Objective: Multicenter clinical trials can be organized in different ways. Multiple centers participated in the HEALTH trial (Hip Fracture Evaluation with ALternatives of Total Hip Arthroplasty versus Hemi- Arthroplasty). For the Dutch sites most study tasks are managed by a central trial coordinator, whereas Canadian and US sites use local study coordinators. The aim of this study was to analyze how these strategies affected trial performance.Design: Prospective observational study.Method: Data related to ethics approval, trial startup time, inclusion rate and percentage of completed follow-ups were collected for each hospital and compared. Data from pre-trial screening were compared with actual inclusion rates.Results: The median start-up time of the trial after obtaining ethics approval was shorter in the Netherlands than in Canada and the US (4.6 versus 11.6 weeks).The inclusion rate was similar in both groups (0.62 versus 0.64/month). The median percentage of enrolled patients in the Netherlands was 27.3% versus 17.0% in Canada/US. The actual inclusion rates were lower than expected from pre-trial screening. The percentage of effectuated follow-up visits was > 90% in both groups.Conclusion: In this study, central trial coordination contributed to faster trial start-up and higher inclusion rates, but had no effect on the effectuated followup visits. Central coordination is therefore a suitable alternative for appointing these tasks to local research assistants and per patient payment. Central coordination enables non-academic hospitals to participate in clinical trials. Limiting conditions for central coordination are budget availability, a manageable number of patients, and a manageable distance between participating sites.