| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
ORIGINAL RESEARCH |
From the Center for Reproductive Health Research and Policy, San Francisco General Hospital, Department of Obstetrics, Gynecology and Reproductive Sciences, University of California, San Francisco, San Francisco, California.
Address reprint requests to: Rebecca A. Jackson, MD, Department of Obstetrics and Gynecology, San Francisco General Hospital, 1001 Potrero Avenue, 6D, San Francisco, CA 94110, E-mail: rjack{at}itsa.ucsf.edu
 |
Abstract |
|---|
Top
 Abstract Materials and MethodsResultsDiscussionReferences |
|---|
Methods: We performed a randomized, double-masked, placebo-controlled trial of intra-amniotic digoxin for second-trimester dilation and evacuation (D&E) involving 126 consecutive women at an inner-city public hospital. Eligible women had gestational ages of 20–23.1 weeks, spoke English or Spanish, and were at least 16 years old. Digoxin (1 mg) or saline was injected intra-amniotically 24 hours before the procedure, at cervical laminaria insertion. The primary outcome was procedure duration. Sample size was based on 80% power to detect a difference of 3.5 minutes between groups.
Results: The two groups were similar in demographic factors, obstetric histories, and gestational duration. The average gestational length was 22.5 weeks. There was no difference in procedure duration (mean ± standard deviation) between groups (placebo 14.7 ± 7.0, digoxin 15.4 ± 8.0). There were no differences in blood loss estimated by surgeons, pain scores, procedure difficulty scores, or complications between groups. Vomiting was significantly more common in those who received digoxin (placebo 3.1%, digoxin 16.1%). Most subjects (91%) reported that they preferred their fetuses were dead before the abortions.
Conclusion: Although digoxin did not increase efficacy of late second-trimester abortion, patient preference might justify its use.
In the United States, dilation and evacuation (D&E) is the most common method of pregnancy termination beyond 13 weeks’ gestation. Approximately 144,000 D&Es are done annually and 7% are completed at 21 weeks’ gestation or greater.1 Although that represents only a small fraction of the 1.2 million legal abortions reported to the Centers for Disease Control and Prevention each year, morbidity and mortality rates are disproportionately higher in this group. Lawson et al2 reported a case fatality rate of 10.4 per 100,000 abortions at 21 weeks’ gestation or more compared with 0.4 per 100,000 at 8 weeks’ gestation or less.
Digoxin has been used to facilitate late second-trimester D&Es with the intent of decreasing procedure risks.3 When injected intrafetally or intra-amniotically in a toxic dose at least 1 day before D&E, digoxin causes fetal death. The result is soft, macerated tissue, which many clinicians believe eases evacuation of the fetus and decreases procedure duration. A shorter procedure is believed to be associated with decreased risk of infection and hemorrhage, the two major complications of D&E procedures.4 Another advantage is that patients, clinicians, and staff might prefer to abort a dead fetus. Although the actual prevalence of digoxin use is not known, in our nonrandom telephone survey of 20 D&E providers nationwide, 95% reported its routine use for terminations at or after 20 weeks’ gestation.
Despite widespread use, a MEDLINE search using the search terms "abortion," "digoxin," "induced," and "pregnancy" from 1966–September 2000 found only one series that described digoxin use.3 We found no comparative studies that examined efficacy of digoxin for facilitating D&E. A Phase I safety study was done before our trial, in which eight women were given 1 mg of intra-amniotic digoxin with no effect on maternal cardiac rhythms or coagulation indices and nontoxic serum levels of digoxin.5 We present the first randomized, placebo-controlled trial of intra-amniotic digoxin for decreasing procedure time of D&Es at 20–24 weeks’ gestation.
|
Materials and Methods |
|---|
|
Top
AbstractMaterials and MethodsResultsDiscussionReferences |
|---|
We used a randomized, double-masked, placebo-controlled design. Subjects were randomly assigned to receive digoxin or placebo. Group assignments were made in blocks of six using a computerized random number generator. A researcher not involved with enrollment placed group assignments in opaque envelopes and labeled them with study numbers. On the enrollment day, the clinic nurse opened an envelope in a private room and filled a syringe with 4 mL (1 mg) of digoxin or 4 mL of identical placebo (normal saline). Group assignments were confirmed by comparison with a sealed nursing medication log and measurement of serum digoxin levels. Patients, physicians, and research assistants were masked to treatment groups throughout the study.
