Table 2.

Characteristics of Included Studies

Source	Study design	Ascertainment of BSOa	Outcome measures	Comparison group	Subject numbers	Population details	Follow-up	Outcome ascertainment	Quality Rating
Colditz et al14 1987	Prospective cohort	Self-report and review of medical records for subset of subjects	CHDb: Nonfatal MIc and death due to CHD	Premenopausal women	Total cohort: 116,258 For nonfatal MI outcomes:112,387	Nurses Health Study cohort, USA Recruited 1976 Excluded: history of angina or MI 98% Caucasian 30–55 years at enrollment	6 years 5% lost to follow-up	Review of medical records attempted for all subjects For death: autopsy or death certificate	Intermediate
Gordon et al15 1978	Prospective cohort	Self-report Medical records when available	CHD: MI, angina, or death due to CHD	Hysterectomy with ovarian conservation, premenopausal and naturally menopausal women	Total cohort: 2,873 BSO: 398	Framingham cohort, USA Recruitment started 1948 29–62 years at enrollment	24 years “Essentially complete follow-up”	Self-report, physical examination, EKG, review of medical records for all cases	High
Howard et al33 2005	Prospective cohort	Self-report	Cardiovascular disease: MI, stroke, CABG/PTCAd, or coronary death	Naturally menopausal women	Total cohort: 89,914 Hysterectomy: 36,865 Hysterectomy/BSO: 18,543	Women’s Health Initiative Observational Study, USA Recruited 1994–98 Excluded subjects: BSO alone, baseline CVDe, angina, CHFf, unknown ethnicity or BSO/hysterectomy status 50–79 years at enrollment	Mean follow-up: 5.1 years	Self-report and review of all medical records	Intermediate
Luoto et al20 1995	Cross-sectional	Self-report Medical records if available (78% of subjects)	Angina and MI, or heart failure	No history of hysterectomy or BSO	BSO: 55 No history of gynecologic surgery: 3,562	Finnish women, random population sample Recruitment 1977–80 Year of hysterectomy 1944–79 30–95 years	N/A	Self-report and exam by physician at entry to study	Poor
Palmer et al21 1993	Case-control	Self-report	First nonfatal MI	Natural menopause	Cases: 858 Controls: 858	Participants from Massachusetts, community age-matched controls Data collected 1986–90 45–69 years at time of MI	N/A	Inpatient hospital records confirmed by physician report for all cases	Intermediate
Ritterband et al34 1963	Retrospective cohort	Medical records	Arteriosclerotic heart disease: MI, positive EKG, or angina	Hysterectomy with ovarian conservation	BSO: 267 Hysterectomy only: 385	New York, USA Subjects identified through hospital records All subjects ≥10 years from surgery	N/A	Self-report And Physical examination	Intermediate
Svanberg35 1982	Double-cohort study	Medical records	MI and angina, other CHD	No history of hysterectomy or BSO	BSO: 32 No history of hysterectomy or BSO: 32	Swedish women Comparison group: age-matched admitted to hospital for uterine prolapse surgery Outcome data collected 1978 15–30 years at BSO 52–84 years at data collection	N/A	Self-report and exam by study physician	Intermediate

^aBilateral salpingo-oophorectomy

^bCHD

^cMyocardial infarction

^dCoronary artery bypass graft/percutaneous coronary angioplasty

^eCardiovascular disease

^fCongestive heart failure