Population surveys using validated questionnaires provided useful information on the prevalence of maternal morbidities

Abstract

Objectives: To evaluate the ability of population surveys to estimate the occurrence of maternal morbidities.

Study design and setting: A literature search was conducted using MEDLINE, EMBASE, POPLINE references from relevant papers and proceedings of scientific meetings. No restrictions were made regarding language, date, design, journal, or country. Potentially relevant papers were independently evaluated by two reviewers. Eligible studies were critically evaluated, particularly with respect to complications: eclampsia and other hypertensive complications, hemorrhages, dystocias, and infections. The questions with the highest combined values for sensitivity and specificity were identified in each study.

Results: Seven hospital-based studies involving 2,907 women were included. The gold standard was the clinical records, and the validation strategy consisted of applying questionnaires and comparing them with the gold standard. Questions regarding eclampsia and other hypertensive complications performed satisfactorily in four studies; questions on dystocia and infection in two studies each, and questions regarding hemorrhagic complications in only one study. In general, when the actual prevalence of the condition is low (<or=5%), surveys tend to overestimate prevalence.

Conclusions: Prior validation of questionnaires on maternal morbidity is fundamental to assure adequate information. Population surveys using validated questionnaires may provide useful information on the prevalence of maternal morbidities.