Treatment took place over 2 days. On day 1, an ultrasound to determine gestational age and standard history and physical examination were done, 10–13 cervical laminaria were placed, and the intra-amniotic injection was given. Proper placement of the needle was confirmed by aspiration of amniotic fluid (AF) before injection. Ultrasound guidance was used if initial amniocentesis failed. On the second day, subjects returned to the clinic for D&E. Cervical dilatation was deemed adequate if a 14-mm cannula and a Bierer forceps passed easily through the internal cervical os. If dilatation was inadequate, subjects received additional laminaria or intravaginal misoprostol and waited in the clinic for 4 hours. If dilatation remained inadequate, D&E was completed the next day (n = 2).
The D&E procedures were done under ultrasound guidance an average of 24 ± 3.8 hours after the study medication was injected. Anesthesia consisted of a paracervical block with 20 mL of 1% chloroprocaine plus 4 units of vasopressin and intravenous conscious sedation with 100 µg of fentanyl plus 1–2 mg of midazolam. Additional midazolam and fentanyl were given as necessary during the procedure. The clinic nurse checked for fetal cardiac motion by ultrasound immediately before D&E and recorded presentation of the fetus after the AF was removed. Postprocedure bleeding was treated when necessary with methylergonovine or carboprost. Each subject received perioperative doxycycline and was monitored in the recovery room for at least 2 hours before discharge.
Dilation and evacuations were done by eight attending physicians and seven third-year resident physicians under direct supervision of an attending physician. Resident physicians were required to demonstrate proficiency in late second-trimester D&Es before participating in the study by doing several at 15–19 weeks’ gestation and completing at least two at over 20 weeks’ gestation, one in less than 15 minutes.
The primary outcome, procedure duration, was measured from the start of suction aspiration of AF until the final instrument was removed from the cervix. A research assistant masked to study group measured procedure time. Blood loss was estimated by the surgeon, and the change in hematocrit was measured between the enrollment day and 1 to 2 days after D&E. Serum digoxin levels were measured on the day of the procedure, an average of 24 hours after intra-amniotic injection. Major complications were defined as major unintended surgery, hemorrhage that required blood transfusion, or any complication that required admission to the hospital. Minor complications included cervical lacerations that required suturing, postabortal endometritis, blood loss estimated at more than 500 mL, and retained products of conception. Late complications, such as retained products of conception or endometritis, were ascertained during a follow-up phone call 2 weeks after D&E. We queried subjects about side effects commonly attributed to digoxin before and 24 hours after injection of study medication. Pain associated with injection, laminaria placement, and D&E was recorded on a six-point visual analog scale. Subjects also were asked whether they preferred fetal death before abortion and open-ended questions on digoxin use with D&E.
We estimated that we needed 63 subjects per group to detect a difference of ±3.5 minutes, assuming a mean procedure duration of 13 minutes, a standard deviation of 7 minutes, 80% power and a two-tailed alpha of 0.05. The estimated mean and standard deviation (SD) were based on a 3-month chart review of D&Es after 20 weeks at our clinic, most with digoxin.
The primary analysis compared the mean procedure duration between groups using Student unpaired t test. Median procedure time was compared using the non-parametric Wilcoxon rank-sum test. Estimated blood loss and other continuous outcomes were evaluated with Student t test. Categoric variables, such as rates of side effects and complications, were evaluated using Fisher exact test. For pain and difficulty ratings, we reported medians and 25th to 75th percentiles. Differences between groups were evaluated with the Wilcoxon rank-sum test. In all cases, groups were considered significantly different at a two-tailed P < .05. To assess effectiveness of masking, surgeons were asked to guess study groups. Based on that, the kappa statistic was calculated; it assesses the degree of agreement beyond that which would have occurred by chance alone.
