Wolpert Miranda, Görzig Anke, Deighton Jessica, Fugard Andrew J B, Newman Robbie, Ford Tamsin
Evidence Based Practice Unit (EBPU), UCL and the Anna Freud Centre, 21 Maresfield Gardens, London, NW3 5SD, UK.
University of West London, London, UK.
Child Adolesc Ment Health. 2015 May;20(2):94-101. doi: 10.1111/camh.12080. Epub 2014 Nov 24.
Establishing what constitutes clinically significant change is important both for reviewing the function of services and for reflecting on individual clinical practice. A range of methods for assessing change exist, but it remains unclear which are best to use and under which circumstances.
This paper reviews four indices of change [difference scores (DS), crossing clinical threshold (CCT), reliable change index (RCI) and added value scores (AVS)] drawing on outcome data for 9764 young people from child and adolescent mental health services across England.
Looking at DS, the t-test for time one to time two scores indicated a significant difference between baseline and follow up scores, with a standardised effect size of d = 0.40. AVS analysis resulted in a smaller effect size of 0.12. Analysis of those crossing the clinical threshold showed 21.2% of cases were classified as recovered, while 5.5% were classified as deteriorated. RCI identified 16.5% of cases as showing reliable improvement and 2.3% of cases as showing reliable deterioration. Across RCI and CCT 80.5% of the pairings were exact (i.e., identified in the same category using each method).
Findings indicate that the level of agreement across approaches is at least moderate; however, the estimated extent of change varied to some extent based on the index used. Each index may be appropriate for different contexts: CCT and RCI may be best suited to use for individual case review; whereas DS and AVS may be more appropriate for case-mix adjusted national reporting.
确定具有临床意义的变化构成因素,对于评估服务功能和反思个体临床实践都很重要。存在一系列评估变化的方法,但哪种方法最适用以及在何种情况下适用仍不明确。
本文利用来自英格兰儿童和青少年心理健康服务机构的9764名年轻人的结果数据,回顾了四种变化指标[差异分数(DS)、跨越临床阈值(CCT)、可靠变化指数(RCI)和增加值分数(AVS)]。
观察差异分数,从时间1到时间2分数的t检验表明基线分数和随访分数之间存在显著差异,标准化效应大小d = 0.40。增加值分数分析得出的效应大小较小,为0.12。对跨越临床阈值者的分析显示,21.2%的病例被归类为康复,而5.5%的病例被归类为恶化。可靠变化指数确定16.5%的病例显示出可靠的改善,2.3%的病例显示出可靠的恶化。在可靠变化指数和跨越临床阈值这两种方法中,80.5%的配对结果一致(即使用每种方法都被归为同一类别)。
研究结果表明,不同方法之间的一致性水平至少为中等;然而,根据所使用的指标,估计的变化程度在一定程度上有所不同。每个指标可能适用于不同的背景:跨越临床阈值和可靠变化指数可能最适合用于个体病例审查;而差异分数和增加值分数可能更适合用于病例组合调整后的全国报告。