Surveillance in a time of changing health care practices: estimating ectopic pregnancy incidence in the United States.

Objectives: Ectopic pregnancy is a common condition with significant health consequences; complications are a major cause of maternal mortality in the United States. Accurate ascertainment of the number of ectopic pregnancies occurring in the United States has been dramatically affected by changing medical practices, causing estimates based on hospital data to be falsely low. This study was performed to identify nationally representative data on ectopic pregnancies and determine overlap of these data, to calculate the annual weighted number of ectopic pregnancies and confidence intervals for these estimates, and to determine barriers to estimation of ectopic pregnancy incidence.

Methods: To assess whether a national estimate of the incidence of ectopic pregnancy could be calculated, we analyzed 1992-99 data from the six nationally representative data sets that include information on ectopic pregnancy. We examined relevant data in each data set and assessed whether any combination of data sets could be used to estimate ectopic pregnancy incidence. We calculated weighted estimates and 95% confidence intervals for hospitalizations, outpatient surgeries, outpatient medical procedures, and physician visits for and self-reports of ectopic pregnancy.

Results: Small sample sizes severely limited calculation of estimates of ectopic pregnancy. Data needed for assessing multiple counting was not available consistently. The likelihood of multiple counting of cases was substantial when data set counts were combined.

Conclusions: A reliable incidence rate for ectopic pregnancy in the United States could not be estimated from existing nationally representative data sources. Major advances in diagnosis and treatment of ectopic pregnancy have affected surveillance in two ways: inpatient hospital treatment of ectopic pregnancy has decreased, and multiple health care visits for a single ectopic pregnancy have increased. Alternate means of surveillance are needed to improve understanding of risk factors and trends for ectopic pregnancy, and we recommend examination of the databases of public and private insurance systems and managed care systems. Similar alternate means of surveillance may be needed for other health conditions with comparable changes in management of care.