Sample size recalculation in multicenter randomized controlled clinical trials based on noncomparative data