Contraception xx (2015) xxx – xxx

Original research article

Early serum human chorionic gonadotropin (hCG) trends after medication abortion☆,☆☆,★ Katherine D. Pocius a, b,⁎, Rie Maurer b, c , Jennifer Fortin d , Alisa B. Goldberg a, b, d , Deborah Bartz a, b, d a

Department of Obstetrics, Gynecology, and Reproductive Biology, Brigham and Women's Hospital, Boston, MA 02115, USA b Harvard Medical School, Boston, MA 02115, USA c Center for Clinical Investigation, Brigham and Women's Hospital, Boston, MA 02115, USA d Planned Parenthood League of Massachusetts, Boston, MA 02115, USA Received 5 August 2014; revised 2 March 2015; accepted 3 March 2015

Abstract Objectives: Despite increased reliance on human chorionic gonadotropin (hCG) for early pregnancy monitoring, there is limited information about hCG trends soon after medication abortion. The purpose of this study was to determine if there is a predictable decline in serum hCG values shortly after medication abortion. Study Design: This is a retrospective study of women with early intrauterine pregnancies who underwent medication abortion with mifepristone and misoprostol and had a serum hCG level on Day 1 (day of mifepristone) and a repeat value on Day 2 to 6. The percent hCG decline was calculated from baseline to repeat measure, with repeat values from the same patient accounted for through repeated measure analysis of variance. Results: Eighty-eight women with a mean gestational age of 5.5 weeks and median baseline hCG of 5220 IU met study criteria over a 3-year period. The mean decline (± SD) in hCG from the Day 1 baseline value was 56.9%±29.5% on Day 3, 73.5%±38.6% on Day 4, 86.1%±8.8% on Day 5, and 92.9%±3.4% on Day 6. Eighty-two women (93% of the cohort) had a complete abortion without further intervention. The least square means hCG decline among these women was 57.6% [95% confidence interval (CI): 50.3–64.9%] on Day 3, 78.9% (95% CI: 75.0– 82.8%) on Day 4 and 86.2% (95% CI: 81.3–91.1%) on Day 5. Conclusion: There is a rapid decline in serum hCG within the first few days after early medication abortion. Further research is needed to delineate how soon after medication abortion this decline may be specific enough to confirm abortion completion. Implications: This study provides the largest cohort of patients followed with serial hCG values in the first few days after medication abortion. Our findings demonstrate the trend in hCG decline in this population, which may be predictable by Day 5. © 2015 Elsevier Inc. All rights reserved. Keywords: Medication abortion; Medical abortion; Mifepristone; Human chorionic gonadotropin; hCG; Follow-up

1. Introduction Medication abortion accounts for more than one third of all US abortions before 63 days gestation [1]. Unfortunately, ☆

Funding: none. Disclosures: none. ★ Clinical trial registration number: N/A. ⁎ Corresponding author at: Department of Obstetrics, Gynecology, and Reproductive Biology, Brigham and Women's Hospital, Boston, MA 02115. Tel.: + 1-908-565-3894; fax: +1-617-730-2833. E-mail addresses: [email protected] (K.D. Pocius), [email protected] (R. Maurer), [email protected] (J. Fortin), [email protected] (A.B. Goldberg), [email protected] (D. Bartz). ☆☆

http://dx.doi.org/10.1016/j.contraception.2015.03.004 0010-7824/© 2015 Elsevier Inc. All rights reserved.

medication abortion requires follow-up to confirm procedure completion, which traditionally consists of an ultrasound to look for the absence of a gestational sac or serial serum human chorionic gonadotropin (hCG) testing with an 80% hCG drop from baseline to follow-up 7 to 14 days later [2,3]. Many women find this abortion follow-up burdensome. As such, follow-up rates are typically poor, as low as 45% [4–6]. Substantial time and resources are utilized in attempts to contact patients who miss their follow-up [5,6]. In an effort to optimize abortion care, recent investigations have sought to simplify medication abortion follow-up [6–16]. Earlier follow-up within the first few days of the medication abortion, especially for patients who may already be returning to clinic

