Bech P
Department P of Psychiatry, Frederiksborg Amts Centralsygehus, Hillerød, Denmark.
Acta Psychiatr Scand. 1991 Nov;84(5):439-45. doi: 10.1111/j.1600-0447.1991.tb03175.x.
In 1960 Ottosson published the first specific depression rating scale (the Cronholm-Ottosson Depression Scale) designed to be sensitive in measuring change during antidepressive therapy. Ottosson and his group have never used factor analysis to validate the scale, as the items of the scale were used factor analysis to validate the scale, as the items of the scale were selected on a preconceived idea for homogeneity, i.e. having a monotonic correspondence to the underlying dimension of severity of depression. The most appropriate method of testing the construct validity of the Cronholm-Ottosson Depression Scale is latent structure analysis. Using the original Ottosson data, a latent structure analysis has been made showing that the 8 items of the scale are homogeneously related, i.e. can be ordered on one dimension of severity of depression. The descriptive statistic of a 50% reduction of pretreatment score (or more) equaled the global clinical score of moderate to excellent improvement. Both scales showed that, already after 4 electroconvulsive treatments, about 80% of depressed patients who received adequate fits had moderate to excellent improvement, whereas only around 40% of depressed patients who received inadequate fits improved moderately or excellently.
1960年，奥托松发表了首个专门用于评估抑郁症的量表（克朗霍尔姆 - 奥托松抑郁量表），该量表旨在能够敏锐地测量抗抑郁治疗过程中的变化。奥托松及其团队从未使用因素分析来验证该量表，因为量表的项目是基于同质性的先入之见而选择的，即与抑郁严重程度的潜在维度具有单调对应关系。检验克朗霍尔姆 - 奥托松抑郁量表结构效度的最合适方法是潜在结构分析。利用奥托松的原始数据进行的潜在结构分析表明，该量表的8个项目具有同质相关性，即在抑郁严重程度的一个维度上可以排序。预处理分数降低50%（或更多）的描述性统计结果等同于中度至极佳改善的整体临床评分。两个量表均显示，在仅接受4次电休克治疗后，约80%接受适当治疗的抑郁症患者有中度至极佳改善，而仅约40%接受不适当治疗的抑郁症患者有中度或极佳改善。
