Arver 2011.

Methods	Retrospective series
Participants	223 high‐risk women (> 20%) without a previous breast malignancy who had BRRM performed at 8 hospitals in Sweden between 1995 and 2005
Interventions	BRRM
Outcomes	BC incidence Physical morbidity ‐ complications, reoperations
Notes	Includes 24 Gahm 2007 subjects
Risk of bias
Bias	Authors' judgement	Support for judgement
Free of selection bias?	Low risk	A consecutive sample of a clearly defined population was chosen. These were “Women without a previous breast malignancy who had undergone BRRM in Sweden between January 1, 1995 and December 31, 2005”.
Free of performance bias?	Low risk	Exposure was confirmed in an objective way. “Oncologists and geneticists at the university hospitals throughout Sweden … were asked to identify women who had been referred for risk‐reducing surgery”.
Free of detection bias?	Low risk	The outcomes were assessed in a valid way. A questionnaire was developed in collaboration with collaborator input. “Data were derived from patient charts, the regional Oncology Centers, and the nationwide Cause of Death Register”.
Free of attrition bias?	Low risk	There was a low dropout rate and dropouts/withdrawals were sufficiently accounted for. Only 15% of cases did not have genetic screening, and 1 woman declined mapping.