Study	Reason for exclusion
Byström 2003	This study analyzed the Norwegian Arthroplasty Register to find risk factors for prosthesis dislocation leading to revision. Seven prosthesis brand combinations were used in 42897 primary hip arthroplasty in 68 hospitals. The study was excluded because we had no idea if groups of surgical approach are similar as for age, comorbidities, and types of prosthesis (shape, fixation,head sizes). In addition, the lateral approach included a mixed of Hardinge, antero‐lateral and trochanterotomy approaches.
Grossmann 1994	This study compared 45 dislocations in 1734 patients with a randomized control group of 61 patients. Factors causing an increasing number of dislocation as surgical approaches were analysed. The study was excluded as it was retrospective and without mention of the clinical diagnosis before total hip replacement.
Kohn 1997	This study reviewed 1238 primary total hip arthroplasties for arthritis. The influence of surgical approaches on dislocation was studied. The study was excluded as it was retrospective.
Moreschini 1996	This study reviewed 19 patients operated on for osteoarthritis (primary and secondary) or femoral neck fractures with clinical and electromyographic assessments. This study was excluded as it was retrospective and without specific data for primary osteoarthritis and adults results.
Moroni 2000	This study reviewed intraoperative femoral fractures after 3566 total hip replacements. Potential risk factors as surgical approaches were reviewed. The study was excluded as it was retrospective and there was not a comparison per surgical approaches.
Pascarel 1989	This randomized study compared 63 hip prostheses implanted with the direct lateral approach with 63 total hip arthroplasties implanted with the posterior approach to assess the functional impact. The study was excluded as the hip replacements were not done for osteoarthritis only (fractures, osteonecrosis, dysplasia,...) and no specific data were available for patients with osteoarthritis.
Zimmerman 2002	The study reviewed 271 patients aged 65 years and older, operated on for osteoarthritis, from numerous surgeons in different hospitals. The study was excluded as lateral and antero‐lateral approaches were mixed in one group, as well as posterior and postero‐lateral approaches in the other group. In addition, it was unrandomized and the groups were different in terms of age and comorbidities.