Safety, Efficacy and Acceptability of - Medical in Vietnam By Nguyen Thi Nhu Ngoc, Beverly Winikoff, Shelley Clark, Charlotte Ellertson, Khong Ngoc Am, Do Trong Hieu and Batya Elul

is inadequate in some clinics, and man- Context: In developing countries where the demand for abortion services is high, such as Viet- agement of pain requires improvement. nam, the need for safe and effective alternatives to surgical abortion is great. Indeed, while some women receive no using mifepristone and misoprostol may be an appropriate option in some of these countries. pain medication, others are medicated be- Methods: In a comparative study of the safety, efficacy and acceptability of medical and surgi- yond the point of conscious sedation and cal abortion, 393 women at two urban clinics chose between a mifepristone-misoprostol med- are consequently unable to respond to ical regimen and the standard surgical procedure offered in each clinic. physical or verbal stimuli. Results: Success rates for both methods were extremely high (96% for medical abortion and Vietnamese officials have responded to 99% for surgical abortion). Medical abortion patients reported many more side effects than women this situation by committing themselves obtaining surgical procedures (most commonly, cramping, prolonged bleeding and nausea), but to offering a broader range of contracep- none of these side effects represented a serious medical risk. Nearly all women, regardless of tives. They have also increased efforts to the method they chose, were satisfied with their abortion experience. Additionally, among women improve the quality of abortion services, who had previously undergone surgical abortion, those who selected medical abortion were including investigating the addition of al- more likely than those who chose surgery to say that their study abortion was more satisfacto- ternatives to surgical abortion. ry than their earlier one (32% vs. 4%). In this article, we describe a study ex- Conclusions: Mifepristone-misoprostol abortion is safe, effective and acceptable for urban Viet- ploring the safety, efficacy and accept- namese women who are given a choice of methods. If similar results are observed for rural areas, ability of mifepristone-misoprostol med- the regimen could help meet the need for abortion services nationwide. ical abortion among women attending International Perspectives 1999, 25(1):10–14 & 33 two clinics in Vietnam. We address three important questions: First, is medical abortion as effective as surgical abortion n the past decade, several nonsurgical Only two studies, however, have focused for women who choose the method? Sec- options have been developed for on the potential use of mifepristone and ond, how do the safety, risks and side ef- women seeking to terminate pregnan- misoprostol for medical abortion in de- fects of medical abortion compare with I 4 cies. To date, however, medical methods of veloping countries, and only one of these those of surgical abortion? Third, do abortion have been officially approved only measured the method’s acceptability to women who choose mifepristone-miso- in several European countries and China. clients.5 Given the potential of medical prostol abortion find the method accept- Although women in developed countries abortion to improve conditions for women able? Answers to these questions can help benefit from these new options, women in in developing countries, these women’s policymakers and providers in Vietnam, the developing world have a greater need perceptions of the method in general and as well as in other developing countries, for safe and effective alternatives to surgi- of the mifepristone-misoprostol regimen determine if medical abortion is a feasi- cal abortion: Nearly all of the estimated in particular is critical to its acceptability. ble and desirable alternative method of 70,000 deaths each year due to unsafe abor- Patients’ attitudes, expectations and toler- pregnancy termination. tion occur in developing countries.1 ance of side effects influence surgical in- The administration of mifepristone, a tervention rates; ultimately, for the method Methods powerful antiprogestin, coupled with a to work successfully, women must com- Study design plays a paramount role in prostaglandin is a highly effective med- plete the regimen and wait while the treat- the reliability and validity of acceptabili- ical method of terminating pregnancy.2 Of ment takes its course. ty data. In randomized clinical trials, the most widely used prostaglandins, In Vietnam, the number of pregnancy which are designed primarily to collect gemeprost and misoprostol, the latter terminations has risen steadily over the safety and efficacy data, women are as- shows the greater