TABLE 2.

Cannabis treatment results by clinical scales.

	n	Before treatment: mean ± SD / median (IQR)	After treatment: mean ± SD / median (IQR)	Test statistic	Effect size	Difference	95% Confidence interval		P
Pittsburgh sleep quality index (PSQI)
Total t	10	14.6 ± 4.1	10.1 ± 4.7	t = −4.4	−1.40	−4.5	−6.8	−2.2	<0.01**
Sleep quality w	14	3.0 (1.0)	1.0 (1.25)	z = −3.1	−0.59	−1.5	−2.0	−1.0	<0.01**
Sleep duration (hours) t	11	4.3 ± 2.4	5.2 ± 1	t = 3.9	1.17	0.9	0.6	2.1	<0.01**
Inability to fall asleep under 30 min w	14	3.0 (1.5)	1.5 (3.0)	z = −1.9	−0.36	−0.5	−1.5	0.0	0.058
Posttraumatic diagnostic scale (PDS)
Total symptom severityt	14	30.3 ± 8.5	22.9 ± 9.0	t = −3.3	−0.86	−7.4	−12.3	−2.5	<0.01**
Intrusiveness t	14	9.1 ± 3.5	7.1 ± 3.2	t = −2.4	−0.65	−2.0	−3.8	−0.2	<0.05*
Avoidance t	14	10.6 ± 3.7	8.0 ± 3.9	t = −2.5	−0.66	−2.6	−4.8	−0.3	<0.05*
Alertness t	14	10.7 ± 2.9	7.6 ± 3.0	t = −3.1	−0.83	−3.1	−5.1	−0.9	<0.01**
Nightmares w	14	2.0 (2.0)	2.0 (1.0)	z = −1.1	−0.21	−0.5	−1.0	0.5	0.27

Participants’ PSQI and PDS symptom severity test scores before and after treatment with cannabis. Lower scores imply improvement (excluding sleep duration, measured in hours). Different sample size between tests is due to the incomplete filling of questionnaires by some of the patients. The statistical tests used were a t-test for ratio and quasi-interval variables, Wilcoxon signed-rank test for ordinal variables, and a Hodges-Lehmann estimator for confidence intervals for ordinal scales variables. Each variable is labeled with the statistical method used according to the following: w = Wilcoxon signed-rank test; difference and confidence intervals were calculated using related-samples Hodges-Lehmann median difference; t = paired samples t-test; difference denotes the difference in means. *p < 0.05 and **p < 0.01.