Chapter 2 30 and means were reported. A meta-analysis of prevalence of UPs amongst women with psychiatric vulnerability was conducted by use of random effects models with the software programmes OpenMetaAnalyst30 and Rstudio31. An I2 test was performed to investigate heterogeneity of the studies in addition to sensitivity analyses to control for robustness of the findings29. A p-value of < 0.05 was considered statistically significant. Separate meta-analyses (forest plots) of specific psychiatric disorder groups were performed in case of ≥4 studies per disorder. Results Study selection The inclusion process is displayed in Figure 2.1. After electronic searches were performed 5429 articles were extracted and consequently transferred to Rayyan QCRI software26. After duplicate removal, screening of title and abstract of 3334 articles was conducted. This resulted in full text reading of 58 articles to assess whether inclusion and/or exclusion criteria were met. Based on the eligibility criteria, eleven articles could be included in the qualitative synthesis. Of the eleven articles, eight articles could be included in the meta-analysis on the prevalence of UPs amongst women with psychiatric vulnerability (Figure 2.3) and four studies in the meta-analysis of OR on UPs between women with and without psychiatric vulnerability (Figure 2.4). Study characteristics The characteristics and results of individual studies are presented in Table 2.1. An overall sample of 18,681 women with (n = 2650) and without (n = 16,031) psychiatric vulnerability were included. Seven categories of psychiatric disorders are represented in this review: eating disorders32,33, mood disorders (depression or bipolar disorder)34-38, anxiety disorders36-38, trauma-related disorders36,37, psychosis and related disorders35,39, substance use disorders37,40, and conduct disorders41. Two studies reported on abortion as an outcome of UPs39,41 and the other nine studies on (live) births. All studies were conducted in high income countries. Some of the included studies inquired for pregnancy intention during pregnancy, however these studies varied in timing of assessment32,36,38,42. Other studies did not report in which trimester women were asked about pregnancy intention33,35,37,40. One prospective cohort study assessed pregnancy intention prior to conception and evaluated the number of positive pregnancy tests over the course of one year34. In case a woman (without pregnancy aspirations at baseline) became pregnant within twelve months, the pregnancy was defined unintended. In addition, some studies made use of (validated)
RkJQdWJsaXNoZXIy MTk4NDMw