RT Journal Article
T1 Analyzing Dropout in Alcohol Recovery Programs: A Machine Learning Approach
A1 Collin, Adele
A1 Díaz Batanero, María Carmen
A1 Fernández Calderón, Fermín
A1 Lozano Rojas, Óscar Martín
A1 Rodríguez González, Alejandro
AB Background: Retention in treatment is crucial for the success of interventions targetingalcohol use disorder (AUD), which affects over 100 million people globally. Most previous studieshave used classical statistical techniques to predict treatment dropout, and their results remaininconclusive. This study aimed to use novel machine learning tools to identify models that predictdropout with greater precision, enabling the development of better retention strategies for thoseat higher risk. Methods: A retrospective observational study of 39,030 (17.3% female) participantsenrolled in outpatient-based treatment for alcohol use disorder in a state-wide public treatmentnetwork has been used. Participants were recruited between 1 January 2015 and 31 December 2019.We applied different machine learning algorithms to create models that allow one to predict thepremature cessation of treatment (dropout). With the objective of increasing the explainability ofthose models with the best precision, considered as black-box models, explainability techniqueanalyses were also applied. Results: Considering as the best models those obtained with one ofthe so-called black-box models (support vector classifier (SVC)), the results from the best model,from the explainability perspective, showed that the variables that showed greater explanatorycapacity for treatment dropout are previous drug use as well as psychiatric comorbidity. Amongthese variables, those of having undergone previous opioid substitution treatment and receivingcoordinated psychiatric care in mental health services showed the greatest capacity for predictingdropout. Conclusions: By using novel machine learning techniques on a large representative sampleof patients enrolled in alcohol use disorder treatment, we have identified several machine learningmodels that help in predicting a higher risk of treatment dropout. Previous treatment for othersubstance use disorders (SUDs) and concurrent psychiatric comorbidity were the best predictorsof dropout, and patients showing these characteristics may need more intensive or complementaryinterventions to benefit from treatment.
PB MDPI
SN 2077-0383 (electrónico)
YR 2024
FD 2024-08
LK https://hdl.handle.net/10272/24116
UL https://hdl.handle.net/10272/24116
LA eng
NO Collin, A., Ayuso-Muñoz, A., Tejera-Nevado, P., Prieto-Santamaría, L., Verdejo-García, A., Díaz-Batanero, C., Fernández-Calderón, F., Albein-Urios, N., Lozano, Ó. M., & Rodríguez-González, A. (2024). Analyzing Dropout in Alcohol Recovery Programs: A Machine Learning Approach. In Journal of Clinical Medicine (Vol. 13, Issue 16, p. 4825). MDPI AG. https://doi.org/10.3390/jcm13164825
NO This study was supported by the grant “COMPARA: Comorbilidad Psiquiátrica en Adiccionesy Resultados en Andalucía. Modelización a través de Big Data”, project P20_00735 in theAndalusian Research, Development, and Innovation Plan, provided by the Fondo Europeo de DesarrolloRegional (EU) and Junta de Andalucía (Spain).
DS Repositorio Institucional de la Universidad de Huelva
RD 1 jun 2026