promise for use in de- past 15 years and is now estimated at more signed to use a particular method. In our veloping countries. Misoprostol can be ad- than one million per year;6 since the early study, which was modeled on research ministered orally and is inexpensive, sta- 1990s, the annual number of has conducted in China, Cuba and India,11 ble at ambient temperatures and widely exceeded the annual number of births.7 A available. By contrast, gemeprost is ex- 1994 nationwide survey found that 13% Nguyen Thi Nhu Ngoc is vice director, Hung Vuong Hos- pensive, not widely available and pro- of women have had at least one abortion.8 pital, Ho Chi Minh City, Vietnam; Khong Ngoc Am is re- tired director, Maternal and Child Health and Family vided in a vaginal suppository that re- Moreover, in 1992, the total abortion rate Planning, Hanoi Obstetric and Gynecology Hospital, quires refrigeration. In 1993, a large French was estimated as 2.5 lifetime abortions per Hanoi, Vietnam; and Do Trong Hieu is chief, Maternal trial confirmed the safety and efficacy of woman, the highest in Asia and the third- and Child Health Department, Ministry of Health, Hanoi. a regimen consisting of mifepristone and highest in the world.9 Beverly Winikoff is director, Reproductive Health; Shel- oral misoprostol.3 This regimen, with a Additionally, the surgical abortion ser- ley Clark is consultant; Charlotte Ellertson is program associate; and Batya Elul is staff program associate—all success rate of 96%, has been used exten- vices available in Vietnam are marked by with the Population Council, New York. This study was 10 sively in France and may be available in a number of safety and quality problems. funded by an anonymous donor. The authors thank the United States by the end of 1999. For example, sterilization of instruments Kurus Coyaji and Andrea Eschen for their assistance.

10 International Family Planning Perspectives women were allowed to choose their abor- two days later, they received 400 mcg of Table 1. Selected characteristics of women ob- tion method. This design reflects more misoprostol orally and were monitored at taining abortions, by method, Hanoi and Ho closely the situation under which the the clinic for at least four hours. Partici- Chi Minh City, Vietnam, 1995–1996 method will be used when offered in a pants were instructed to return for a fol- Characteristic Medical Surgical clinic. Thus, a sample of women who have low-up exam and an exit interview 14 (N=260) (N=133) chosen between medical and surgical days later, and were told to come to the Mean age 26.4 27.9** abortion constitute the correct population clinic at any time before then if they were Mean weight (kg) 46.4 46.6 from which to generalize about the ac- worried or if they changed their mind Mean height (cm) 155.8 154.5** Mean education (yrs.) 11.6 10.6** ceptability of both methods. A drawback, about the method. The women were not Mean gestational however, is that safety and efficacy data given any medication to control pain, age (wks.) 5.9 6.1* can be generalized only to women who since such medications are easily available % with first pregnancy 35.4 30.1 % married/in union 73.1 84.2* choose between methods. over the counter in Vietnam. % who had used The study was conducted from January Generally, if the abortion was not com- contraceptives 37.7 58.6*** % who had had 1995 to April 1996 in the two largest urban plete at the follow-up visit, surgical abor- previous abortion 48.5 43.6 centers in Vietnam, Hanoi and Ho Chi tion was performed as a backup. Among Minh City; one clinic in each city partici- the 10 women who had backup proce- *Difference between medical and surgical abortion patients is sig- nificant at p≤.05. **Difference between medical and surgical abor- pated. Both facilities had legal, established dures, five underwent tion patients is significant at p≤.01. ***Difference between med- surgical abortion services. Although abor- and three had sharp curettage; the method ical and surgical abortion patients is significant at p≤.001. Note: While the number of medical patients was roughly equally dis- tion services in Vietnam generally are of was unknown for the other two. Three tributed by site (48% from Hanoi, 52% from Ho Chi Minh City), the rather poor quality, these clinics had women whose abortions were incomplete distribution of surgical patients was quite uneven (72% from Hanoi, 28% from Ho Chi Minh City). Thus, the background data present- among the best services. at the follow-up visit were permitted to ed for surgical clients are more heavily weighted toward Hanoi. Both sites followed a uniform study pro- keep waiting rather than receiving surgi- tocol. Women seeking abortions could par- cal abortions. They