Cohort studyindependent cross-checkfull text checked★ key study I-ORG-134 in-depth review
Removing ovaries at hysterectomy and long-term survival
Association of bilateral salpingo-oophorectomy with all cause and cause specific mortality: population based cohort study.
Overview
Cusimano MC et al.. BMJ 2021 · PMID 34880044 · DOI 10.1136/bmj-2021-067528
Removing both ovaries before age 50 was linked to more deaths over 20 years (about 1 extra death per 71 women under 45), but not after 50.
Does the study support the specific claim? Partly · Methodological quality: high
Plain-language summary
Study explained simply
- What did the researchers want to find out?
Check whether ovary removal at hysterectomy affects survival
- How was the study done?
Population cohort with propensity weighting
- How many people, animals or samples were studied?
200,549 Ontario women
- What was actually found?
Higher mortality under 50, not over 50
- How large was the observed effect?
HR 1.31 (<45), NNH 71 at 20 y
- What do the results mean in everyday life?
Keeping healthy ovaries matters mainly before menopause
- What are the study's limitations?
Observational; some confounders missing
- What does it say about Elawady's specific claim?
Supports the claim with an age condition; contradicts its size
- Do the results apply to Germany and Europe?
Applicable
- What can explicitly not be concluded from this study?
Cannot prove causation
Scientific in-depth review
1. Full reference
- Authors
- Cusimano MC et al.
- Journal
- BMJ
- Year
- 2021
- Identifier
- PMID 34880044 DOI 10.1136/bmj-2021-067528
- Verification
- verified – PubMed efetch 2026-10-26; Crossref works API 2026-10-26
- Retracted
- no
- Corrections
- none
- Extent checked
- full text
- Conflicts of interest and funding
Full text: funded by ICES (Ontario Ministry of Health) and academic scholarships; funders had no role; no relevant competing interests.
2. Research question
Is bilateral salpingo-oophorectomy (BSO) at non-malignant hysterectomy associated with all-cause and cause-specific death, by age at surgery?
3. Study design and population
- Study design
- Cohort study – Population-based cohort with overlap propensity-score weighting
- Population
Women aged 30-70 undergoing non-malignant hysterectomy
- Country
Canada (Ontario)
- Period
1996-2015, follow-up to 2017 (median 12 years)
- Sample size
200,549 women; 76,383 with BSO
- Inclusion and exclusion criteria
Abdominal hysterectomy for non-malignant indication; excluded emergency, prior cancer, genetic predisposition, prior BSO
4. Methods and endpoints
- Exposure / intervention
BSO
- Comparator
Ovarian conservation
- Endpoints
All-cause death (primary); non-cancer and cancer death
- Statistical methods
Overlap propensity score weighted survival models; P < 0.0125 for 4 age strata
5. Actual results
- Results
All-cause death HR 1.31 (1.18-1.45; NNH 71 at 20 years) at age < 45 and 1.16 (1.04-1.30; NNH 152) at 45-49; not increased at 50-54 (0.83, 0.72-0.97) or >= 55 (0.92, 0.82-1.03). Excess in < 50 driven by non-cancer death. HR declined during the menopausal transition and stayed around 1 thereafter.
- Absolute risks
NNH 71 (age < 45) and 152 (45-49) over 20 years; no excess at >= 50.
- Authors' conclusion
BSO appeared associated with increased all-cause mortality in women < 50, not in those >= 50; caution in premenopausal women without indication.
6. Strengths
- Complete population data
- Validated exposure codes
- Propensity weighting
- Age-stratified
- Multiplicity correction
7. Weaknesses
- No data on menopausal status, smoking, BMI, family history (full text)
- Observational
8. Risk of bias
- Orienting tool
- ROBINS-E
- Type of assessment
- simplified assessment (abstract only)
- Full formal assessment
- no
- Overall risk of bias
- some concerns
- Rationale
Strong design; residual confounding by unmeasured lifestyle factors possible.
9. Assessment of the statistical interpretation
CIs log-symmetric; NNH 71 corresponds to about 1.4% absolute excess risk of death over 20 years (stats/C-ORG-09_10_11_gyn.py).
10. Relevance to the specific claim
Best recent evidence on the claim's main outcome; defines the age condition missing in the claim.
11. Independent contrary findings
Duan 2012 (I-ORG-135) found no increase.
12. Transferability to Germany and Europe
Canadian; directly applicable to Europe.
13. Overall methodological judgement
Supports the claim for premenopausal women (< 50) and contradicts it for women >= 50; the effect is far smaller than 'one death per nine operations'.
| Multidimensional assessment | |
|---|---|
| Authenticity of the source | verified |
| Methodological quality | high |
| Risk of bias | some concerns |
| Precision of results | high |
| Directness | direct |
| Reproducibility | replicated |
| Consistency with the overall evidence | consistent |
| Regional transferability | high |
| Adequacy of the interpretation (by the archive) | not applicable |
14. Open questions
Role of estrogen therapy.
15. Linked fact checks
- Bilateral oophorectomy at hysterectomy for benign disease increases all-cause mortality, coronary disease and lung cancer — roughly one extra death per nine operations (Parker 2013, Nurses' Health Study). – Studies with mixed or unclear results
- Long-term registry data link removal of tonsils, appendix, gallbladder, uterus, ovaries and spleen to higher rates of respiratory disease, autoimmunity, cardiovascular events, gastrointestinal cancers and dementia; each organ does work the body never fully replaces. – Studies with mixed or unclear results