Joost Peerbooms

89 2-year follow-up of PRP in lateral epicondylitis Course of VAS Pain Scores (Intention-to-Treat Principle) As shown in Figure 2A, the course of the VAS scores across assessment points is different for the group treated with corticosteroids and the PRP group ( P < .0001). Table 2 shows the mean scores and standard deviations. The base-line scores of the corticosteroid group were significantly higher compared with all subsequent time points ( P < .0001), except for 26 weeks ( P = .029). Between 8 weeks and 26 weeks, pain scores temporarily got worse ( P = .007). In contrast, compared with baseline scores, the scores of the PRP group significantly improved during the entire duration of the study ( P < .002). Overall, the average VAS scores differed significantly between the 2 groups ( F 1,98 = 6.3, P = .014). Figure 2. The course of visual analogue scale (VAS) pain scores across assessment points. Bars present 95% confidence intervals. Patients with chronic lateral epicondylitis were ran- domly assigned to the platelet-rich plasma (PRP) group or the corticosteroid group. A, inten- tion-to-treat; B, reintervention excluded. When VAS scores were compared at each assessment point separately, the PRP group scored significantly worse at 4 weeks after the injection ( P < .023), while the opposite was found at 26 weeks ( P < .0001), 52 weeks ( P < .0001), and 104 weeks ( P < .0001) after treatment. No differences between the PRP group and the corticosteroid group were found at baseline, 8 weeks, and 12 weeks. In general, the results of the intention- to-treat analysis and the as-treated analysis were comparable (Figure 2B). 6