Factors associated with provision of elective abortions among US
obstetrician-gynecologists
J. Steinauer, U. Landy, R. Jackson, P. Darney
Department of Obstetrics, Gynecology and Reproductive Sciences, University of
California, San Francisco, USA
Introduction: The number of abortion providers
has continued to decrease in the US. We sought to investigate what factors
correlate with elective abortion provision.
Aims and Methods: An anonymous,
validated, two-page survey mailed to all 5,000 ob-gyns who became
board-certified between 1998 and 2001 inquired about demographics, career,
abortion attitudes, residency training, intention to provide abortions prior to
residency, and current abortion provision. We performed logistic regression
models including all variables significantly associated (p<. 1) with abortion provision in univariate analyses, including a previously validated abortion attitude score.
Results: Two thousand one hundred thirty-eight (43%) surveys
were returned. Seventy-two percent (1540) stated that some abortion training had
been available in residency, of whom 901 (62%) participated in training.
Elective abortions were currently provided by 502 (23%). Thirty-four percent
worked in hospitals that prohibit abortion provision, and 16% worked in
practices that prohibit abortion provision. Prior to residency 33% had planned
to provide abortions, 50% had planned to not provide abortions, and 17% had been
undecided. In multivariate analyses, residency training availability increased
the odds of elective abortion provision (OR 2.60, p<.001.). Factors that negatively correlated with provision included religiosity (OR 0.70, p=.005), membership in a practice that prohibits abortion provision (OR 0.45, p<.001), association with a hospital that prohibits provision (OR 0.42, p<.001), and practice in a rural setting (OR 0.51, p=.001). Residency training availability was independently correlated with provision regardless of intention to provide abortion prior to residency.
Conclusions: Residency abortion training
availability increased the likelihood of abortion provision when controlling for
abortion beliefs, religiosity, practice location, subspecialty status, and
hospital/practice policy, regardless of intention to provide abortions before
residency.
