Objective
To compare time from misoprostol initiation to fetal expulsion for mifepristone–misoprostol versus misoprostol-alone regimens of medication abortion performed at ≥24 weeks’ gestation.
Study design
We conducted a retrospective study of medication abortion performed at ≥24 weeks’ gestation between May 2016 and January 2018 at one site, comparing outcomes of patients receiving mifepristone–misoprostol versus misoprostol alone during two periods. All patients received feticidal injection and laminaria; the mifepristone–misoprostol group also received mifepristone 200 mg orally around the time of initial laminaria. Beginning 24–72 h later (depending on cervical assessment), both groups received misoprostol buccally every two hours.
Results
Analyses included 257 patients in the mifepristone–misoprostol group and 152 patients in the misoprostol-alone group. Median time from misoprostol initiation to fetal expulsion was similar between groups (4.8 h vs. 4.9 h; p = 0.43). Patients in the mifepristone–misoprostol group received less misoprostol overall (median [IQR]: 800 mcg [800–1200 mcg] vs. 1200 mcg [800–1600 mcg]; p < 0.01) and fewer patients received a second round of laminaria (n = 56, 22% vs. n = 58, 33%; p < 0.01) than the misoprostol-alone group. Seven patients (2%) were transferred to a hospital for complications; this proportion did not vary by regimen.
Conclusions
Addition of mifepristone was not associated with a reduction in induction interval at ≥24 weeks. However, patients in the mifepristone–misoprostol group received a lower total dose of misoprostol and were less likely to require two days of laminaria. The clinical significance of these differences is unclear, but may have implications for patient experience. Both regimens had low rates of complications.
Implications
A randomized controlled trial comparing the mifepristone–misoprostol and misoprostol-alone regimens at ≥24 weeks is needed, as is evidence on patient perspectives on these regimens. Given the existing evidence, either regimen is reasonable.