Objective: To compare outcomes associated with expectant vs. interventionist care of severe preeclampsia in observational studies.

Data sources: Medline (01/1980-07/2007), bibliographies of retrieved papers, personal files, Cochrane Database of Systematic Reviews.

Study selection: Expectant or interventionist care of preeclampsia at <34 wk. TABULATION, INTEGRATION, RESULTS: Data abstraction independently by two reviewers. Median [IQR] of clinical maternal/perinatal outcomes presented.

Results: 72 publications, primarily from tertiary care centres in Dutch and developed world sites. Expectant care of severe preeclampsia <34 wk (39 cohorts, 4,650 women), for which 40% of women are eligible, is associated with pregnancy prolongation of 7-14 d, and few serious maternal complications (median <5%), similar to interventionist care (2 studies, 42 women). Complication rates are higher with HELLP <34wk (12 cohorts, 438 women) and severe preeclampsia <28wk (6 cohorts, 305 women), similar to interventionist care (6 cohorts, 467 women and 2 cohorts, 70 women, respectively). Expectant care of HELLP <34 wk (12 cohorts, 438 women) is associated with fewer days gained (median 5), but more serious maternal morbidity (e.g., eclampsia, median 15%). More than half of women have at least temporary improvement of HELLP. In the developed world, expectant (vs. interventionist) care of severe preeclampsia or HELLP <34 wk is associated with reduced neonatal death and complications. Stillbirth is higher in Dutch and developing world sites where viability thresholds are higher. For preeclampsia <24wk (4 cohorts), perinatal mortality is >80%. No predictors of adverse maternal/perinatal outcomes were identified (13 studies).

Conclusions: Future research should establish the best maternal/fetal monito regimen and indications for delivery with expectant care. A definitive RCT is needed.