I congratulate ter Heide et al on conducting an extremely timely and important clinical trial. ^{Reference ter Heide, Mooren, van de Schooot, de Jongh and Kleber1} As stated by the authors, there is a conspicuous lack of clinical trials investigating the effect of eye movement desensitisation and reprocessing (EMDR) in treating traumatised refugees.

However, in order to be able to appraise/interpret the results more fully, it would be of great value if the authors could analyse their data with different cut-off points for defining clinically significant change. In their article, the authors define clinically significant change as a decrease of 10 points or more on the Clinician-Administered PTSD Scale (CAPS). ^{Reference Weathers, Keane and Davidson2} The authors report that approximately 40% of patients in both conditions achieved clinically significant improvement (Table 2).

However, there is currently no widely agreed threshold for defining clinically significant change on the CAPS and different clinical trials have used a range of different cut-off scores. For example, in two major publications Schnurr et al ^{Reference Schnurr, Friedman, Engel, Foa, Shea and Chow3,Reference Schnurr, Friedman, Foy, Shea, Hsieh and Lavori4} defined clinically significant change as a decrease of at least 10 points in total CAPS scores. Hinton et al ^{Reference Hinton, Chhean, Pich, Safren, Hofmann and Pollack5} used the rationally derived 15-point change ^{Reference Weathers, Keane and Davidson2} as a marker of clinically significant change. In line with the method set forth by Jacobson & Truax, ^{Reference Jacobson and Truax6} Taylor et al ^{Reference Taylor, Thordarson, Maxfield, Fedoroff, Lovell and Ogrodniczuk7} defined clinically significant change as a reduction in total CAPS score of at least two standard deviations. Hien et al ^{Reference Hien, Wells, Jiang, Suarez-Morales, Campbell and Cohen8} used a 30-point or greater improvement on the CAPS to determine clinically significant improvement of PTSD symptoms, whereas Bryant et al ^{Reference Bryant, Moulds, Guthrie, Dang, Mastrodomenico and Nixon9} defined clinically significant change as a cut-off of <45 on the CAPS at follow-up.

It is also important to take into consideration the measurement variability of the instrument. ^{Reference Jacobson and Truax6} A change of 10 points on the CAPS is not necessarily even reliable. Monson et al ^{Reference Monson, Schnurr, Resick, Friedman, Young-Xu and Stevens10} calculated a reliable change score of 12.22 points on the CAPS, and we previously have calculated a conservative reliable change score of 14.3 points. ^{Reference Halvorsen and Stenmark11} So a change of 10 points might simply be within the measurement variability of the CAPS.

These issues taken together, it would be very informative if the authors made use of different cut-off scores for defining clinically significant change. The Institute of Medicine ¹² notes that common methods to define clinically significant change on the CAPS in the treatment literature are to define it as a ➮10-point decrease, a ➮30 percent decrease, or as two standard deviations below pre-treatment level. Using a 30% decrease in total CAPS score for the present sample entails a cut-off score for clinically significant change of approximately 23 points. Using two standard deviations below pre-treatment level ^{Reference Jacobson and Truax6} as a marker for clinically significant change entails a cut-off score of approximately 36 points. It would be very informative if the authors re-analysed their data with at least these two thresholds for defining clinically significant change in addition to the 10-point cut-off score reported in their paper.