returned later for ad- ticipate if bimanual examination showed ditional follow-up. may have been influenced by their earli- that they were no more than eight weeks Patients who chose surgical abortion er experiences, these women were asked pregnant (or if it had been no more than had the procedure on their first visit, in ac- to compare their study abortion and their 56 days since their last menstrual period), cordance with the clinics’ regular prac- prior abortion. they had no contraindications to medical tices. Nearly all of these women (98%) re- Data entry and analysis were per- or surgical abortion, they lived within one ceived vacuum aspiration without formed using standard statistical software hour of the clinic and they were willing to dilation. (Two women had vacuum aspi- (SPSS) and procedures. All means testing return for follow-up visits. Women aged ration with dilation, and one woman un- used t-tests, with Levene’s tests conduct- 35 or older were ineligible if they smoked derwent sharp curettage.) In Ho Chi Minh ed to determine whether pooled or sepa- 10 or more cigarettes per day. City, all surgical abortion patients received rate variance estimates were appropriate. If a woman met the study criteria and local anesthesia, while in Hanoi, most did Chi-square tests were used to analyze cat- wished to participate, a trained provider not receive any anesthesia. Fourteen days egorical data. All tests were two-tailed. explained both abortion methods. All after the procedure, patients returned to women received standardized counseling the clinic for a checkup and exit interview. Results about both procedures and their most Clinic physicians were already trained Sample Characteristics common side effects. For example, women in providing surgical abortions and The sample consisted of 393 women—221 were told that medical abortion is a rela- received additional training in medical in Hanoi and 172 in Ho Chi Minh City. tively new method, that it requires taking abortion for the study. They provided all Overall, 260 women chose medical abor- two sets of pills orally and that after the of the surgical procedures, administered tion and 133 opted for a surgical proce- second set of pills, most women experi- about half of the medical abortions and dure (Table 1).† ence cramping for several hours and supervised the nurses who administered Women who selected the medical bleeding for several days.* Moreover, they the other half. The in-country principal method were slightly younger than those were informed that in French studies, this investigators closely monitored the study who decided on surgical abortion (26.4 vs. medical abortion regimen was about 95% to ensure standardized treatment. Before 27.9 years) and had had more years of effective. The provider also explained the the main study began, each site conduct- schooling (11.6 vs. 10.6). Both groups types of surgical abortion available at the ed a pilot study of 10 medical patients. sought to terminate their pregnancies quite clinic and that this method was nearly Data on these women are included in our early, but the mean gestational age was 100% effective. Explicit comparisons be- analyses, since no significant changes somewhat lower among women who tween medical and surgical abortion were were made to the protocol following re- chose the medical method (5.9 weeks) than avoided, however, so as not to bias view of their experiences. among those who opted for surgery (6.1 women’s selection. After hearing about Providers collected clinical and experi- weeks). Women undergoing medical abor- both methods, women chose between ential data from each patient. Questions them. Any women who could not decide covered procedures, medications, side ef- *If a woman asked how long a medical abortion takes, would have been randomized to a fects or problems, and the woman’s reac- she was informed that while the majority of women ex- perience a complete abortion within several hours of tak- method, but no participants were unde- tion to the abortion experience. Addi- ing the second set of pills, some wait up to two weeks to cided. All women gave informed consent. tionally, women completed a daily diary have a complete expulsion. Women who chose medical abortion re- of all side effects during the weeks of the †This ratio is not meaningful, because many women who ceived 600 mg of mifepristone at their ad- study and indicated when they thought preferred surgical abortion (particularly in Ho Chi Minh mission visit and remained under obser- their abortion had occurred. Finally, since City) saw no reason to enroll in the study rather than sim- vation for 30 minutes. At a second visit, women who had had previous abortions ply to undergo the standard procedure.

