Prediction of Positive Patient Health Questionnaire-2 Screening Using Area Deprivation Index in Primary Care.

Depression is recognized as a significant public health issue in the United States. The National Survey on Drug Use and Health reports that 21.0 million adults aged 18 or older had major depressive disorder in 2020, including 14.8 million experiencing a major depressive episode with severe impairment. The aim is to predict the positivity of Patient Health Questionnaire-2 (PHQ-2) outcomes among patients in primary care settings by analyzing a range of variables, including socioeconomic status, demographic characteristics, and health behaviors, thereby identifying those at increased risk for depression. Employing a machine learning approach, the study utilizes retrospective data from electronic health records across 15 primary care clinics in South Florida to explore the relationship between social determinants of health (SDoH), including area of deprivation index (ADI) and PHQ-2 positivity. The study encompasses 15 primary care clinics located in South Florida, where a diverse patient population receives care. Analysis included 94,572 patient visits; 74,636 records were included in the study. If a zip+4 was not available or an ADI score did not exist, the visit was not included in the final analysis. Screening involved the PHQ-2, assessing depressed mood and anhedonia, with a cutoff >2 indicating positive screening. ADI was used to assess SDoH by matching patients' residential postal codes to ADI national percentiles. Demographics, sexual history, tobacco use, caffeine intake, and community involvement were also evaluated in the study. Over 40 machine learning algorithms were explored for their accuracy in predicting PHQ-2 outcomes, using software tools including Scikit-learn and stats models in Python. Variables were normalized, scored, and then subjected to predictive regression models, with Random Forest showing outstanding performance. Feature engineering and correlation analysis identified ADI, age, education, visit type, coffee intake, and marital status as significant predictors of PHQ-2 positivity. The area under the curve and model accuracies varied across clinics, with specific clinics showing higher predictive accuracy and others (p > .05). The study concludes that the ADI, as a proxy for SDoH, alongside other individual factors, can predict PHQ-2 positivity. Health organizations can use this information to anticipate health needs and resource allocation.

Prediction of patient admission and readmission in adults from a Colombian cohort with bipolar disorder using artificial intelligence.

Introduction: Bipolar disorder (BD) is a chronically progressive mental condition, associated with a reduced quality of life and greater disability. Patient admissions are preventable events with a considerable impact on global functioning and social adjustment. While machine learning (ML) approaches have proven prediction ability in other diseases, little is known about their utility to predict patient admissions in this pathology. Aim: To develop prediction models for hospital admission/readmission within 5 years of diagnosis in patients with BD using ML techniques. Methods: The study utilized data from patients diagnosed with BD in a major healthcare organization in Colombia. Candidate predictors were selected from Electronic Health Records (EHRs) and included sociodemographic and clinical variables. ML algorithms, including Decision Trees, Random Forests, Logistic Regressions, and Support Vector Machines, were used to predict patient admission or readmission. Survival models, including a penalized Cox Model and Random Survival Forest, were used to predict time to admission and first readmission. Model performance was evaluated using accuracy, precision, recall, F1 score, area under the receiver operating characteristic curve (AUC) and concordance index. Results: The admission dataset included 2,726 BD patients, with 354 admissions, while the readmission dataset included 352 patients, with almost half being readmitted. The best-performing model for predicting admission was the Random Forest, with an accuracy score of 0.951 and an AUC of 0.98. The variables with the greatest predictive power in the Recursive Feature Elimination (RFE) importance analysis were the number of psychiatric emergency visits, the number of outpatient follow-up appointments and age. Survival models showed similar results, with the Random Survival Forest performing best, achieving an AUC of 0.95. However, the prediction models for patient readmission had poorer performance, with the Random Forest model being again the best performer but with an AUC below 0.70. Conclusion: ML models, particularly the Random Forest model, outperformed traditional statistical techniques for admission prediction. However, readmission prediction models had poorer performance. This study demonstrates the potential of ML techniques in improving prediction accuracy for BD patient admissions.