Study design/study setting
This study was an open-labeled randomized controlled study of 203 consenting women with incomplete abortion who presented at the gynecological emergency unit of the University of Nigeria Teaching Hospital (UNTH) Ituku- Ozalla, Enugu and Julius Memorial Specialist Hospital Enugu, Nigeria. The UNTH is a teaching hospital that offers 24-hour emergency obstetrics care as part of its comprehensive multi-specialist emergency services. Julius Memorial Specialist Hospital is a private specialist hospital situated in Abakpa Nike - a suburb and the most densely inhabited part of Enugu metropolis. It runs a 24-hour comprehensive emergency obstetric care and serves as a referral center for maternity homes and health centers in Enugu East local government area of Enugu State, Nigeria. Enugu is the capital city of Enugu state which is in the South-eastern geopolitical zone of Nigeria.
Sample recruitment
All consenting women at the study centers with first-trimester incomplete abortion, confirmed by a trans-abdominal ultrasound scan of the uterus, were eligible for the study. Women with excessive vaginal bleeding, severe anemia (Hb < 7.0g%), or suspected ectopic pregnancy were excluded from the study. Other exclusion criteria were a history of allergy to prostaglandins, evidence of genital infection such as offensive vaginal discharge, uterine tenderness, and pyrexia.
As illustrated in Figure 1, 212 consecutive consenting eligible women were randomized into two treatment groups; intervention group A (n=110) received single-dose sublingual misoprostol 400mcg while the control group B (n=102) had immediate manual vacuum aspiration (MVA) of the uterus. Participants’ randomization was by simple random allocation - after obtaining informed consent, each participant was assigned a sequential number (1 to 212) then, she picked one of the two balls (same size and texture) from a black bag. Each ball was marked A or B representing the two treatment groups. The participant was allocated to the treatment group written on the ball she picked. The sample size on each arm was adequate to identify a 95.5% difference in the incidence of complete uterine evacuation between the two treatment groups, based on the assumed incidence of complete abortion of 90% for the MVA (p0), power of 80%, alpha of 0.05, and attrition rate of 10%.
Data collection method
Each participant in the misoprostol group was given 400mcg of misoprostol (Cytotec®) placed under their tongues for 30 minutes to enable the tablets to dissolve. They were observed for about 8 hours after the drug administration and allowed home. Each participant was given the mobile phone number of the principal investigator for the reporting of any side effects or complications. They were discharged on analgesics – oral ibuprofen 400mg twelve hourly for 5 days, and prophylactic antibiotics - oral amoxicillin/clavulanic acid 625mg 8 hourly, and metronidazole 400mg 8 hourly for 1 week. Follow-up at the gynecology clinic visit was scheduled for each participant after one week. During the follow-up visit, they were clinically and sonologically evaluated for complete abortion; according to the study protocol, those with incomplete uterine evacuation were offered immediate manual vacuum aspiration or further observation for an extra week after which surgical evacuation would be carried out if the uterine cavity was not empty.
Participants in the MVA group (B) were offered immediate manual vacuum aspiration in the theater by trained medical practitioners, using standard procedure. [13] After the procedure, each was observed for 2 hours then, discharged home on the analgesics and prophylactic antibiotics described above. Each was scheduled for a follow-up clinic visit after 1 week of treatment, during which they were evaluated by ultrasound scan for complete uterine evacuation. In the case of incomplete uterine evacuation, the protocol was a repeat uterine evacuation by MVA.
At the initial treatment visit, participants’ baseline data on socio-demographics, gestational age, contraceptive history, and assigned treatment groups were documented. Their mobile phone numbers or that of a close relative was obtained. During the follow-up visit, information on the completeness of the uterine evacuation, treatment side effects and their severity, and satisfaction with treatment were obtained and recorded on the study’s datasheet for all participants. All documentations were made by the principal investigator and his trained assistants. Participant’s perceived severity of side-effect was assessed with a three-point Likert scale (i.e. mild/tolerable, moderate/less tolerable, and severe/intolerable). Also, the participants’ satisfaction with the treatment received was assessed using the visual analogue scale (VAS) (0 - 100), as described by Mba et.al.[14] All participants were also offered post-abortion contraceptive counseling and service and linked to other reproductive health services as appropriate. For the study, incomplete abortion was diagnosed when a pregnant woman presented with a history of vaginal bleeding with dilated cervix on physical examination as well as abdominopelvic ultrasound scan features of a retained product of conception.
Data variable
The primary outcome measure was the incidence of complete uterine evacuation (complete abortion) after one week of treatment with single-dose sublingual misoprostol 400mcg or MVA. The secondary outcome measures included incidence, types, and tolerability of treatment side effects as well as participants’ satisfaction with the treatment received.
Ethical approval
The study was approved by the Human Research Ethics Committee of the University of Nigeria Teaching Hospital, Enugu, Nigeria.
Data analysis
Data for 102 and 101 women were analyzed in the intervention (misoprostol) and control (MVA) groups respectively (figure 1) using the statistical package for social sciences (SPSS) computer software, version 20.0 for Windows (IBM Corporation). Data analysis was per protocol. Continuous variables were compared using the student’s t-test while the discrete variables were analyzed using proportions, and their associations compared with Pearson’s chi-square or Fisher’s exact where applicable. A probability value of less than 0.05 was considered statistically significant.