Volume 25, Number 1, March 1999 11 Mifepristone-Misoprostol Medical Abortion in Vietnam

because it entailed fewer Table 2. Percentage of abortion patients Table 4. Percentage of medical abortion patients experiencing citing various reasons for selecting their visits (28%) or was con- various side effects, by segment of the regimen method, by method venient (26%). Fear of side effects was not a Side effect After mifepri- During obser- After obser- Reason Medical Surgical stone, before vation after vation, until (N=258) (N=131) major concern to women misoprostol misoprostol exit in either group when Effective 5.4 64.1 (N=258) (N=259) (N=257) they selected their Simpler and faster † 67.9 Nausea 37.6 6.9 6.2 Less pain 58.9 † method. Vomiting 15.9 0.8 2.3 Safer 40.4 47.3 Cramping/abdominal pain 38.8 93.8 37.7 Avoids surgery/anesthesia 43.4 † Only three medical Easier emotionally 30.2 † abortion patients did not Diarrhea 1.2 3.1 2.7 Fewer visits † 27.5 Prolonged bleeding 0.0 0.0 80.5 Convenient 7.8 26.0 complete the protocol. Profuse bleeding 2.7 4.2 2.3 Less bleeding † 7.6 One woman, feeling Increased bleeding 0.0 94.6 0.0 More natural 6.2 † worried and fatigued, Private 5.8 † Note: The observation period after administration of misoprostol was at least four hours. Fewer side effects † 3.8 went to another clinic be- fore taking misoprostol †Cited by one woman or no women. Note: Women could cite up to three reasons. and obtained a surgical abortion. Another surgical patients (7%) and three medical woman did not return to the clinic in time abortion patients (1%) were lost to follow- to receive misoprostol and had a surgical up. All available data from these 12 tion were less likely than those having sur- intervention. The third woman requested women are included in our analysis. gical procedures to be married (73% vs. a surgical abortion at another clinic after Side effects—nausea, vomiting, cramp- 84%) and to have been using a contracep- taking misoprostol because she had expe- ing, pain, diarrhea and bleeding—were far tive (38% vs. 59%). The differences in age rienced only spotting and not heavy bleed- more common among the medical abor- and length of gestation, however, were no ing. All three are included in the analysis. tion clients than among the women who longer statistically significant once we con- chose surgery (Table 3). However, al- trolled for study site (not shown). Efficacy and Safety though we have included cramping and Since medical abortion clients selected bleeding as side effects, they may be Method Choice and Adherence to Protocol their method to avoid surgery, we con- symptoms of a medical abortion; indeed, Upon enrollment in the study, women were sidered any of these women who under- if they do not occur, the woman is unlikely asked to name up to three reasons for their went a surgical procedure for any reason to have a successful medical abortion. method selection. Among women who se- to represent a treatment failure.12 All sur- Furthermore, medical abortion patients lected the medical method, 59% did so to gical abortion patients who had more than were observed on more occasions (at least avoid pain (Table 2). Substantial proportions one surgical procedure were also deemed three visits vs. at least two) and for a also chose the medical method to avoid to represent treatment failures. longer period of time (17 vs. 15 days) than surgery or anesthesia (43%), or because they Three types of failures can occur among were surgical abortion patients. More im- believed that it was the safer option (40%) medical patients: user choice, provider portant, even for medical clients, none of or that it would be less traumatic (30%). choice (or error) and true drug failures. the observed side effects represented a se- In contrast, women choosing surgical User choice failure occurs when a woman rious medical risk. abortion did so mainly because they per- asks for surgical intervention prior to the Side effects of medical abortion varied ceived it to be simpler and faster (68%) or end of the study or is unable or chooses at different stages of the procedure (Table more effective (64%) than medical abortion. not to take the complete medical treat- 4). Women were more likely to report nau- As with the medical patients, safety con- ment. Provider choice failure occurs when sea and vomiting after taking mifepristone cerns loomed large in the minds of surgi- a provider performs or recommends med- than later in the abortion process, but this cal patients (47%). Large proportions of ically unwarranted surgical interventions may reflect symptoms of pregnancy. (In- women also decided to undergo surgery (either out of impatience or in reaction to deed, upon enrollment in the study, 43% a concern with no clear medical basis). of all women reported nausea—42% who Table 3. Percentage distribution of abortion True drug failure occurs when an adverse chose medical abortion and 46% who patients, by outcome, and percentage of pa- event requires surgical intervention dur- opted for surgical—and 6% reported vom- tients citing various side effects, by method ing the study period or when an abortion iting.) Cramping and abdominal pain in- Measure Medical Surgical is not complete by the end of the study. creased sharply during the four-hour ob- (N=257) (N=124) Failure rates for both abortion methods servation period immediately after Outcome were extremely low (Table 3). Only one administration of misoprostol, but sub- Successful abortion 96.1 99.2 surgical patient (1%) required a backup in- sided later. Profuse bleeding, although Failure 3.9 0.8 Total 100.0 100.0 tervention. Among medical patients, there never experienced by more than 5% of the were 10 failures (for a rate of 4%): six user medical abortion clients, was also most Side effects Nausea 39.3 0.8*** choice, one provider choice and three true likely during these four hours. Vomiting 17.1 2.4*** drug failures.* Cramping/abdominal pain 96.1 37.1*** Diligent efforts were made to minimize *At the follow-up visit, three medical abortion patients Diarrhea 5.8 0.0** loss to follow-up. All women who did not had had incomplete abortions and were permitted to Profuse bleeding 8.9 4.8 keep waiting for their abortions to become complete. Two Prolonged bleeding 80.5 25.8*** report for a scheduled appointment were of these women had complete abortions confirmed when sent up to three reminder letters. Only after **Difference between medical and surgical abortion patients is they returned for an additional follow-up visit, a few days significant at p≤.01. ***Difference between medical and surgical providers made home visits in an effort to to one month after the first; the third woman received a abortion patients is significant at p≤.001. Note: Patients who were lost to follow-up are excluded. trace these patients were the women des- surgical intervention, because her abortion still was not ignated as lost to follow-up. In total, nine complete three days after her initial follow-up visit.

