Dor 2002.

Methods	Retrospective cohort.
Participants	Cohort of 5026 women who received at least one treatment cycle from 1981 to 1992 at 2 IVF units operated by the same physicians, who used similar treatment protocols, in Israel. Participants were identified by meticulous review of the medical records of the units since their foundation. Mean follow‐up, 3.6 years.
Interventions	IVF treatment with three main ovarian hyperstimulation protocols: combined treatment with clomiphene citrate followed by human menopausal gonadotropin, FSH and LH; human menopausal gonadotropin; gonadotropin‐releasing hormone analogue, followed by human menopausal gonadotropin. Human chorionic gonadotropin was administered when the appropriate ovarian response was achieved.
Outcomes	Histopathologically confirmed endometrial cancer. The study cohort computer file was linked to the nationwide Israel Cancer Registry to identify cancer cases through 1996 and observed cases were compared to expected cases calculated from the general population. Incident cases in the cohort, 2.
Notes	Causes of subfertility were mechanical, ovulatory, male factor.
Risk of bias
Bias	Authors' judgement	Support for judgement
Selection bias (comparability)	High risk	Quote: "Cases of cancer that were diagnosed within 1 year of initiation of IVF treatment were excluded from analysis to allow a minimal latency period between exposure and development of cancer". Comment: Outcome not likely to be present at start. Quote: "Expected cases of cancer were computed on the basis of age, sex, place of birth, and year‐specific national cancer incidence rates". Comment: Non‐exposed drawn from the same population as the exposed cohort. Comment: No comparability on cause of subfertility, diabetes mellitus, polycystic ovary syndrome (PCOS), and obesity was ensured.
Selection bias (confounding)	High risk	Quote: "Standardized incidence ratios (SIRs) were computed as a ratio of observed to expected cases of cancer, along with estimated 95% CIs. Expected cases of cancer were computed on the basis of age, sex, place of birth, and year‐specific national cancer incidence rates". Comment: Analyses adjusted only for age.
Performance bias	High risk	Quote: "Patients were identified by meticulous review of the medical records of the units since their foundation". Comment: Exposure to ovary‐stimulation drugs was ascertained by a secure source, namely, medical records. Blindness regarding the allocated interventions not guaranteed.
Detection bias	Low risk	Quote: "The study cohort computer file was linked to the Israel National Cancer Registry to identify cancer cases through December 1996". Comment: Ascertainment of outcome by record linkage.
Attrition bias	Low risk	Quote: "Depending on the cancer site, cancer ascertainment during internal verifications through the Israel Cancer Registry was found to be 90% to 95% complete". Comment: Follow‐up was expected to be rather complete.
Selective reporting (reporting bias)	High risk	Comment: Analyses on endometrial cancer by type of subfertility, number of IVF cycles, and treatment outcome not reported.
Other bias	High risk	Quote: "mean follow‐up, 3.6 ± 3.4". Comment: Inadequate length of follow‐up (< 10 years). Comment: Non‐RCT study.