İlker BOĞA, Mete Tuğcan ÜÇDAL
Bağımlılık Dergisi - 2026;27(3):116-129
Objective: Alcohol use disorder (AUD) is underdiagnosed in internal medicine. We evaluated a local large language model (LLM), fine-tuned on a Turkish medical corpus, for detecting AUD across outpatient, emergency and inpatient settings. Method: We retrospectively analyzed 42 adult patients at a tertiary state hospital (January-May 2024); the de-identified dataset followed the MIMIC-IV schema. The reference standard was the clinician-confirmed DSM-5 diagnosis; comparators were the Alcohol Use Disorders Identification Test-Consumption (AUDIT-C), gamma-glutamyl transferase (GGT), mean corpuscular volume and the AST/ALT ratio. An in-hospital Llama-3-8B-Instruct model fine-tuned via low-rank adaptation produced binary predictions, probabilities and severity categories. Analyses used Wilson 95% confidence intervals, bootstrap AUC, Cohen's kappa and paired DeLong tests. Results: Mean age was 53.0 +/- 15.0 years; 50.0% were female. AUD prevalence was 47.6% (20/42). The LLM achieved sensitivity 85.0% (95% CI, 64.0-94.8), specificity 90.9% (72.2-97.5), accuracy 88.1%, area under the curve (AUC) 0.864 (0.720-0.977) and kappa = 0.761. In all 17 true positives, severity matched the reference exactly (quadratic-weighted v = 1.000). In this enriched cohort AUDIT-C and GGT separated perfectly (AUC = 1.000) and the LLM AUC was lower (DeLong p = 0.038), a ceiling effect of this case mix, not a general marker property. Conclusion: The local LLM achieved clinically acceptable accuracy and exact severity classification within correctly flagged cases; its principal value lies where AUDIT-C is not routine. Because the cohort was enriched for AUD, predictive values and discrimination may differ substantially in routine populations. Multicenter prospective validation is warranted.