12 International Family Planning Perspectives Among the most serious risks of abor- medically induced abortions took place on Table 6. Percentage distribution of abortion tion, regardless of the method used, is ex- the day the women received misoprostol, patients, by measure of satisfaction with their cessive blood loss during and following and 8% took place throughout the next two method, according to method the procedure. On average, the women in weeks. However, medical abortion early both groups experienced minimal blood in gestation can escape detection; 10% of Measure Medical Surgical loss (Table 5). Only 2% of women who had medical abortion patients did not recog- Satisfaction (N=257) (N=124) Highly satisfied 5.4 2.4 medical abortions and 1% of their coun- nize when their abortions occurred. Satisfied 91.8 92.7 terparts who had surgical procedures ex- Most medical patients could identify Not satisfied 2.7 4.8 perienced a reduction in their hemoglo- where they were when the abortion oc- Would choose bin levels of greater than 2 g per deciliter curred (even if they could not pinpoint the method again*** (N=256) (N=123) (which is considered clinically meaning- time of the abortion). Nearly three-quarters Yes 95.7 51.6 No 4.3 48.4 ful blood loss), and none required a trans- (72%) reported that their abortions occurred fusion (not shown). at the clinic, but many (20%) said theirs oc- Would recommend method*** (N=251) (N=124) Analysis of participants’ diaries showed curred at home. About 1% reported other Medical 95.2 37.1 that medical abortion clients reported locations, and the rest were unsure. Surgical 2.0 28.2 more blood loss than did surgical abortion At the exit visit, all but one patient (who Either 2.8 34.7 patients. The mean number of days of had had a surgical procedure) stated that Comparison with previous bleeding (i.e., heavy, normal or light) was the explanation they had received about abortion*** (N=121) (N=57) More satisfactory 32.2 3.5 significantly greater for women who had their method adequately prepared them As satisfactory 64.5 86.0 medical abortions than for those who had for the abortion experience. The remain- Less satisfactory 3.3 10.5 surgical abortions.* For both groups, how- ing woman reported that the experience Total 100.0 100.0 ever, heavy bleeding accounted for only was worse than she had expected it to be. ***Difference in distributions between medical and surgical abor- a small number of total bleeding days. The vast majority of women were satis- tion patients is significant at p≤.001. Expectations about both the amount fied with their abortion experience—97% and the duration of bleeding also differed of those who had medical procedures and between the medical and surgical groups. 95% who had surgical abortions (Table 6). surgical abortion clients believed that the Medical abortion patients were more like- Of the 13 women who were not satisfied alternative procedure was preferable to the ly than surgical patients to have bled more with the experience, five had had method one they had chosen, perhaps because of and longer than they had expected to. failures. Nevertheless, about half of women discussions they had with women who ob- who had failures remained satisfied with tained the other procedure. Acceptability their abortions. A patient who had under- At their final visit, women were asked Where and when an abortion occurs after gone a surgical intervention after the med- to describe the best and worst aspects of a medical procedure may significantly in- ical