|
Name of study | Type | Mean of followup (years) | Size of endometriosis cohort | Ovarian cancer cases identified in the cohort | Confounding factors considered | Main limitations |
|
Aris 2010, Canada [17] | Retrospective cross-sectional | 9 | 2521 | 41 | Age, pregnancies, family history, race, oral contraceptive, tubal ligation, hysterectomy and breastfeeding | Retrospective collection of data using a coded computerised system. Selection bias |
Borgfeldt and Andolf 2004, Sweden. [18] | Case-control | 10 | 28,163 | 81 | Age and parity | Use of cohort of women discharged from hospital with a diagnosis of endometriosis. This may lead to including women with moderate and severe endometriosis (hospital stay patients) without minimal and mild cases. This may overestimate the risk ratio. Selection bias. |
Brinton et al. 1997, Sweden. [19] | Retrospective cohort study | 11.4 | 20,686 | 29 | Age and length of history of endometriosis |
Melin et al. 2006, Sweden. [20] | Retrospective cohort study | 12.7 | 64,492 | 122 | Age and type of surgery performed. |
Melin et al. 2007, Sweden. [21] | Retrospective cohort study | 13.4 | 63 630 | 134 | Age, parity, and type of surgery performed |
Modugno et al. 2004, USA. [22] | Case-control | | | 177 | Age, parity, oral contraceptive use, tubal ligation, family history of ovarian cancer, and study site, gynaecological surgical procedures | Recall and selection bias, as the authors pooled data on the history of endometriosis reported by patients. |
Ness et al. 2000. USA. [23] | Case-control | | | 66 | Age, history of ovarian cancer, parity, breastfeeding, type and length of each contraceptive, tubal ligation, hysterectomy, ovarian operations. | Data was collected by case interviews. Recall bias. Low participation rates among cases and controls. |
Olsen et al. 2002, USA [24] | Prospective cohort study. | 13 | 1,392 | 3 | Age, education, marital status, alcohol intake, physical activity, smoking, parity, oral contraceptive use, HRT, history of hysterectomy or dilatation and curettage, BMI and waist to hip ratio. | Reliance on self-reports of endometriosis in this cohort (a questionnaire). Recall bias. The number of women who developed ovarian cancer is quite limited. |
|