are more likely to use self-management behaviors regularly; however, there is no scientific evidence that these behaviors produce satisfactory results. IMPLICATIONS FOR PRACTICE: Healthcare providers should identify bowel symptoms after SPS and develop supportive interventions for the self-management of symptoms.

Validation of the Chinese version of the low anterior resection syndrome score for measuring bowel dysfunction after sphincter-preserving surgery among rectal cancer patients.

PURPOSE: The low anterior resection syndrome (LARS) score is a simple and valid instrument for measuring bowel dysfunction after sphincter-preserving surgery (SPS) among rectal cancer patients. We aimed to translate the LARS score into Chinese, and to test its reliability and validity among Chinese rectal cancer patients. METHODS: The LARS score was translated into Chinese by using internationally recognized forward- and back-translation procedures. In total, 102 patients completed the questionnaire; a subgroup of 20 patients answered the survey twice. The reliability was estimated through the test-retest reliability method. The convergent and discriminant validities were confirmed by measuring the relation of the LARS score with the European Organization for Research and Treatment of Cancer (EORTC) QLQ-C30 and QLQ-CR29 domains, respectively, and testing its ability to differentiate among patients with different clinical characteristics. RESULTS: The Spearman correlation coefficient of the LARS-scores at the two surveys was 0.86 (p < .001), and the linear-weighted kappa values of the five items of the LARS score were 0.38, 0.76, 0.79, 0.77, and 0.78, respectively. The LARS score showed significant correlations with all the assumptive domains of QLQ-C30 and QLQ-CR29, especially flatulence, fecal incontinence, and stool frequency (all p < .05). It could also detect differences between female and male patient groups (p = .033), patients who had/had not undergone radiation therapy (p = .007), those who had undergone surgery in the last <12.0 or ≥12.0 months (p = .002), and those with low or high tumor edge level (p = .017). CONCLUSIONS: The Chinese version of the LARS score had good psychometric properties and can be used in clinical and research settings in the Chinese population.