This study aims to assess the quality and performance of predictive models for colorectal cancer liver metastasis (CRCLM). A systematic review was performed to identify relevant studies from various databases. Studies that described or validated predictive models for CRCLM were included. The methodological quality of the predictive models was assessed. Model performance was evaluated by the reported area under the receiver operating characteristic curve (AUC). Of the 117 articles screened, seven studies comprising 14 predictive models were included. The distribution of included predictive models was as follows: radiomics (n = 3), logistic regression (n = 3), Cox regression (n = 2), nomogram (n = 3), support vector machine (SVM, n = 2), random forest (n = 2), and convolutional neural network (CNN, n = 2). Age, sex, carcinoembryonic antigen, and tumor staging (T and N stage) were the most frequently used clinicopathological predictors for CRCLM. The mean AUCs ranged from 0.697 to 0.870, with 86% of the models demonstrating clear discriminative ability (AUC > 0.70). A hybrid approach combining clinical and radiomic features with SVM provided the best performance, achieving an AUC of 0.870. The overall risk of bias was identified as high in 71% of the included studies. This review highlights the potential of predictive modeling to accurately predict the occurrence of CRCLM. Integrating clinicopathological and radiomic features with machine learning algorithms demonstrates superior predictive capabilities.

This study aims to assess the quality and performance of predictive models for colorectal cancer liver metastasis (CRCLM). A systematic review was performed to identify relevant studies from various databases. Studies that described or validated predictive models for CRCLM were included. The methodological quality of the predictive models was assessed. Model performance was evaluated by the reported area under the receiver operating characteristic curve (AUC). Of the 117 articles screened, seven studies comprising 14 predictive models were included. The distribution of included predictive models was as follows: radiomics (n = 3), logistic regression (n = 3), Cox regression (n = 2), nomogram (n = 3), support vector machine (SVM, n = 2), random forest (n = 2), and convolutional neural network (CNN, n = 2). Age, sex, carcinoembryonic antigen, and tumor staging (T and N stage) were the most frequently used clinicopathological predictors for CRCLM. The mean AUCs ranged from 0.697 to 0.870, with 86% of the models demonstrating clear discriminative ability (AUC > 0.70). A hybrid approach combining clinical and radiomic features with SVM provided the best performance, achieving an AUC of 0.870. The overall risk of bias was identified as high in 71% of the included studies. This review highlights the potential of predictive modeling to accurately predict the occurrence of CRCLM. Integrating clinicopathological and radiomic features with machine learning algorithms demonstrates superior predictive capabilities.

This study aims to assess the quality and performance of predictive models for colorectal cancer liver metastasis (CRCLM). A systematic review was performed to identify relevant studies from various databases. Studies that described or validated predictive models for CRCLM were included. The methodological quality of the predictive models was assessed. Model performance was evaluated by the reported area under the receiver operating characteristic curve (AUC). Of the 117 articles screened, seven studies comprising 14 predictive models were included. The distribution of included predictive models was as follows: radiomics (n = 3), logistic regression (n = 3), Cox regression (n = 2), nomogram (n = 3), support vector machine (SVM, n = 2), random forest (n = 2), and convolutional neural network (CNN, n = 2). Age, sex, carcinoembryonic antigen, and tumor staging (T and N stage) were the most frequently used clinicopathological predictors for CRCLM. The mean AUCs ranged from 0.697 to 0.870, with 86% of the models demonstrating clear discriminative ability (AUC > 0.70). A hybrid approach combining clinical and radiomic features with SVM provided the best performance, achieving an AUC of 0.870. The overall risk of bias was identified as high in 71% of the included studies. This review highlights the potential of predictive modeling to accurately predict the occurrence of CRCLM. Integrating clinicopathological and radiomic features with machine learning algorithms demonstrates superior predictive capabilities.

