Prediction of therapeutic dropout in patients with addictions: Development and validation of the Predictors of Dropout from Addiction Treatment (PDAT) scale

by Carlos Miguel Sirvent-Ruiz, María Miranda, María de la Villa Moral-Jiménez

Background

Withdrawal from addiction treatment is a frequent but difficult-to-predict contingency. We clarify and contextualize the concept of dropouts in addiction treatment, as well as the external and internal elements that most frequently lead to such dropouts. The main instruments used to measure dropout are summarized, after which a new tool, Predictors of Dropout from Addiction Treatment (PDAT) scale, is presented. The PDAT consists of four factors: 1) Motivation: desire to recover and to actively engage in current treatment; 2) Craving: longing for the use of substances and/or the substance addiction environment; 3) Problem awareness: level of insight, or degree of knowledge, and ability to objectify the problem and the disease, with the renunciations and limitations that this entails; and 4) Dysphoria: dyade inner restlessness – moodiness, i.e., emotional disturbance and depressive anticipation that precedes treatment withdrawal.

Methods

The sample consisted of 243 addicted subjects in residential treatment, ranging in age from 18 to 63 years (average = 38.43, standard deviation = 10.95), who completed an initial 26-item PDAT questionnaire. The factor structure of the PDAT was determined by factor analysis. Mixed effects logistic regressions and receiver operating characteristics curve (ROC) analyses were applied to assess the predictive validity of the PDAT. Results: The 13-item PDAT showed adequate reliability and convergent and discriminant validity, with both the general scale and each of its factors having predictive validity 7 and 15 days after administration.

Conclusion

The scale is a useful instrument with proven clinical efficacy and brevity of application. In addition, its four factors are useful for targeting interventions based on the unbalanced factors.

Sex prediction through machine learning utilizing mandibular condyles, coronoid processes, and sigmoid notches features

by Isabela Bittencourt Basso, Pedro Felipe de Jesus Freitas, Aline Xavier Ferraz, Ana Julia Borkovski, Ana Laura Borkovski, Rosane Sampaio Santos, Rodrigo Nunes Rached, Erika Calvano Küchler, Angela Graciela Deliga Schroder, Cristiano Miranda de Araujo, Odilon Guariza-Filho

Characteristics of the mandible structures have been relevant in anthropological and forensic studies for sex prediction. This study aims to evaluate the coronoid process, condyle, and sigmoid notch patterns in sex prediction through supervised machine learning algorithms. Cephalometric radiographs from 410 dental records of patients were screened to investigate the morphology of the coronoid process, condyle, and sigmoid notch and the Co-Gn distance. The following machine learning algorithms were used to build the predictive models: Decision Tree, Gradient Boosting Classifier, K-Nearest Neighbors (KNN), Logistic Regression, Multilayer Perceptron Classifier, Random Forest Classifier, and Support Vector Machine (SVM). A 5-fold cross-validation approach was adopted to validate each model. Metrics such as area under the curve (AUC), accuracy, recall, precision, and F1 Score were calculated for each model, and ROC curves were constructed. All tested variables demonstrated statistical significance (p