Objective To investigate the diagnostic significance of fine needle aspiration cytology (FNAC) combined with BRAFV600E gene detection in the diagnosis of cervical lymph node metastasis of thyroid cancer. Methods Atotal of 140 patients with suspected cervical lymph node metastasis of thyroid cancer were collected as the research objects, and all patients were given ultrasound-guided FNAC and detection of BRAFV600E gene. The significance of the diagnosis was analyzed according to the gold standard after pathological examination. Results All the 140 patients underwent surgical treatment. For FNAC, the sensitivity was 63.6% (84/132), the specificity was 100% (8/8), the accuracy was 65.7% (92/140), the positive predictive value was 100% (84/84), and the negative predictive value was 14.3% (8/56). For detection of BRAFV600E gene, the sensitivity was 84.8% (112/132), the specificity was 100% (8/8), the accuracy was 85.7% (120/140), the positive predictive value was 100% (112/112), and the negative predictive value was 28.5% (8/28). For FNAC combined with BRAFV600E gene detection, the sensitivity was 90.9% (120/132), the specificity was 100% (8/8), the accuracy was 91.4% (128/140), the positive predictive value was 100% (120/120), and the negative predictive value was 40.0% (8/20). The area under curve of receiver operating characteristic for FNAC, detection of BRAFV600E gene, and FNAC combined with BRAFV600E gene detection were 0.818, 0.924, and 0.955, respectively. Conclusion FNAC combine with BRAFV600E gene detection improves the accuracy of neck lymph node metastasis in patients with thyroid cancer, which is worthyof performed.
Objective To systematically review the performance of postpartum hemorrhage risk prediction models, and to provide references for the future construction and application of effective prediction models. Methods The CNKI, WanFang Data, VIP, CBM, PubMed, EMbase, The Cochrane Library, Web of Science, and CINAHL databases were electronically searched to identify studies reporting risk prediction models for postpartum hemorrhage from database inception to March 20th, 2022. Two reviewers independently screened the literature, extracted data, and assessed the risk of bias and applicability of the included studies. Results A total of 39 studies containing 58 postpartum hemorrhage risk prediction models were enrolled. The area under the curve of 49 models was over 0.7. All but one of the models had a high risk of bias. Conclusion Models for predicting postpartum hemorrhage risk have good predictive performance. Given the lack of internal and external validation, and the differences in study subjects and outcome indicators, the clinical value of the models needs to be further verified. Prospective cohort studies should be conducted using uniform predictor assessment methods and outcome indicators to develop effective prediction models that can be applied to a wider range of populations.