Predicting Outpatient Follow‐Up Retention After Inpatient Treatment in Patients With Alcohol Use Disorder: A Data‐Driven Random Forest Approach
Addiction Biology · 2026-06 · 31 (6)
DOI: 10.1111/adb.70169 ↗Kurz gefasst
Fachartikel in der Zeitschrift Addiction Biology, online erschienen am 12.06.2026. Verfasst von Jennifer J Barb und 8 weiteren Autor:innen. Beteiligte Einrichtung: Translational Biobehavioral and Health Promotion Branch, Clinical Center, National Institutes of Health, Bethesda, Maryland, USA. Hauptthemen laut MEDLINE: Alcoholism, Ambulatory Care. Die Arbeit nennt 37 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: National Institutes of Health (NIH), NIAAA DICBR Office of the Clinical Director (OCD), NIH Clinical Center.
Themen: Sucht und Abhängigkeit
Abstract (Originalsprache)
Longer treatment engagement is associated with improved recovery outcomes in alcohol use disorder (AUD), making patient retention a critical determinant of reduced return to drinking. This study aimed to identify factors predicting outpatient treatment engagement, operationalized as the number of follow‐up visits, among individuals with AUD following inpatient care. We applied a five‐step analytic framework integrating random forest modelling (RFM) and Least Absolute Shrinkage and Selection Operator (LASSO) regression to identify predictors of outpatient visit frequency. Clinical, psychological and physiological variables ( n = 177 per participant) collected during inpatient treatment prior to discharge were included. RFM ranked variables associated with follow‐up visits, with LASSO used for validation and complementary selection. Over a 5‐year period, 119 treatment‐seeking patients with AUD (mean age = 45.8) returned for outpatient care, averaging 5.14 visits, following a medical treatment inpatient stay. Positive urgency (VIP = 44.31) and positive life events (VIP = 41.19) emerged as the strongest predictors; both inversely associated with visit frequency. LASSO confirmed positive urgency as a significant predictor (coefficient: −0.03296). Greater alcohol use severity and higher haemoglobin levels were also associated with fewer outpatient visits, whereas higher depressive symptom severity predicted increased follow‐up engagement. Using complementary machine learning and regression approaches, this study identified affective traits, alcohol use severity and physiological factors as key determinants of outpatient engagement following inpatient AUD treatment. Interestingly, positive urgency and positive life events, often considered markers of recovery or resilience, were linked to reduced outpatient attendance, suggesting that certain personality or motivational factors may diminish perceived need for continued care. These findings highlight the importance of integrating psychological and motivational variables into postdischarge planning to enhance retention and improve early recovery outcomes.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Translational Biobehavioral and Health Promotion Branch, Clinical Center, National Institutes of Health, Bethesda, Maryland, USA.
- Division of Intramural Clinical and Biological Research, National Institute on Alcohol Abuse and Alcoholism, Bethesda, Maryland, USA.
- Clinical Psychoneuroendocrinology and Neuropsychopharmacology Section, Translational Addiction Medicine Branch, National Institute on Drug Abuse Intramural Research Program and National Institute on Alcohol Abuse and Al…
- Center for Alcohol and Addiction Studies, Department of Behavioral and Social Sciences, Brown University, Providence, Rhode Island, USA.
- Division of Addiction Medicine, Department of Medicine, School of Medicine, Johns Hopkins University, Baltimore, Maryland, USA.
- Department of Neuroscience, Georgetown University Medical Center, Washington, DC, USA.
ORCID-Profile: Jennifer J Barb, Mehdi Farokhnia, Lorenzo Leggio
Förderung
- National Institutes of Health (NIH)
- NIAAA DICBR Office of the Clinical Director (OCD)
- NIH Clinical Center
Zitieren
APA
Barb, J. J., King, L. C., Hughes, A. N., Yarmovsky, J., Diazgranados, N., Farokhnia, M., Wallen, G. R., Schwandt, M., & Leggio, L. (2026). Predicting Outpatient Follow‐Up Retention After Inpatient Treatment in Patients With Alcohol Use Disorder: A Data‐Driven Random Forest Approach. Addiction Biology, 31(6). https://doi.org/10.1111/adb.70169
Vancouver
Barb JJ, King LC, Hughes AN, Yarmovsky J, Diazgranados N, Farokhnia M, et al. Predicting Outpatient Follow‐Up Retention After Inpatient Treatment in Patients With Alcohol Use Disorder: A Data‐Driven Random Forest Approach. Addiction Biology. 2026;31(6). doi:10.1111/adb.70169
BibTeX
@article{Barb2026,
author = {Barb, Jennifer J and King, Lillian C and Hughes, Alexandria N and Yarmovsky, Jenny and Diazgranados, Nancy and Farokhnia, Mehdi and Wallen, Gwenyth R and Schwandt, Melanie and Leggio, Lorenzo},
title = {Predicting Outpatient Follow‐Up Retention After Inpatient Treatment in Patients With Alcohol Use Disorder: A Data‐Driven Random Forest Approach},
journal = {Addiction Biology},
year = {2026},
volume = {31},
number = {6},
doi = {10.1111/adb.70169}
}
EUVIMED