procedure failed concluded that there their abortion method (Table 7, page 14). fluence the method’s acceptability. Ac- was nothing wrong with the medical Each was permitted up to three answers. cording to participants’ diaries, 82% of method, but that she was simply “unlucky.” For medical abortion, the features most In all, 178 women had had a previous frequently cited by patients were that the Table 5. Measures of bleeding experienced by surgical abortion—60% vacuum aspira- method is less painful than surgical abor- abortion patients, by method tion, 37% , and 3% tion (35%), is safer (30%), does not involve

Measure Medical Surgical some other surgical procedure. When surgery (20%) and is effective (14%). The asked how their experience during the emphasis on less pain is not surprising, MEANS Hemoglobin level (g/dl) (N=253) (N=123) study compared with their previous abor- given that surgical abortion is delivered At entry 11.8 11.6 tion experience, women who had medical with minimal anesthesia in Vietnam. At exit 11.7 11.6 abortions were significantly more likely Prolonged heavy bleeding was most Change –0.1 –0.1 than those who had surgical procedures commonly reported as the worst feature Days of bleeding*** (N=257) (N=124) to say that their study experience was of medical abortion (mentioned by 39% of Heavier than usual menses 1.3 (2.2) 0.4 (0.8) more satisfactory (32% vs. 4%). Medical women). A substantial proportion of med- Like normal menses 3.1 (2.7) 2.2 (1.2) Lighter than usual menses 6.2 (3.5) 3.1 (1.7) clients were less likely than surgical ical clients (17%) also reported that the clients to report that the study abortion method involved too many visits and too PERCENTAGE DISTRIBUTIONS was not as satisfactory as their previous lengthy a follow-up. Some 30% of women Amount of bleeding* (N=257) (N=124) More than expected 25.3 16.9 abortion (3% vs. 11%). who had medical abortions, however, As much as expected 57.2 65.3 Women who had medical abortions were unable to offer any negative features Less than expected 16.0 11.3 were significantly more likely to say they of the method. Not sure/do not know 1.6 6.5 would select the same method again than Women who chose surgical abortion Duration of bleeding*** (N=257) (N=124) were those who selected surgical abortion clearly appreciated the method’s effec- Longer than expected 49.0 24.2 As long as expected 34.2 58.1 (96% vs. 52%). Nearly all (95%) medical tiveness (46%), as well as the ease and sim- Shorter than expected 14.8 11.3 abortion clients would recommend their plicity of the procedure (23%). Yet 23% Not sure/do not know 1.9 6.5 method, compared with only 28% of sur- *Analysis of the mean number of days of bleeding, how- Total 100.0 100.0 gical abortion clients. Additionally, 37% of surgical abortion ever, overestimates the total number of days of bleed- *Difference in distribution between medical and surgical abortion clients would recommend medical abortion ing, since diary entries recording different types of bleed- patients is significant at p≤.05. ***Difference in distribution be- ing on a single day were counted as separate days of tween medical and surgical abortion patients is significant at to friends, while only 2% of medical abor- bleeding. Thus, for example, if a woman recorded both p≤.001. Notes: For days of bleeding, numbers in parentheses are standard deviations. tion clients would recommend surgical normal and heavy bleeding one day, she was counted as abortion. Thus, in hindsight, some of the having had a full day of each.

