Data selection and analysis
Articles were screened on reporting an association between endometriosis
and depressive or anxiety symptoms. Furthermore, factors correlated with
or contributing to this association were listed. Data from studies using
validated outcome measures and reporting a mean and standard deviation
(SD) concerning depression or anxiety scores in endometriosis patients
compared with a control group, were pooled. The corresponding author was
contacted if a mean and/or SD was not provided in the article, and
additional data were added to pooling. Values were not included when not
following a normal distribution. A random effects meta-analysis was
carried out using STATA version 14 Software. Because of the
heterogeneity in assessment tools, a standardized mean difference (SMD)
with a 95% confidence interval (CI) was calculated. The heterogeneity
was tested through I-squared tests with a significance level of 5%.
Mean age (and SD) of cases and controls are provided in order to assess
comparability.
Results
Selection process The database search initially identified 1,837 records. After removal
of duplicates and adding articles from reference lists, 1,289 remained
for screening on title and abstract. 127 articles were assessed on full
text. Consensus was obtained in 90%, the remaining 10% were handled by
discussion and a third opinion. Finally, 47 articles were included and a
systematic qualitative review was performed [27-73]. Due to
variations in study design, lack of control groups, different methods of
measuring outcome and due to missing data, only 17 articles were
eligible for data-pooling and meta-analysis
[28,29,35,38,43,46,38,51,54-56,57,60-62,67,68]. Figure 1 shows a
flowchart of the literature review process.