Joost Peerbooms

21 Concentrations of blood components in PRP separation systems 1. Humans 2. Platelet-rich plasma 3. 1 AND 2 4. Blood platelets or platelet count 5. Leukocytes or leukocyte count 6. Platelet-derived growth factor 7. 3 AND 4 AND 5 8. 3 AND 6 9. 7 OR 8 The search was performed by one of the authors (B.W.O.). References of retrieved publications were also used to add studies potentially meeting the inclusion criteria that were missed by the electronic search. Abstracts from scientific meetings and review articles were excluded. Review Process To identify relevant articles for this review, the title and abstract of the articles found by the abovementioned search strategy were reviewed. After selection, the full articles were reviewed for definitive selection. All identified studies were independently reviewed by 2 reviewers (B.W.O. and J.C.P.) for inclusion using the abovementioned criteria. In case of disagreement, a third reviewer (A.J.H.V.) was consulted to resolve the disagreement. Data Collection The following data were extracted from the included trials: study design (comparative study or study describing one separation device), study characteristics (e.g., number of blood samples), concentration analysis methods, type of outcome, results of the study, and main conclusion(s) of the study. This information was extracted by one author (B.W.O.). If necessary, authors were contacted for additional information about their specific article. The companies producing the PRP separation systems were contacted to gain information about the specific preparation protocols. In case a company did not respond to the request, the literature was searched for the preparation protocol. STATISTICAL ANALYSIS First, 95% CIs were calculated for each of the blood components studied in the included studies using the mean concentration, SD, and number of samples. The following formula was used: where is the mean concentration, γ the critical value of the t distribution based on the sample size of the study, the standard deviation and the number of samples studied. Forest plots were created using the mean and 2