Background: Induction of labor is an increasingly common obstetrical procedure, with approximately 20–34% of women undergoing labor induction in the United States annually.
Objective: To determine the extent of labor induction in the absence of standard medical indications and to assess possible associations with maternal and infant characteristics and hospital factors.
Methods: We ascertained induction of labor and associated details as part of a medical record validation study of 4541 women with live, singleton births in 2000 in Washington State using medical record, birth certificate, and hospital discharge data. In this analysis, we report findings for the 1473 women (33% of original cohort) whose medical records indicated that their labors were induced.
Results: Among women with induced labor, 7.9% had no clinical information providing an indication for the induction, and 6.4% had only “nonstandard” indications recorded. Compared with women delivering in moderate volume hospitals, women who delivered at lower volume (odds ratios [OR] 3.9; 95% confidence intervals [CI] 1.8–8.6) or higher volume hospitals (OR 4.2; 95% CI 2.4–7.2) had significantly increased risk for undocumented indication of labor. Women who had undocumented indication for induction were at significantly decreased risk of giving birth at a teaching hospital and a public nonfederally owned hospital, and were at greater risk to give birth at a private religious hospital. Factors that remained independently associated with nonstandard indication for induction of labor were primiparas (OR 2.4; 95% CI 1.3–4.2); multiparas (OR 4.3; 95% CI 2.5–7.4), pregnancy-induced hypertension (OR 0.2; 95% CI 0.1–0.4), hospital volume ≥2000 births annually (OR 19.9; 95% CI 6.7–58.6), primary (OR 11.7; 95% CI 4.1–33.6), and tertiary level hospital (OR 0.4; 95% CI 0.2–0.7).
Conclusions: Our findings suggest that nearly 15% of inductions either were not clinically indicated according to standard protocols or indications were incompletely documented. At minimum, further studies are needed to explore how best to improve documentation of indications of labor because accurately describing, among other things, the process of labor induction, is a basic benchmark of care.
From the *Department of Family Child Nursing, School of Nursing; the Departments of †Health Services, ‡Epidemiology, and §Biostatistics, School of Public Health and Community Medicine, and ¶Department of Obstetrics and Gynecology, School of Medicine, University of Washington, Seattle, Washington; the ∥Program in Epidemiology, Division of Public Health Sciences, Fred Hutchinson Cancer Research Center; Seattle, Washington; and the **Center for Health Studies, Group Health Cooperative, Seattle, Washington.
Supported by a grant from the Association of Schools of Public Health, Centers for Disease Control, Prevention Cooperative Agreement (S1838-21/21), and the University of Washington’s School of Nursing Research and Intramural Funding Program.
Reprints: Mona T. Lydon-Rochelle, PhD, Mailstop 357262, University of Washington, Seattle, WA 98195-7262. E-mail: email@example.com.