Psychiatric vulnerability and the risk for unintended pregnancies, a systematic review and meta-analysis 29 01-2000 and more recent and invited them to share data in case this was not available for the meta-analysis in published papers. Data extraction Two independent reviewers (NS and NR) screened the identified articles separately based on title and abstract using Rayyan QCRI software26. Subsequently, full text screening was performed independently by NS and NR to see whether the articles fulfilled all inclusion and exclusion criteria. If no agreement was reached, a third reviewer (BB) resolved conflicts. Data synthesis was performed by use of a custommade form that entailed all information necessary to compare studies. Variables analyzed in this review were authors and year of publication, presence and type of psychiatric disorder, presence and type of comparison group (if available), study design, sample size, age of participants, timing and tool used to measure UPs and prevalence of UPs in the study population. NS conducted the full data extraction and NR verified this. Assessment of risk of bias The Grading of Recommendations Assessment, Development and Evaluation (GRADE)27 method was used to assess quality of the outcome UP. The National Institute of Health (NIH) tools for quality assessment28 were used to assess the risk of bias in individual studies according to study type. Studies were qualified as ‘good’, ‘fair’ or ‘poor’ considering the risk of bias in that study for our specific outcome ‘UPs’. Hence, studies were assessed solely on the ability to report data on the outcome of interest in this review. Inconsistency was evaluated according to the following levels of heterogeneity by use of I2 tests: 25% was considered low, 50% moderate and 75% substantial heterogeneity29. A cut-off p-value of < 0.05 was used to determine statistical significance of the test. Indirectness was based on the ability of the data to relate to UP rates and imprecision was based on the confidence intervals of the presented results. Publication bias was assessed by evaluating a funnel plot for possible asymmetry. Also, we considered the absence of (un) published articles (with negative findings) in this field. The quality assessments were performed by two individual reviewers (NS and NR), and a third reviewer was involved to resolve conflicts (BB). Procedure for data synthesis Odds ratios (ORs), relative risks (RRs) and risk differences (RDs) were reported if present. In case of observational studies without comparative designs, percentages
RkJQdWJsaXNoZXIy MTk4NDMw