Secondary analyses examined the possible effect of gestational age, provider type (resident or attending physician), and fetal presentation on procedure time. We used linear regression to calculate procedure times adjusted for those three factors and compared those times using the unpaired t test. We also examined procedure times in the following three subgroups: gestational age of 22 weeks or greater, among attending physicians, and in nulliparas. In those subgroups we had less power to detect differences between groups because sample sizes were approximately halved. Using a one-sided hypothesis that digoxin would decrease procedure time, a post-hoc power calculation found 80% power to detect a 5-minute decrease in the digoxin group for attending physicians (one-sided alpha = 0.05) or a 4-minute decrease for gestational ages at or above 22 weeks (one-sided alpha = 0.05). All analyses were done with STATA Statistical Software release 6.0 (Stata-Corp., College Station, TX).
|
Results |
|---|
|
Top
AbstractMaterials and MethodsResultsDiscussionReferences |
|---|
). Fetal presentation after AF evacuation was recorded in only half of participants and was not balanced between groups, with more breech presentations in the placebo group. We adjusted for this in later analyses.
|
). The average time was 0.7 minutes longer in the digoxin group (95% confidence interval [CI] -1.9, 3.3). The CI indicates that at best, treatment with digoxin shortened the procedure no more than 1.9 minutes and at worst, lengthened it by 3.3 minutes.
|
The average serum digoxin level in the treated group was 0.73 ± 0.2 mg/dL (range 0.2–1.2 mg/dL) compared with 0 mg/dL in all placebo subjects, which confirmed correct group assignment. Surgeons were able to guess which treatment the subject had received 74% of the time; 50% would have been expected to guess correctly by chance alone (kappa = 0.46). Surgeon rating of procedure difficulty was similar in both groups (digoxin 2.6, placebo 2.7 on a six-point scale).
There was one major complication, a uterine perforation that required laparotomy in a subject who received digoxin. There were six minor complications, including two cervical lacerations that required suturing (both placebo), one readmission for postabortal endometritis (placebo), one with retained products of conception (placebo), and two with more than 500 mL of estimated blood loss (one digoxin, one placebo). We were unable to contact 30 of 126 (23.8%) women by phone for the 2-week follow-up; however, a review of hospital records at our institution found only one readmission, as noted. There were no complications associated with intra-amniotic injection.
Procedure pain was not reduced with digoxin as measured by the six-point visual analog scale or amount of intravenous pain medication received (Table 2
). However, use of 10 mL or more of total medications (95th percentile) was more common in the digoxin group (9.7% versus 0%, P = .01). General anesthesia was required by 3.2% of women in the digoxin group and 1.6% in the placebo group (P = .62) because of inability to tolerate the procedure with local anesthetic plus conscious sedation.
Subjects reported a strong preference for fetal death before abortion (digoxin 92%, placebo 92%, P = .72). Although they did not know whether they had received digoxin or placebo, women in both groups understood that digoxin caused fetal death. We asked them to describe why they might want a digoxin injection for D&E in the future. The reasons were similar between groups. Among those who would want an injection (n = 107), reasons included desire for fetal death before the procedure (35%), belief that it made the procedure easier (29%) or less painful (19%), the injection was not painful (5%), unsure (4%), and no answer given (9%). For those who would decline an injection (n = 13), four stated the injection was painful, two preferred no fetal death before the procedure, three believed the injection did not work, and four gave no answer. Symptoms after injection were reported equally by both groups, except that vomiting was significantly more frequent in the digoxin group (Table 3, 16.1% versus 3.1%, P = .02).
|
). Among attending physicians with the most D&E experience, there was no difference in procedure times between groups. We used linear regression to control for gestational age, provider type, and fetal presentation. The adjusted procedure times were 15.0 minutes for placebo and 15.2 minutes for digoxin (P = .85).
