Skip to main content

Table 3 Treatment outcome variables at pre-treatment, post-treatment and 6-month follow-up

From: Emotion regulation individual therapy for adolescents with nonsuicidal self-injury disorder: a feasibility study

Outcome Pre-treatment Post-treatment 6-mo f-u Pre-to post-treatment comparison Post- to 6-mo follow-up comparison Pre-to 6-mo follow-up comparison
Count- data Median (IQR) Median (IQR) Median (IQR) Z Percentage change [95% CI] Z Percentage change [95% CI] Z Percentage change [95% CI]
DSHI-9-f 8 (4 to 13) 2.5 (0 to 4.25) 0 (0 to 3) 2.33* 42 [8 to 64] 3.82*** 63 [39 to 78] 4.77*** 79 [60 to 89]
BSL 3 (2 to 3) 1 (0 to 3) 1 (0 to 2.75) 2.16* 44 [5 to 67] 0.35 13 [−91 to 60] 1.99* 52 [1 to 76]
Continuous Mean (SD) Mean (SD) Mean (SD) Z Cohen’s d [95% CI] Z Cohen’s d [95% CI] Z Cohen’s d [95% CI]
DSHI-9-v 2.12 (1.54) 1.35 (1.22) 0.80 (1.01) 2.02* 0.50 [0.07 to 0.78] 2.24* 0.40 [0.09 to 0.80] 2.61** 0.89 [0.47 to 1.23]
DERS 122.71 (17.18) 109.69 (25.21) 100.57 (17.89) 2.91** 0.81 [0.23 to 1.50] 2.66** 0.59 [0.15 to 1.05] 4.44*** 1.40 [0.87 to 2.03]
BPFSC 70.35 (10.72) 66.88 (12.61) 63.36 (11.34) 1.45 0.33 [−0.03 to 1.11] 1.47 0.36 [−0.16 to 0.87] 2.73** 0.69 [0.18 to 1.19]
CGAS 51.24 (4.94) 56.59 (7.97) 61.07 (11.21) 3.7*** 1.08 [0.51 to 1.62] 1.72 0.85 [−0.22 to 2.13] 4.74*** 1.94 [1.03 to 3.12]
  1. Note. Count data. Test statistics are based on generalized estimation equation models using either a negative binomial or normal distribution with log link function for count and continuous data, respectively. Effect sizes for count data were based on log-transformed scores. Confidence intervals for effect sizes are based on 5000 bootstrap replications. Abbreviations: BPFSC = Borderline Personality Feature Scale for Children, BSL = Borderline Symptom List, CGAS = Children’s Global Assessment Scale, DERS = Difficulties in Emotion Regulation Scale, DSHI-9-f = Deliberate Self-Harm Inventory – frequency past month, DSHI-9-v = Deliberate Self-Harm Inventory – versatility past month. *p < .05, **p < .01, ***p < .001