2

K.D. Pocius et al. / Contraception xx (2015) xxx–xxx

as part of a work-up for pregnancy of unknown location, might improve patient ease and compliance with abortion completion confirmation. There are published early declines in hCG following spontaneous abortion [17–19] and surgically treated ectopic pregnancy [20,21]. However, similar studies in medication abortion are based on historic medication abortion regimens or have been largely limited to hCG values beyond Day 7, with Day 1 defined as day of mifepristone administration [2,22,23]. Three published studies on the current medication abortion regimen of mifepristone and misoprostol report serum hCG values collected before Day 7 [2,24,25]. The largest study included 34 women but did not report the clinically useful hCG decline from initial hCG at time of mifepristone administration [24]. The other two studies together include data from a total of 14 women and demonstrate a mean hCG decline of 60% between Day 1 and Day 4 and 91% between Day 1 and Day 6 [2,25]. The purpose of this study was to provide a more detailed examination of the decline in serum hCG values on Day 2 to 6 after medication abortion with the current evidencebased regimen.

2. Materials and methods This is a retrospective review of patients with early intrauterine pregnancies who underwent medication abortion with mifepristone and misoprostol at Planned Parenthood League of Massachusetts (PPLM) between January 2008 and November 2011. We queried the clinic billing database to identify all medication abortion patients with an early intrauterine pregnancy who had a Day 1 (day of mifepristone administration) serum hCG level and a repeat serum hCG level drawn on Day 2 to 6. All patients had a transvaginal ultrasound prior to administration of mifepristone to diagnose gestational age by mean gestational sac diameter using the Hadlock formula. For this study, an early intrauterine pregnancy is defined as an intrauterine gestational sac without a yolk sac on transvaginal ultrasound examination. All patients received 200-mg oral mifepristone in the clinic on Day 1 followed by 800-mcg buccal misoprostol self-administered 24 to 48 h later at home. In accordance with clinic protocol, patients without a yolk sac visualized on ultrasound were asked to return for a repeat hCG within 72 h after misoprostol administration. Further work-up with additional hCG testing was sometimes requested at the discretion of the follow-up clinician, resulting in repeat serial hCG testing in some patients. All hCG assays were performed at PPLM or an affiliated laboratory with the same hCG assay to ensure data comparability between sites. The primary outcome was percent decline between baseline and repeat serum hCG values on Days 2 through 6. The time between hCG evaluations was measured in whole days since specific times of mifepristone administration were not documented. For this study, successful medication abortion was defined as complete abortion after a single dose of

misoprostol with no further intervention, such as repeat misoprostol or uterine evacuation. This study was approved by the Partners Healthcare Institutional Review Board. The percent hCG decline in relation to Day 1 hCG values was calculated for all repeat values for study participants. Mean values are presented with standard deviation (SD). Repeated measure analysis of variance was used to estimate least square means and 95% confident intervals for the percent decline in hCG values in order to account for multiple measurements from the same woman. This test also accounts for missing data on each day from the full cohort, allowing for a more precise mean with less variability from the full population. Each repeat hCG value for each patient was plotted against time using a locally weighted scatter plot smoothing method. All statistical tests were performed with SAS statistical software, release 9.3 (SAS Institute Inc., Cary, NC, USA).

