RESUMO
BACKGROUND: Current postpartum hemorrhage (PPH) risk stratification is based on traditional statistical models or expert opinion. Machine learning could optimize PPH prediction by allowing for more complex modeling. OBJECTIVE: We sought to improve PPH prediction and compare machine learning and traditional statistical methods. METHODS: We developed models using the Consortium for Safe Labor data set (2002-2008) from 12 US hospitals. The primary outcome was a transfusion of blood products or PPH (estimated blood loss of ≥1000 mL). The secondary outcome was a transfusion of any blood product. Fifty antepartum and intrapartum characteristics and hospital characteristics were included. Logistic regression, support vector machines, multilayer perceptron, random forest, and gradient boosting (GB) were used to generate prediction models. The area under the receiver operating characteristic curve (ROC-AUC) and area under the precision/recall curve (PR-AUC) were used to compare performance. RESULTS: Among 228,438 births, 5760 (3.1%) women had a postpartum hemorrhage, 5170 (2.8%) had a transfusion, and 10,344 (5.6%) met the criteria for the transfusion-PPH composite. Models predicting the transfusion-PPH composite using antepartum and intrapartum features had the best positive predictive values, with the GB machine learning model performing best overall (ROC-AUC=0.833, 95% CI 0.828-0.838; PR-AUC=0.210, 95% CI 0.201-0.220). The most predictive features in the GB model predicting the transfusion-PPH composite were the mode of delivery, oxytocin incremental dose for labor (mU/minute), intrapartum tocolytic use, presence of anesthesia nurse, and hospital type. CONCLUSIONS: Machine learning offers higher discriminability than logistic regression in predicting PPH. The Consortium for Safe Labor data set may not be optimal for analyzing risk due to strong subgroup effects, which decreases accuracy and limits generalizability.
RESUMO
IMPORTANCE: There is a paucity of evidence-based, physician-authored content available on social media. Data are lacking on physicians use of social media, including intended audience and content. OBJECTIVE: The aim of this study was to explore the patterns of Twitter and Instagram use for popular urogynecology hashtags between physicians, patients, and allied health professionals (AHPs). STUDY DESIGN: Twelve hashtags derived from the Urogynecology Tag Ontology project were used as search terms to select Twitter and Instagram posts. Up to 5 top posts per hashtag per author type (physician, patient, or AHP) were included. Posts were analyzed using Dedoose qualitative analytic software by author, hashtag, intended audience, and themes. RESULTS: On Twitter, 109 posts met inclusion criteria: 41% written by physicians, 40% patients, and 18.3% AHPs. For Instagram, 72 posts were included: 50% written by patients, 39% AHPs, and 11% physicians. Twitter physician posts were mainly intended for health professionals (64%) with only 18% for patients. Patients posted to the general public (57%) and patients (36%). Instagram physician posts were intended for health professionals (49%), whereas 62% of AHPs posted to patients. Most patient posts were directed to other patients (90%). Physicians posted about academic peer discussions, medical education, and advocacy. Patients posted about personal experiences, treatments, or dissatisfaction. CONCLUSIONS: Physicians are more likely to post on Twitter than Instagram, with content focused primarily on their peer group, and physicians/patients are unlikely to engage with each other. There is an opportunity to improve social media interactions between physicians and the public while increasing high-quality patient education.
