From: A case study of binary outcome data extraction across three systematic reviews of hip arthroplasty: errors and differences of selection