|
|
Discussion |
|---|
|
Top
AbstractMaterials and MethodsResultsDiscussionReferences |
|---|
The results of our trial are applicable to D&Es for gestational ages older than 20 weeks. It is unlikely that digoxin would be beneficial in earlier gestations because those procedures are often completed in less than 10–12 minutes, even without digoxin.3 In our practice, D&Es are done by many providers with varying levels of experience. Nevertheless, the average operative times for these D&Es were similar to those reported for highly experienced clinicians,3,7 making it more likely that our results apply to all providers. We injected digoxin intra-amniotically, 24 hours before the procedure. Although a longer waiting period before D&E might have produced different results, in the pilot stage of our study we found an unacceptably high rate of spontaneous abortion if we injected digoxin 48 hours before the procedure. Intrafetal digoxin injection might induce death more quickly than intra-amniotic injection, which could result in a reduction in procedure time by providing a longer period for fetal softening. Further study should address dose and route of administration.
According to our MEDLINE search, ours is the only clinical trial that evaluated efficacy of digoxin for late second-trimester abortions. It was designed to detect a difference of ± 3.5 minutes between groups. Although a difference of less then 3.5 minutes cannot be excluded by our results, such a small difference has questionable clinical relevance. Balanced distribution of baseline factors between groups in our trial was evidence of effective randomization. Masking was not fully effective because 74% of physicians guessed the correct treatment group, compared with 50% expected. However, procedure duration should not have been affected by a provider’s knowledge of treatment group, and measurement of the outcome was made by a research assistant who was masked to study groups.
Despite the lack of evidence that digoxin decreases procedure time or difficulty, it might be preferred by patients because it induces fetal death before D&E. Women in our study stated a strong preference for fetal death before abortion. Those results should be interpreted cautiously because only patients who were willing to accept digoxin entered the trial. Further study is needed to determine whether patients would truly prefer digoxin, even if it does not decrease procedure time, difficulty, or pain.
Patient preference is a justifiable reason to continue to offer digoxin, but only if it is safe. Our study did not have adequate power to detect differences in rare complications such as uterine perforation or excessive blood loss requiring transfusion. At least 866 subjects per group would be required to detect a change in complication rate from 1% to 3%. We also had limited power to detect differences in side effects, but we found that they were reported more commonly in the digoxin group, with significantly more vomiting in women who received it. Future investigation should focus on those uncertainties and patient preference to clarify the appropriate role of digoxin in second-trimester abortions.
|
Footnotes |
|---|
Received July 24, 2000. Received in revised form October 10, 2000. Accepted October 26, 2000.
|
References |
|---|
|
Top
AbstractMaterials and MethodsResultsDiscussionReferences |
|---|
2. Lawson HW, Frye A, Atrash HK, Smith JC, Shulman HB, Ramick M. Abortion mortality, United States, 1972 through 1987. Am J Obstet Gynecol 1994;171:1365–72.[Medline]
3. Hern WM, Zen C, Ferguson KA, Hart V, Haseman MV. Outpatient abortion for fetal anomaly and fetal death from 15–34 menstrual weeks’ gestation: Techniques and clinical management. Obstet Gynecol 1993;81:301–6.
4. Grimes DA, Schulz KF. Morbidity and mortality from second-trimester abortions. J Reprod Med 1985;30:505–14.[Medline]
5. Drey EA, Thomas LJ, Benowitz NL, Goldschlager N, Darney PD. Safety of intra-amniotic digoxin administration before late second-trimester abortion by dilation and evacuation. Am J Obstet Gynecol 2000;182:1063–6.[Medline]
6. Hern WM. Correlation of fetal age and measurements between 10 and 26 weeks of gestation. Obstet Gynecol 1984;63:26–32.
7. Shulman LP, Ling FW, Meyers CM, Shanklin DR, Simpson JL, Elias S. Dilation and evacuation for second-trimester genetic pregnancy termination. Obstet Gynecol 1990;75:1037–40.
This article has been cited by other articles:
|
|
 |
|
  P. G. Stubblefield, S. Carr-Ellis, and L. Borgatta Methods for Induced Abortion Obstet. Gynecol., July 1, 2004; 104(1): 174 - 185. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