3. Results During the study period, 16,385 medication abortions were performed at PPLM. Eighty-eight patients were identified to have had a medication abortion for an early intrauterine pregnancy, a serum hCG level drawn on Day 1 and at least one repeat value on Day 2 through 6. The range of gestational ages on Day 1 by transvaginal ultrasound was 4.0 to 7.4 weeks with a median of 5.5 weeks. The range of baseline hCG varied widely from 352 to 153,952 IU with a median and upper and lower quartiles of 5220 (2181–9919) IU. All Day 1 hCG values were less than 76,000 IU except one. The percent of hCG decline in relation to each patient's Day 1 hCG is presented in Fig. 1 and Table 1. Fifteen (17%) of the total 88 patients had a repeat hCG value on Day 3. Of these patients with Day 3 hCG values, the mean decline in hCG from Day 1 was 56.9%±29.5% SD. One patient had an up-trending Day 3 hCG value, but through continued hCG monitoring, she was ultimately diagnosed to have had a successful medication abortion without further intervention at final follow-up. Forty patients (45%) had a repeat hCG value on Day 4. Of these, the mean decline in hCG from Day 1 was 73.5%±38.6% SD. One patient had a steeply up-trending Day 4 hCG value (240% of baseline) in the setting of a presumed failed medication abortion, but she was lost to further follow-up. A second patient had an up-trending hCG value on Day 4 (107% of baseline) but ultimately went on to have a successful abortion without further intervention. The mean percent decline by Day 5 was 86.1±8.8 and by Day 6 was 92.9±3.4. Eighty-two of the 88 patients (93%) had a successful medication abortion with one dose of misoprostol without need for further intervention. The least square means with 95% confidence interval (CI) of hCG decline for the successful abortion group was 57.6% (50.3–64.9%) on Day 3, 78.9% (75.0–82.8%) on Day 4 and 86.2% (81.3– 91.1%) on Day 5. The least square means of hCG decline on Day 3 for the failed abortion group was 54.7% (42.6–66.8%),

K.D. Pocius et al. / Contraception xx (2015) xxx–xxx

Fig. 1. Percent of serum hCG decline from Day 1 for all patients (N=88). Each repeat serum hCG value is plotted as a percent of the initial hCG value drawn on Day 1. A line of best fit is shown, created with a locally weighted scatter plot smoothing method. Repeat values from a patient with a successful medication abortion are represented with (+), while the values from the six patients with medication abortion failure are represented with (•).

as calculated from four patients' hCG values. A least square mean could not be calculated for Day 4 or Day 5 in the failed abortion group as only one patient with a failed procedure had an hCG value for each of those days. Of the six patients diagnosed as abortion failure requiring intervention, one had an up-trending hCG pattern. This patient was presumed to have an ongoing pregnancy in the setting of a 240% rise in hCG between Day 1 and Day 4 but was lost to further follow-up. One patient received a repeat dose of misoprostol based on lack of bleeding after her first dose and did not have a repeat hCG prior to this treatment. The remaining four patients were diagnosed with incomplete abortions based on follow-up ultrasound findings despite declining hCG levels (63%, 63%, 67% and 86% hCG decline between Day 1 and 3, respectively) [Fig. 1]. Of the 82 patients who had enough follow-up to confirm medication abortion success, many patients missed one or more of the recommended follow-up lab draws throughout the abortion evaluation. Forty-eight patients (55%) completed all provider-requested follow-up. Twenty-nine (33%) missed one, and 11 (13%) missed two or more follow-up appointments. Documented reasons for missed visits included confusion about follow-up recommendations, travel issues and patient belief that she had passed the pregnancy and that follow-up was therefore unnecessary.

3

medication abortion regimens [22]. These findings expand upon smaller studies of early hCG trends after medication abortion [2,24,25]. This may help guide the clinical management of those who present for care soon after taking misoprostol, including patients undergoing concurrent pregnancy of unknown location evaluation. Ultimately, the SDs in our data on Day 3 and Day 4 are too large to comment on the percent of hCG decline on those days to predict abortion success. By Day 5, 24 of the 25 patients had at least a 50% decline in hCG values, with the mean and median decline of over 80%. The specificity of serial hCG values in the first week following medication abortion to predict procedure success or failure has yet to be determined. Our study was not equipped to fill this gap in the literature because our sample size was too small to account for the low occurrence of procedure failures. Our study was restricted to a relatively homogenous population of women with early intrauterine pregnancies (defined as presence of a gestational sac and absence of a yolk sac), and therefore, our findings may not be applicable to other populations such as women with later gestations or pregnancies of unconfirmed location. In addition, the retrospective nature of our data precludes the ability to obtain precise timing of hCG serum collection in relation to mifepristone and misoprostol administration and pregnancy expulsion, which may increase the variability in the percent hCG declines between our patients. Lastly, it is worth noting the circular reasoning inherent to this type of study. Because repeat hCG measures are often used to help diagnose medical abortion success and failure, any comparisons of hCG trends between successful and failed medication abortion patients may be problematic when studied retrospectively if hCG trends are used to define the abortion success or failure. In this study, however, with the exception of the one patient with the up-trending hCG who was lost to follow-up, the other five patients were diagnosed with failed medication abortion based on ultrasound findings or symptoms as opposed to hCG trends. In conclusion, there is a rapid decline in serum hCG within the first few days after successful medication abortion with mifepristone and misoprostol, which becomes predictable with little variability as of Day 5. Ultimately, a prospective study with more standardized protocols, documentation of timing of medication administration and suspected expulsion, and defined criteria for success is needed to more precisely describe early hCG trends after successful medication abortion. Such information could help simply medication abortion follow-up and, thus, improve access to early, safe abortion.