Volume 25, Number 1, March 1999 13 Mifepristone-Misoprostol Medical Abortion in Vietnam

ployed and suggests that the date of the 25(6, part 1):342–352. Table 7. Percentage of abortion patients cit- ing various features as their method’s best follow-up visit can be successfully delayed 7. Johansson A et al., Abortion in context: women’s ex- and worst characteristic, by method beyond the current standard of two perience in two villages in Thai Binh Province, Vietnam, weeks, which has been adopted from the International Family Planning Perspectives, 1996, 22(3): Feature Medical Surgical 103–107. (N=257) (N=124) surgical regimen. Side effects were more common among 8. Statistical Publishing House, Vietnam Intercensal De- Best medical abortion clients than among sur- mographic Survey, 1994: Major Findings, Hanoi, Vietnam: Effective 14.4 46.0 Statistical Publishing House, 1995. Less pain 34.6 † gical clients, but they did not jeopardize Safer/less risk the safety of the medical regimen and 9. Goodkind D, 1994, op. cit. (see reference 6). of complication 30.4 7.3 Faster/easier/simpler 5.1 23.4 were tolerable for the vast majority of 10. Vietnamese Ministry of Health, A strategic assessment None/not sure 8.2 22.6 women who chose that method. Howev- of policy, programme and research issues relating to abor- Avoids surgery 19.8 † er, women who had medical abortions re- tion in Vietnam: a draft report, Hanoi, Vietnam, 1997. Less mental stress/ healthier 7.4 † ported bleeding more and longer than 11. Winikoff B et al., The acceptability of medical abor- Convenient/compatible they had expected and more frequently tion in China, Cuba and India, International Family Plan- with duties 6.6 4.0 than women who obtained surgical pro- ning Perspectives, 1997, 23(2):73–78 & 89; and Winikoff B et al., 1997, op cit. (see reference 4). Worst cedures. Since women’s expectations may Pain † 57.3 significantly affect their comfort and sat- 12. Winikoff B et al., Analysis of failure in medical abor- Prolonged heavy tion, Contraception, 1996, 54(6):323–327. bleeding 38.9 22.6 isfaction with a method, medical abortion 13. None/not sure 30.0 18.5 patients must receive appropriate advance Winikoff B et al., 1997, op. cit. (see reference 11); and Too many visits/lengthy information to prepare them for the Winikoff B et al., 1997, op. cit. (see reference 4). follow-up 16.7 † 14. Winikoff B et al., 1997, op. cit. (see reference 4); Pey- Fear of surgery † 10.5 method’s potential side effects. ron R et al., 1993, op. cit. (see reference 3); and Aubény More mental stress † 8.9 This trial was conducted in major clin- E et al., Termination of early pregnancy (up to and after Long waiting time until ics in large urban areas, where backup fa- abortion 7.8 † 63 days of amenorrhea) with mifepristone (RU 486) and Fatigue/dizziness 5.1 † cilities are easily accessible and of rea- increasing doses of misoprostol, International Journal of sonably high quality. Studies in rural areas Fertility, 1995, 40(Suppl. 2):85–91. †Cited by one woman or no women. Note: Women could cite up to three reasons. with more basic facilities are needed be- fore the method’s safety, effectiveness and Resumen acceptability for women throughout the Contexto: En los países en desarrollo donde were unable to name any good charac- country can be judged. Additionally, since es elevada la demanda de servicios de aborto, teristics of the method. Although surgi- many medical abortion clients reported tales como Vietnam, es enorme la necesidad que cal abortion clients reported far less pain that the regimen involved too many vis- existe de contar con alternativas seguras y efi- during the study than did medical clients, its and many surgical clients chose their caces para evitar la intervención quirúrgica. 57% considered pain the method’s worst method because it entailed fewer visits, Una buena opción en algunos de estos países feature. Surgical clients also included fear research into a simplified protocol in- puede ser el aborto médico realizado median- of surgery and mental stress among the volving fewer clinic visits is important. te el uso del mifepristone y el misoprostol. worst features of their method. Nevertheless, our results indicate that Métodos: En un estudio comparativo realiza- mifepristone-misoprostol medical abor- do sobre la seguridad, la eficacia y la aceptabili- Discussion tion can complement available surgical dad de los abortos médico y quirúrgico, 393 mu- Our findings suggest that mifepristone- services and help meet the pressing need jeres de dos clínicas urbanas eligieron entre el misoprostol medical abortion is a safe, ef- for safe, effective and acceptable abortion método médico en base a mifepristone y miso- fective and desirable alternative to surgical services in Vietnam. prostol y el procedimiento quirúrgico estándar. abortion in Vietnam. The method’s success Resultados: Las tasas de éxito para ambos rate in our study (96%) is the highest docu- References métodos resultaron extremadamente elevadas mented in a developing country13 and is 1. World Health