Joost Peerbooms

129 Discussion were lost in follow-up. The same can be concluded with regard to follow-up in the arthroplasty study. In this part of the study, we were unable to follow up with 36% (18 out of 50) in the PRP group and 21% (11 out of 52) in the control group due to incomplete primary outcome (wound healing) datasets. In both studies on tendinopathy ( Chapter 5, 6, 7 ), we did not use ultrasound-guided injections for either group. Ultrasound-guided techniques were used in some other studies. 64 However, Kane et al. did not find advantages in using ultrasound guidance over direct palpation of the most tender area in terms of guidance of the injections. 65 All injections were given via the same technique for both corticosteroids and PRP. In the plantar fasciitis study ( Chapter 7 ), 16 patients were treated with a 30 ml PRP kit instead of the 60 ml PRP kit described in the protocol. This was due to logistical reasons in one of the treatment centres. This was discussed with the Medical Ethics Committee (METC), and we analysed it statistically. The influence of dosage on the effect of treatment was assessed by inspecting the three-way interaction effect among treatment group, time and injection dosage. However, this conclusion is based on a small sample size (only 16 patients). Several studies have demonstrated that the reaction of cells depends on the dosage of the growth factors, but that very high concentrations are not essential for optimal stimulation of cell processes and may have an opposite effect. 34,66,67 Because the results suggest that the injection dosage did not affect the differences between the treatment groups in the changes in their outcomes over time, we did not exclude them from this study. Control group In our tennis elbow studies ( Chapter 5, 6 ), the recurrence rate and need for repeated injection or surgery was greater in the corticosteroid group than in the PRP group. This may suggest that corticosteroid injections may have a detrimental effect on tendons. This is supported by two high-quality reviews on treatment effects in chronic tendinopathies. 68,69 These studies showed that corticosteroids provided inferior clinical outcomes compared to a wait-and-see policy after a six-month follow-up. These results were also found in a high-quality double-blind RCT. 70 This raises the question of whether the outcome seen in favour of PRP in our studies is due to the beneficial effect of PRP or due to the detrimental effect of corticosteroids. The difference between our studies and those of Coombes et al. (2013) and Smith et al. (2002) is that in our study the duration of complaints was chronic (more than six months), while in the Coombes et al. and the Smith et al. studies, there was a much shorter duration of complaints (six weeks or more). 69,70 Therefore, it would be valuable to have RCTs with normal saline as a control group because of the suggested placebo effect of normal saline. In a systematic review and meta-analysis, normal saline injections 8