Fig. 2.

Mantel–Haenszel statistical method with random-effects analysis model and OR as output only for included observational studies and RCT, and funnel plots assessing respective variance

Forest plots analysing crude event numbers between robotic nipple-sparring mastectomy and conventional nipple-sparing mastectomy. a Infection. b Seroma. c Haematoma. d Implant loss. Overall heterogeneity for the respective outcomes was considered acceptable (less than 30 per cent) given the nature of the included studies. RNSM, robotic nipple-sparing mastectomy; CNSM, conventional nipple-sparing mastectomy; M-H, Mantel–Haenszel; SE, standard error.