← Return to Article Details Cancer, meta-analysis and reporting biases: the case of erythropoiesis-stimulating agents