Comparative risk of death from legally induced abortion in hospitals and nonhospital facilities

Abstract

The rapid emergence of nonhospital abortion facilities as alternatives to traditional hospital-based surgical care has raised important questions concerning their safety. Using 1974--1975 data from the Center for Disease Control's nationwide surveillance of abortion mortality and the Alan Guttmacher Institute's nationwide surveys of abortion providers, we have calculated the risk of death from legally induced abortion in the United States at less than or equal to 12 menstrual weeks' gestation in hospitals and nonhospital facilities (clinics and physicians' ofices). The adjusted death-to-case rate for hospitals was 1.1 deaths/100,000 abortions, compared with a crude rate of 1.0 for nonhospital facilities. The risk of death from legal abortion in nonhospital facilities is apparently similar to the risk in hospitals.

PIP: To compare the risk of death from first trimester induced abortions in hospitals and nonhospital facilities (outpatient clinics and physicians' offices), the death-to-case rates by type of facility were calculated for abortions performed in the U.S. in 1974-1975. Data were derived from the Center for Disease Control's nationwide surveillance of abortion mortality and the Alan Guttmacher Institute's surveys of abortion providers. The data indicated 12 deaths from abortion-related complications in hospitals and 11 deaths in nonhospital settings (crude death-to-case rates of 2.1 and 1.0 deaths/100,000 abortions, respectively). However, 2 distorting factors were found to elevate the hospital rate: preexisting medical conditions, and concurrent sterilization. Adjustments for these factors yielded a hospital death-to-case rate of 1.1 deaths/100,000 abortions, indicating that the mortality risk is similar in both settings. These rates have 2 important limitations: 1) small changes in the already small number of deaths result in relatively large increases in the death-to-case rates, restricting extrapolation from these statistics; and 2) confounding variables influencing abortion mortality, e.g., patient's age and gestational age, cannot be controlled. Causes of death common to both facilities included adverse reaction to anesthesia, sepsis, hemorrhage, and pulmonary embollism, with no one cause disproportionally represented in either setting. Since over 90% of life-threatening complications developed either during the abortion or within 3 days, earlier patient follow-up is advised.