Purpose: We compared the incidence of venous thromboembolism (VTE) among Asian populations with localized colorectal cancer undergoing curative resection with and without the use of pharmacological thromboprophylaxis (PTP). Methods: A comprehensive literature search was undertaken to identify relevant studies published from January 1, 1980 to February 28, 2022. The inclusion criteria were patients who underwent primary tumor resection for localized nonmetastatic colorectal cancer; an Asian population or studies conducted in an Asian country; randomized controlled trials, case-control studies, or cohort studies; and the incidence of symptomatic VTE, deep vein thrombosis, and/or pulmonary embolism as the primary study outcomes. Data were pooled using a random-effects model. This study was registered in PROSPERO on October 11, 2020 (No. CRD42020206793). Results: Seven studies (2 randomized controlled trials and 5 observational cohort studies) were included, encompassing 5,302 patients. The overall incidence of VTE was 1.4%. The use of PTP did not significantly reduce overall VTE incidence: 1.1% (95% confidence interval [CI], 0%–3.1%) versus 1.9% (95% CI, 0.3%–4.4%; P = 0.55). Similarly, PTP was not associated with significantly lower rates of symptomatic VTE, proximal deep vein thrombosis, or pulmonary embolism. Conclusion The benefit of PTP in reducing VTE incidence among Asian patients undergoing curative resection for localized colorectal cancer has not been clearly established. The decision to administer PTP should be evaluated on a case-bycase basis and with consideration of associated bleeding risks.

This study aims to assess the quality and performance of predictive models for colorectal cancer liver metastasis (CRCLM). A systematic review was performed to identify relevant studies from various databases. Studies that described or validated predictive models for CRCLM were included. The methodological quality of the predictive models was assessed. Model performance was evaluated by the reported area under the receiver operating characteristic curve (AUC). Of the 117 articles screened, seven studies comprising 14 predictive models were included. The distribution of included predictive models was as follows: radiomics (n = 3), logistic regression (n = 3), Cox regression (n = 2), nomogram (n = 3), support vector machine (SVM, n = 2), random forest (n = 2), and convolutional neural network (CNN, n = 2). Age, sex, carcinoembryonic antigen, and tumor staging (T and N stage) were the most frequently used clinicopathological predictors for CRCLM. The mean AUCs ranged from 0.697 to 0.870, with 86% of the models demonstrating clear discriminative ability (AUC > 0.70). A hybrid approach combining clinical and radiomic features with SVM provided the best performance, achieving an AUC of 0.870. The overall risk of bias was identified as high in 71% of the included studies. This review highlights the potential of predictive modeling to accurately predict the occurrence of CRCLM. Integrating clinicopathological and radiomic features with machine learning algorithms demonstrates superior predictive capabilities.

Purpose: This study compared the short- and long-term clinical outcomes of laser hemorrhoidoplasty (LH) vs. conventional hemorrhoidectomy (CH) in patients with grade II/III hemorrhoids. Methods: PubMed/Medline and the Cochrane Library were searched for randomized and nonrandomized studies comparing LH against CH in grade II/III hemorrhoids. The primary outcomes included postoperative use of analgesia, postoperative morbidity (bleeding, urinary retention, pain, thrombosis), and time of return to work/daily activities. Results: Nine studies totaling 661 patients (LH, 336 and CH, 325) were included. The LH group had shorter operative time (P<0.001) and less intraoperative blood loss (P<0.001). Postoperative pain was lower in the LH group, with lower postoperative day 1 (mean difference [MD], –2.09; 95% confidence interval [CI], –3.44 to –0.75; P=0.002) and postoperative day 7 (MD, –3.94; 95% CI, –6.36 to –1.52; P=0.001) visual analogue scores and use of analgesia (risk ratio [RR], 0.59; 95% CI, 0.42–0.81; P=0.001). The risk of postoperative bleeding was also lower in the LH group (RR, 0.18; 95% CI, 0.12– 0.28; P<0.001), with a quicker return to work or daily activities (P=0.002). The 12-month risks of bleeding (P>0.999) and prolapse (P=0.240), and the likelihood of complete resolution at 12 months, were similar (P=0.240). Conclusion LH offers more favorable short-term clinical outcomes than CH, with reduced morbidity and pain and earlier return to work or daily activities. Medium-term symptom recurrence at 12 months was similar. Our results should be verified in future well-designed trials with larger samples.