Agreement and classification performance of the heaviness of smoking index (HSI) compared with the Fagerström test for cigarette dependence (FTCD): A subgroup analysis

Background The Heaviness of Smoking Index (HSI) is a brief tool for assessing nicotine addiction. This study evaluated the agreement and classification performance of the HSI relative to the Fagerström Test for Cigarette Dependence (FTCD). Methods In this cross-sectional study of 1147 adult smokers, HSI scores were derived from the FTCD. Agreement and apparent classification performance were examined using Pearson correlation, intraclass correlation coefficient (ICC), Cohen's kappa, receiver operating characteristic (ROC) analysis, and Bland–Altman plots across demographic and behavioral subgroups. Results HSI showed a strong correlation with FTCD ( r = 0.904) and excellent continuous agreement (ICC = 0.95). Categorical agreement for high nicotine addiction was substantial (κ ≈ 0.65). ROC analysis demonstrated excellent apparent classification performance (AUC = 0.942) for identifying FTCD-defined severe addiction. An HSI cut-off of ≥3 (derived from an optimal threshold of 2.5) provided a favorable balance between sensitivity and specificity for initial screening, whereas a cut-off of ≥4 showed high specificity for severe addiction. Performance remained generally consistent across subgroups, although some subgroup-specific variation was observed. Bland–Altman analysis indicated minimal mean bias, although HSI proportionally underestimated addiction at higher severity levels. Conclusion HSI demonstrates strong agreement with FTCD and may serve as a rapid screening measure for assessing nicotine addiction. A lower cut-off (≥3) may be useful for initial screening, whereas a higher cut-off (≥4) may help identify severe addiction. Caution is warranted at higher addiction levels and among poly-tobacco users.