4. Discussion

Acknowledgements

We found that there is a rapid decline in serum hCG within the first few days after successful medication abortion with mifepristone and misoprostol in early pregnancies similar to that seen with spontaneous abortion [17–19], surgically treated ectopic pregnancy [20,21] and other

The authors thank Elizabeth Driehaus of PPLM for creating code to search the billing data base to identify patients. The findings and conclusions in this article are those of the authors and do not necessarily represent the views of Planned Parenthood Federation of America, Inc.

4

K.D. Pocius et al. / Contraception xx (2015) xxx–xxx

Table 1 Percent hCG decline compared to Day 1 for all subjects (N= 88). Day after mifepristone

n

Mean % hCG decline from Day 1±SD

Median % hCG decline from Day 1 [range]

Least Square Mean % hCG decline from Day 1 (95% CI)

3 4 5 6

15 40 25 10

56.9±29.5 73.5±38.6 86.1±8.8 92.9±3.4

69.5 [− 14.6 to 90.8] 82.1 [− 140.7 to 93.0] 88.5 [55.0 to 94.6] 93.9 [85.9 to 96.8]

56.7 (46.3 to 67.18) 73.7 (67.2 to 80.1) 86.0 (78.0 to 94.2) 92.7 (79.9 to 105.5)

References [1] Jones RK, Jerman J. Abortion incidence and service availability in the United States, 2011. Perspect Sex Reprod Health 2014;46(1):3–14. [2] Fiala C, Safar P, Bygdeman M, Gemzell-Danielsson K. Verifying the effectiveness of medical abortion: ultrasound versus hCG testing. Eur J Obstet Gynecol Reprod Biol 2003;109(2):190–5. [3] American College of Obstetricians and Gynecologists. Medical management of first-trimester abortion. Practice bulletin no. 143. Obstet Gynecol 2014;123:676–92. [4] Haimov-Kochman R, Arbel R, Sciaky-Tamir Y, Brezezinski A, Laufer N, Yagel S. Risk factors for unsuccessful medical abortion with mifepristone and misoprostol. Acta Obstet Gynecol 2007;86:462–6. [5] Horning WL, Chen BA, Meyn LA, Creinin MD. Comparison of medical abortion follow-up with serum human chorionic gonadotropin testing and in-office assessment. Contraception 2012;85:402–7. [6] Clark W, Bracken H, Tanenhaus J, Schweikert S, Lichtenberg ES, Winikoff B. Alternatives to a routine follow-up visit for early medical abortion. Obstet Gynecol 2010;115(2 Pt 1):264–72. [7] Grossman D, Grindlay K. Alternatives to ultrasound for follow-up after medication abortion: a systemic review. Contraception 2011;83:504–10. [8] Clark W, Panton T, Hann L, Gold M. Medication abortion employing routine sequential measurements of serum hCG and sonography only when indicated. Contraception 2007;75(2):131–5. [9] Clark WH, Gold M, Grossman D, Winikoff B. Can mifepristone medical abortion be simplified? A review of the evidence and questions for future research. Contraception 2007;75(4):245–50. [10] Perriera LK, Reeves MF, Chen BA, Hohmann HL, Hayes J, Creinin MD. Feasibility of telephone follow-up after medical abortion. Contraception 2010;81:143–9. [11] Bracken H, Clark W, Lichtenberg ES, Schweikert SM, Tanenhaus J, Barajas A, et al. Alternatives to routine ultrasound for eligibility assessment prior to early termination of pregnancy with mifepristonemisoprostol. BJOG 2011;118(1):17–23. [12] Dayananda I, Maurer R, Fortin J, Goldberg AB. Medical abortion follow-up with serum human chorionic gonadotropin compared with ultrasonography: a randomized controlled trial. Obstet Gynecol 2013;121(3):607–13. [13] Blum J, Shochet T, Lynd K, Lichtenberg S, Fischer D, Arnesen M, et al. Can at-home semi-quantitative pregnancy tests serve as a replacement for

