Search

Background Previous studies have provided conflicting results regarding the effect of drospirenone-containing oral contraceptive pills ( OCPs) on the risk of venous and arterial thrombosis. Objectives To conduct a systematic review to assess the risk of venous thromboembolism ( VTE), myocardial infarction ( MI), and stroke in individuals taking drospirenone-containing OCPs. Search strategy We systematically searched CINAHL, the Cochrane Library, Dissertation & Abstracts, EMBASE, HealthStar, Medline, and the Science Citation Index from inception to November 2012. Selection criteria We included all case reports, observational studies, and experimental studies assessing the risk of venous and arterial thrombosis of drospirenone-containing OCPs. Data collection and analysis Data were collected independently by two reviewers. Main results A total of 22 studies [six case reports, three case series (including 26 cases), and 13 comparative studies] were included in our systematic review. The 32 identified cases suggest a possible link between drospirenone-containing OCPs and venous and arterial thrombosis. Incidence rates of VTE among drospirenone-containing OCP users ranged from 23.0 to 136.7 per 100 000 woman-years, whereas those among levonorgestrel-containing OCP users ranged from 6.64 to 92.1 per 100 000 woman-years. The rate ratio for VTE among drospirenone-containing OCP users ranged from 4.0 to 6.3 compared with non-users of OCPs, and from 1.0 to 3.3 compared with levonorgestrel-containing OCP users. The arterial effects of drospirenone-containing OCPs were inconclusive. Author's conclusions Our systematic review suggests that drospirenone-containing OCP use is associated with a higher risk for VTE than both no OCP use and levonorgestrel-containing OCP use. [ABSTRACT FROM AUTHOR]

The article highlights the use of machine learning (ML) for the prediction of postpartum complications. Topics discussed include use of ML tools and routinely collected administrative data, such as electronic medical records, to predict health outcomes; the benefit that these tools are available to the physician and do not require any additional information; and the need of further study and demonstration in the field.