Organization, Abortion: A Tabulation of (96% para el aborto médico y 99% para el abor- comparable to the rate found in developed Available Data on the Frequency and Mortality of Unsafe Abor- to quirúrgico). Las pacientes del aborto médi- 14 tion, second ed., Geneva: World Health Organization, countries. Moreover, while the medical 1994. co indicaron un número mucho mayor de efec- abortion failure rate in our study exceeds that tos secundarios que las que se sometieron a 2. Bygdeman M et al., Progesterone receptor blockage: of the surgical method, many Vietnamese effect on uterine contractility and early pregnancy, Con- procedimientos quirúrgicos (más comúnmente women apparently are willing to accept an traception, 1985, 32(1):45–51. dolores, sangrado prolongado y náuseas), aun- increased risk of failure, since most said they 3. Peyron R et al., Early termination of pregnancy with que ninguno de estos efectos secundarios re- would choose medical abortion again and mifepristone (RU 486) and the orally active prostaglandin presentó un riesgo médico serio. Casi todas las would recommend it to their friends. misoprostol, New England Journal of Medicine, 1993, mujeres, fuere cual fuere el método escogido, Three women whose pregnancies had 328(21):1509–1513. se mostraron satisfechas con su experiencia. not yet terminated as of their exit visits 4. He C et al., Study on safety and efficacy of mifepris- Además, entre las mujeres que previamente se were advised to return for additional fol- tone plus misoprostol for termination of early pregnan- habían sometido a un aborto quirúrgico, aque- low-up rather than receive surgical inter- cy, Reproduction and Contraception, 1992, 3:1–10; and llas que escogieron un aborto médico eran más Winikoff B et al., Safety, efficacy, and acceptability of med- vention. Two had had complete abortions ical abortion in China, Cuba, and India: a comparative proclives que las que dicidieron de someterse by the time they returned and thus re- trial of mifepristone-misoprostol versus surgical abor- a un aborto quirúrgico a indicar que su abor- quired no backup procedure, while the tion, American Journal of Obstetrics and Gynecology, 1997, to actual era más satisfactorio que el anterior third eventually received sharp curettage 176(2):431–437. (32% contra 4%). to complete her abortion. This experience 5. Winikoff B et al., 1997, op. cit. (see reference 4). Conclusiones: El aborto médico en base a mi- confirms that the method’s failure rate is 6. Goodkind D, Abortion in Vietnam: measurements, fepristone y misoprostol es seguro, eficaz y largely a function of the protocol em- puzzles, and concerns, Studies in Family Planning, 1994, (continued on page 33)

14 International Family Planning Perspectives Medical Abortion in Vietnam... option viable dans certains de ces pays. ment de la méthode choisie, presque toutes les (continued from page 14) Méthodes: Dans une étude comparative de femmes se sont déclarées satisfaites de leur ex- la sécurité, de l’efficacité et de l’acceptabilité de périence. De celles qui avaient subi un avor- aceptable para las mujeres vietnamitas de l’avortement médical et chirurgical, 393 tement chirurgical précédent, celles ayant choi- zonas urbanas que tienen la opción de escoger femmes rencontrées dans deux cliniques ur- si la procédure médicale se sont du reste un método. Si se observan resultados simila- baines ont choisi entre un régime médical à base révélées plus susceptibles, par rapport à leurs res en las zonas rurales, este sistema podría sa- de mifepristone-misoprostol et la procédure chi- homologues qui avaient de nouveau choisi la tisfacer la necesidad insatisfecha de servicios rurgicale ordinaire offerte dans chaque clinique. méthode chirurgicale, de qualifier la procédu- de aborto que existe a nivel nacional en el país. Résultats: Les taux de succès des deux mé- re incluse dans l’étude de plus satisfaisante que thodes se sont avérés extrÍmement élevés (96% la précédente (32% par rapport à 4%). Résumé pour l’avortement médical et 99% pour la mé- Conclusions: L’avortement provoqué par mi- Contexte: Dans les pays en voie de dévelop- thode chirurgicale). Les patientes ayant choi- fepristone-misoprostol offre une méthode súre, pement qui présentent une demande de services si la procédure médicale ont signalé beaucoup efficace et acceptable aux yeux des Vietna- d’avortement élevée (le Viet Nam, par exemple), plus d’effets secondaires que celles qui avaient miennes auxquelles un choix de méthode est il existe un besoin important de solutions súres demandé l’intervention chirurgicale (douleurs offert. Si des résultats comparables étaient ob- et efficaces autres que les procédures chirurgi- abdominales, saignements prolongés et nau- servés dans les milieux ruraux, le régime pour- cales. L’avortement médical à base de mife- sées, surtout), mais aucun de ces effets ne pré- rait aider à répondre au besoin de services pristone et de misoprostol pourrait offrir une sentait de risque médical grave. Indépendam- d’avortement à l’échelle nationale.

Volume 25, Number 1, March 1999 33