{"doi":"10.1145/3721201.3721390","title":"State-Specific Explainable Machine Learning for Predicting Premature Dropout in Medication for Opioid Use Disorder","abstract":"Opioid use disorder (OUD) presents a pressing global public health challenge, with Medication for Opioid Use Disorder (MOUD) serving as an effective treatment. However, the success of MOUD is largely dependent on patient adherence, as premature dropout increases the risks of relapse and worsens health outcomes. Motivated by large variations in treatment outcomes across different states in U.S., this study proposes a novel, explainable machine learning (ML) framework to predict premature dropout from MOUD, enabling state-specific customization through multi-task learning. The proposed multi-task learning model integrates shared layers to capture general dropout patterns and state-specific layers to highlight regional differences. To enhance interpretability, we employ Shapley Additive Explanations (SHAP) to identify key predictive features for each state and use Local Interpretable Model-Agnostic Explanations (LIME) to offer individualized insights into adherence factors. Evaluations on real-world datasets demonstrate superior performance compared to baseline models, validating the framework's ability to address both general and state-specific factors. This work underscores the need for targeted interventions and policies to improve MOUD retention and tackle the opioid crisis across diverse states.","journal":null,"year":2025,"id":580882,"datarank":0.0,"base_score":0.0,"endowment":0.0,"self_citation_contribution":0.0,"citation_network_contribution":0.0,"self_endowment_contribution":0.0,"citer_contribution":0.0,"corpus_percentile":null,"corpus_rank":null,"citation_count":0,"citer_count":0,"citers_with_citation_signal":0,"citers_with_endowment":0,"datacite_reuse_total":0,"is_dataset":false,"is_dataset_confidence":0.9525,"is_data_producer":false,"deposit_databanks":null,"is_oa":true,"file_count":0,"downloads":0,"has_version_chain":false,"published_date":"2025-01-01","fair_score":null,"fair_percentile":null,"algorithm_id":"datarank_citation_only_1hop_v6","ranking_scope":"data_only","authors":[{"id":1491954,"name":"Tongnian Wang","orcid":"0000-0003-3928-1754","position":1,"is_corresponding":false},{"id":1491955,"name":"Yuanxiong Guo","orcid":"0000-0003-2241-125X","position":2,"is_corresponding":false},{"id":1491956,"name":"Carolina Vivas-Valencia","orcid":"0009-0003-4071-5051","position":3,"is_corresponding":false},{"id":712839,"name":"Cici Bauer","orcid":"0000-0002-2337-7965","position":4,"is_corresponding":false},{"id":1491957,"name":"Yanmin Gong","orcid":"0000-0002-1761-2834","position":5,"is_corresponding":false},{"id":1491953,"name":"Xiangxing Guo","orcid":"0009-0001-2732-0333","position":0,"is_corresponding":true}],"reference_count":15,"raw_metadata":null,"created_at":"2026-07-19T02:58:43.046112Z","pmid":null,"pmcid":null,"fwci":null,"citation_percentile":null,"influential_citations":0,"oa_status":null,"license":null,"views":0,"total_file_size_bytes":0,"version_count":0,"fair_f":null,"fair_a":null,"fair_i":null,"fair_r":null,"fair_zscore":null,"fair_rationale":null,"fair_model":null,"fair_agent_version":null,"fair_fulltext_source":null,"fair_has_llm":null,"fair_computed_at":null,"clinical_trials":[],"software_tools":[],"db_accessions":[],"linked_datasets":[],"topics":[]}