Joost Peerbooms

107 PRP in plantar fasciitis Treatment was considered being a success if patients showed a FFI pain score reduction of 25% between baseline and 12-month follow-up. In addition, patients should not have required escape therapy and pain medication beyond the protocol defined allowable period. Patients that obtained a different treatment were classified as unsuccessful. To determine the percent of change, first the baseline pain score was subtracted from the endpoint pain score. Subsequently, this difference score was divided by the baseline pain score and multiplied by 100. If a patient was lost to follow-up, the last available measurement was used to determine the treatment success. Function and Quality of Life The secondary outcome measures of this study were the FFI Disability, FFI Activity 2,4,14 and American Orthopaedic Foot & Ankle Society (AOFAS) score. 13,26 Lastly, patients’ quality of life was assessed using the World Health Organization Quality of Life (WHOQOL-BREF). 20, 31 This is the short version of the WHOQOL-100. The WHOQOL- BREF consists of 4 domains (Physical Health, Psychological Health, Social Relationships, and Environment) and 2 items assessing overall quality of life and general health. The response scale consists of 5-point Likert scales. Higher scores indicate better quality of life. All outcomes in this study were measured at baseline, 1, 3, 6 months and at 1 year after the procedure. Determination of Sample Size Our main hypotheses will be tested by investigating the interaction effect between treatment and measurement occasion, indicating whether the treatments differ in their change in the outcome over time. We are not aware of earlier research comparing PRP with corticosteroid treatments for chronic plantar fasciitis on pain, function and quality of life with a follow up of at least 1 year. Therefore, we take a conservative stance by assuming a small partial eta-squared effect size of 0.02 and a correlation between the repeated measurements of 0.3. To detect such effect sizes with a power of 0.80 and a significance level of .05, at least 84 participants are required (42 in each group). Statistical Analysis For dichotomous baseline characteristics, frequencies and percentages were reported. Means and standard deviations were calculated for continuous and normally distributed baseline characteristics. For nonnormally distributed continuous characteristics, the median and interquartile range were reported. To test the null hypothesis that the treatment groups do not differ in their change on the outcome measures over time, linear mixed modelling analyses were used, focusing on the interaction effect between treatment group and time. The influence of dosage on this treatment effect was assessed by inspecting the 3-way interaction effect between treatment group, time and injection dosage. For all outcome measures, 7