Fig. 3. The funnel plot shows publication bias for osteoporotic fracture related to the use of bisphosphonate. There was symmetry, suggesting that there was no significant publication bias in the twenty-four studies.