Risk factors for failure in medical abortion
Marie Duriez, MD, Philippe Lefebvre, MD
Service d’Orthogénie (Hospital Family Planning Service), Roubaix, France
Aim: To identify potential risk factors of inefficiency for elective medical termination of
pregnancy based on records of failures of this technique in a hospital environment.
Patients and methods: A retrospective study was conducted on elective medical
pregnancy terminations performed up to 49 days post amenorrhoea in the Family Planning
Service of Roubaix hospital between January 1st 2001 and December 31st 2005. The
service's termination protocol consists in an oral dose of 600mg mifepristone, followed by
an oral dose of 400µg misoprostol 48 hours later. A 2nd oral dose of misoprostol (400µg) is
given 3 hours later if there has been no expulsion.
Every patient is required to return 15 days later to check their βHCG levels.
Failure is defined as ongoing pregnancies, total or partial retentions, and cases requiring
emergency surgery. Success is defined as complete abortion requiring no additional
surgical or medical treatment.
Five items were analysed: patient age, patient parity, duration of pregnancy, βHCG levels
on the day mifepristone (D1) was given, and the dose of misoprostol received.
Results: 1,975 medical terminations were performed during this period. 125 (6.33%) of
these patients did not return to be checked and have been excluded from the study. The
analysis was thus performed on 1,850 patients.
The method was a success in 97.08% of cases (1,796/1,850). 54 failures (2.92%) were
recorded, including 7 ongoing pregnancies (0.38%) .
Patients for whom the method resulted in a success compared to patients who had failures
have a significantly lower age.The duration of pregnancy was not different for the two
Nulliparous (873/1,850) patients had significantly fewer failures (12/873: 1.37%) than
multiparous patients (42/977: 4.30%).
Age is significantly higher for failures amongst the nulliparous patients. Conversely, for
patients who have had at least one child, age is not a determining factor.
28 patients did not receive any misoprostol because they expulsed prior to the 48th hour
(1.51%). Amongst the 673 patients who received only a single dose of misoprostol, 11
(1.63%) required additional actions including one emergency admission for haemorrhage.
Amongst those who received two doses of misoprostol, 43 failed (3.74%), including 2 re-
admitted the same day for haemorrhages and 1 for pelvic pains.
Discussion and conclusion: The overall efficiency results for the method are excellent
despite an exhaustive and detailed analysis of the failures.
The various studied factors have demonstrated that there is an increase in failure rates for
the method with parity and, to a lesser extent, with the patient's age.
High plasma beta HCG levels also seem to be more often associated with failures of the
method. The addition of a second dose of misoprostol is likely to increase the chances of
an expulsion during the hospital stay but, this non-comparative retrospective study can not
conclude on the beneficial effect of a second dose of misoprostol on overall efficiency.
Finally, it should be noted that none of the criteria evaluated in this study can be used as a
diagnosis factor to predict the outcome of an elective termination as none of them has the
sensitivity / specificity that is required to identify 'at risk' patients from amongst other