5. Needle-free electronically-controlled jet injector treatment with bleomycin is efficacious and well-tolerated in patients with severe keloids: results of a randomized, double-blind, placebo-controlled trial. 121 5. Outcome assessments The outcome assessments used are as follows. (i) Change in volume (in mm3) and height (in mm) of keloid tissue measured by a three-dimensional (3D) camera (LifeViz Micro; Quantificare, Sophia Antipolis, France) at baseline compared with follow-up. (ii) Change in POSAS at baseline compared with follow-up. (iii) Change in skin perfusion measured by laser speckle contrast imaging (LSCI; Perimed PeriCam LSCI; Perimed AB, Järfälla, Sweden) at baseline compared with follow-up. (iv) Average spilled volume assessed by weighing a filtration paper before and after each EPI-assisted injection. Postinjection weight was determined by weighing the filtration paper after dabbing it at the injection site. (v) Procedure-related pain score measured with an 11-point numerical rating scale (NRS pain) directly after every EPI-assisted treatment. (vi) Incidence and type of adverse events, assessed every 4 weeks by the treating physician, and by the patient, who was instructed to take photographs daily and report adverse events with an e-diary app. (vii) Treatment satisfaction measured with the five-point Likert scale, evaluated by the patient at follow-up. Statistical analysis Sample size calculation was based on prior studies that investigated bleomycin treatment in keloids. We employed a 2-sided paired t-test with α = 0.05 to detect a significant difference of at least 35% in volume reduction between treatments. To reach a statistical power of 90% and account for a corresponding coefficient variance of the difference of 40%, a sample size of 11 patients was determined to be necessary. Anticipating a dropout rate of 25%, 14 patients were needed to demonstrate a treatment effect. Descriptive statistics and tables are presented as least squares mean (LSM) with confidence intervals (CI). A mixed effects model with a random subject factor and pre-value as covariate was
RkJQdWJsaXNoZXIy MTk4NDMw