[14]

[15]

[16]

[17]

[18]

[19]

[20]

[21]

[22]

[23]

[24]

[25]

clinical follow-up of medical abortion? A US study. Contraception 2012;86:757–62. Ngoc NT, Bracken H, Blum J, Nga NT, Minh NH, van Nhang N, et al. Acceptability and feasibility of phone follow-up after early medical abortion in vietnam: a randomized controlled trial. Obstet Gynecol 2014;123(1):88–95. Bracken H, Lohr P, Taylor J, Morroni C, Winikoff B. RU OK? The acceptability and feasibility of remote technologies for follow-up after early medical abortion. Contraception 2014;90:29–35. Michie L, Cameron ST. Simplified follow-up after ealy medical abortion:12-month experience of a telephone call and self-performed low-sensitivity urine pregnancy test. Contraception 2014;89:440–5. Marrs RP, Kletzky OA, Howard WF, Mishell DR. Disapearance of human chorionic gonadotropin and resumption of ovulation following abortion. Am J Obstet Gynecol 1979;135(6):731–6. Barnhart K, Sammel MD, Chung K, Zhou L, Hummel AC, Guo W. Decline of serum human chorionic gonadotropin and spontaneous complete abortion: defining the normal curve. Obstet Gynecol 2004;104(5 Pt 1):975–81. Barnhart KT, Bader T, Huang X, Frederick MM, Timbers KA, Zhang JJ. Hormone pattern after misoprostol administration for a nonviable first-trimester gestation. Fertil Steril 2004;81(4):1099–105. Letterie GS, Hibbert M. Serial serum human chorionic gonadotropin (hCG) levels in ectopic pregnancy and first trimester miscarriage. Arch Gynecol Obstet 2000;263(4):168–9. Rempen A, Haubitz I. Longitudinal observations of human chorionic gonadotropin in serum following surgery for ectopic pregnancy. Arch Gynecol Obstet 1996;258(4):181–92. Creinin MD. Change in serum beta-human chorionic gonadotropin after abortion with methotrexate and misoprostol. Am J Obstet Gynecol 1996;174(2):776–8. Parashar P, Iversen OE, Midboe G, Myking O, Bjorge L. Medical abortion in the first trimester: the use of serum hCG and endometrial thickness as markers of completeness. Eur J Contracept Reprod Health Care 2007;12(4):366–71. Honkanen H, Ranta S, Ylikorkala O, Heikinheimo O. The kinetics of serum hCG and progesterone in response to oral and vaginal administration of misoprostol during medical termination of pregnancy. Hum Reprod 2002;17(9):2315–9. Walker K, Schaff E, Fielding S, Fuller L. Monitoring serum chorionic gonadotropin levels after mifepristone abortion. Contraception 2001;64(5):271–3.

Early serum human chorionic gonadotropin (hCG) trends after medication abortion.

Despite increased reliance on human chorionic gonadotropin (hCG) for early pregnancy monitoring, there is limited information about hCG trends soon af...
186KB Sizes 1 Downloads 8 Views