Rush Augustus John, South Charles, Jain Shailesh, Agha Raafae, Zhang Mingxu, Shrestha Shristi, Khan Zershana, Hassan Mudasar, Trivedi Madhukar H
Department of Psychiatry, Texas Tech University Health Science Center, Midland, TX, USA.
Department of Psychiatry and Behavioral Sciences, Duke University School of Medicine, Durham, NC, USA.
Neuropsychiatr Dis Treat. 2021 Jul 14;17:2333-2345. doi: 10.2147/NDT.S305331. eCollection 2021.
To develop clinically meaningful improvement thresholds in both the 17-item and the 6-item Hamilton Rating Scale for Depression (HRSD) total scores in depressed outpatients.
The post-hoc analysis included all adult outpatients with non-psychotic major depressive disorder in the STARD trial who entered and exited the first treatment step (up to 14 weeks of citalopram) with a complete set of study measures at baseline and exit and at least one post-baseline measure. Within-patient change and linear regression anchor-based analyses were conducted to define meaningful and substantial changes in the HRSD and HRSD using three patient-reported outcomes [Work and Social Adjustment Scale (WSAS), Quality of Life Enjoyment and Satisfaction-Short Form (Q-LES-Q-SF); Mini-Q-LES-Q] obtained at baseline and exit from the first treatment step in STARD.
Linear regression analyses identified a meaningful change threshold for the HRSD as 3.9 [3.7-4.1] [lower, upper 95% CI] and a substantial change as 7.8 [7.4-8.3] with the WSAS. Analogous thresholds based on the Q-LES-Q-SF were 5.8 [5.5-6.1] and 11.6 [11.0-12.2], respectively, and 4.9 [4.7-5.2] and 9.9 [9.3-10.4] for the Mini-QLES-Q, respectively. For the HRSD, linear regression analyses with the WSAS identified a meaningful change as 2.2 [2.1-2.4], while a substantial change was 4.5 [4.2-4.7]. Analogous figures based on the Q-LES-Q-SF were 3.2 [3.0-3.4] and 6.4 [6.1-6.8]. Similarly, based on the Mini-QLESQ, results were 2.8 [2.6-2.9] and 5.6 [5.3-5.9]. For both the HRSD and the HRSD, within-patient analyses produced less precise estimates of the same change thresholds with substantial overlap between groups. Based on the WSAS, a clinically meaningful change in the HRSD total score was 9.6 (SD = 6.5), while a substantial change was 15.0 (SD = 6.7). Analogous change thresholds based on the Q-LESQ-SF were 12.9 (SD = 6.2) and 16.8 (SD = 6.4), respectively. For the Mini-Q-LES-Q, thresholds were 10.9 (SD = 6.5) and 16.1 (SD = 6.2).
A 4-6 point change in the HRSD is clinically meaningful; a 7-12 point change is clinically substantial. For the HRSD, analogous estimates were 2-3 and 4-7 point changes, respectively.
制定门诊抑郁症患者17项及6项汉密尔顿抑郁量表（HRSD）总分具有临床意义的改善阈值。
事后分析纳入了STARD试验中所有患有非精神病性重度抑郁症的成年门诊患者，这些患者进入并完成了首个治疗阶段（长达14周的西酞普兰治疗），在基线期、治疗结束时均有全套研究测量指标，且至少有一项基线后测量指标。采用患者内变化分析和基于线性回归锚定的分析方法，利用STARD试验首个治疗阶段基线期和结束时获得的三项患者报告结局[工作与社会适应量表（WSAS）、生活质量享受与满意度简表（Q-LES-Q-SF）；简易Q-LES-Q]来定义HRSD及HRSD中有意义和显著的变化。
线性回归分析确定，基于WSAS，HRSD有意义的变化阈值为3.9[3.7 - 4.1][下限，上限95%CI]，显著变化阈值为7.8[7.4 - 8.3]。基于Q-LES-Q-SF的类似阈值分别为5.8[5.5 - 6.1]和11.6[11.0 - 12.2]，基于简易Q-LES-Q的分别为4.9[4.7 - 5.2]和9.9[9.3 - 10.4]。对于HRSD，基于WSAS的线性回归分析确定有意义的变化为2.2[2.1 - 2.4]，显著变化为4.5[4.2 - 4.7]。基于Q-LES-Q-SF的类似数值分别为3.2[3.0 - 3.4]和6.4[6.1 - 6.8]。同样，基于简易Q-LES-Q的结果分别为2.8[2.6 - 2.9]和5.6[5.3 - 5.9]。对于HRSD和HRSD，患者内分析对相同变化阈值的估计不太精确，组间有大量重叠。基于WSAS，HRSD总分有临床意义的变化为9.6（标准差 = 6.5），显著变化为15.0（标准差 = 6.7）。基于Q-LES-Q-SF的类似变化阈值分别为12.9（标准差 = 6.2）和16.8（标准差 = 6.4）。对于简易Q-LES-Q，阈值分别为10.9（标准差 = 6.5）和16.1（标准差 = 6.2）。
HRSD总分变化4 - 6分具有临床意义；变化7 - 12分具有临床显著性。对于HRSD，类似的估计分别为变化2 - 3分和4 - 7分。
