To prevent the overdiagnosis and overtreatment of prostate cancer (PC), therapeutic strategies have been established such as active surveillance and focal therapy, as well as methods for clarifying the diagnosis of high-grade prostate cancer (HGPC) (defined as a Gleason score ≥7), such as multiparametric magnetic resonance imaging and new markers such as the 4Kscore test (4KsT). By means of a pilot study, we aim to test the ability of the 4KsT to identify HGPC in prostate biopsies (Bx) and compare the test with other multivariate prognostic models such as the Prostate Cancer Prevention Trial Risk Calculator 2.0 (PCPTRC 2.0) and the European Research Screening Prostate Cancer Risk Calculator 4 (ERSPC-RC 4). Fifty-one patients underwent a prostate Bx according to standard clinical practice, with a minimum of 10 cores. The diagnosis of HGPC was agreed upon by 4 uropathologists. We compared the predictions from the various models by using the Mann-Whitney U test, area under the ROC curve (AUC) (DeLong test), probability density function (PDF), box plots and clinical utility curves. Forty-three percent of the patients had PC, and 23.5% had HGPC. The medians of probability for the 4KsT, PCPTRC 2.0 and ERSPC-RC 4 were significantly different between the patients with HGPC and those without HGPC (p≤.022) and were more differentiated in the case of 4KsT (51.5% for HGPC [25-75 percentile: 25-80.5%] vs. 16% [P 25-75: 8-26.5%] for non-HGPC; p=.002). All models presented AUCs above 0.7, with no significant differences between any of them and 4KsT (p≥.20). The PDF and box plots showed good discriminative ability, especially in the ERSPC-RC 4 and 4KsT models. The utility curves showed how a cutoff of 9% for 4KsT identified all cases of HGPC and provided a 22% savings in biopsies, which is similar to what occurs with the ERSPC-RC 4 models and a cutoff of 3%. The assessed predictive models offer good discriminative ability for HGPCs in Bx. The 4KsT is a good classification model as a whole, followed by ERSPC-RC 4 and PCPTRC 2.0. The clinical utility curves help suggest cutoff points for clinical decisions: 9% for 4KsT and 3% for ERSPC-RC 4. This preliminary study should be interpreted with caution due to its limited sample size.

Objective: The objective of this study is to analyze the usefulness of thrombophilia and antithrombotic drugs in combination with materno-fetal characteristics to generate a predictive model of placenta-mediated pregnancy complications (PMPC) for counselling treatment. Methods: A retrospective analysis was performed in women with singleton pregnancy that required a thrombophilia study, including 222 patients with unknown cause PMPC and 151 women with no complications at current pregnancy in Hospital Clínico Universitario, Lozano, Blesa, Zaragoza, Spain. Chi-squared and Mann-Whitney test were applied to analyze univariate risk factors. Multivariate analysis was performed using logistic regression model with candidate variables: maternal characteristics, obstetric history, thrombophilia and treatment with LMWH and/or with ASA. The calibration, discrimination, and best cut-off point for the clinical application of the model was analyzed. Results: Maternal characteristics showed differences in median body mass index (BMI), OR: 0.4, smoking habit, OR: 8.5, and hypertension, OR: 11.4, appearing all of them as risk factors. In our study, a prior pregnancy that ended in a child alive was a protective factor OR: 0.02-.0.4, and having a previous preterm child was a strong risk factor OR: 4.2. Thrombophilia was not a risk factor. Patients under LMWH treatment (15%) and/or ASA (6.2%) had better pregnancy outcomes, showing both as protective factors: ASA OR: 0.32 and LMWH OR: 0.16. The model has an AUC value of 0.847, with good calibration. A nomogram and an app is provided for this adjusted model with high discrimination ability in internal validation (AUC = 0.833). Our clinical utility analysis guide us to choose 40% as the best threshold probability. Conclusions: We found risk and protective factors associated with PMPC, but our data were not conclusive to demonstrate its relation with maternal thrombophilia. However, the challenger finding is the clinical utility of antithrombotic drugs as a protective factors in PMPC prevention. It is possible to identify patients with high risk of PMPC through a combined predictive model, for counselling treatment.

Electronic fetal monitoring (EFM) is widely used in intrapartum care as the standard method for monitoring fetal well-being. Our objective was to employ machine learning algorithms to predict acidemia by analyzing specific features extracted from the fetal heart signal within a 30 min window, with a focus on the last deceleration occurring closest to delivery. To achieve this, we conducted a case–control study involving 502 infants born at Miguel Servet University Hospital in Spain, maintaining a 1:1 ratio between cases and controls. Neonatal acidemia was defined as a pH level below 7.10 in the umbilical arterial blood. We constructed logistic regression, classification trees, random forest, and neural network models by combining EFM features to predict acidemia. Model validation included assessments of discrimination, calibration, and clinical utility. Our findings revealed that the random forest model achieved the highest area under the receiver characteristic curve (AUC) of 0.971, but logistic regression had the best specificity, 0.879, for a sensitivity of 0.95. In terms of clinical utility, implementing a cutoff point of 31% in the logistic regression model would prevent unnecessary cesarean sections in 51% of cases while missing only 5% of acidotic cases. By combining the extracted variables from EFM recordings, we provide a practical tool to assist in avoiding unnecessary cesarean sections.