Original Research ajog.

org

GYNECOLOGY
Risk of complication during surgical abortion
in obese women
Katrina S. Mark, MD; Barbara Bragg, CRNP; Tara Talaie, MS; Kiran Chawla, MD; Latasha Murphy, MD;
Mishka Terplan, MD, MPH

BACKGROUND: Surgical abortion is a generally safe procedure.
Obesity is a known risk factor for complications in other surgical
procedures, but insufficient information exists to determine the effects of
increasing body mass index on the risk of surgical abortions.
OBJECTIVE: The purpose of this study was to determine whether
obesity is a risk factor for major complications in surgical abortions.
METHODS: A quality control database from a single outpatient center
was analyzed to determine rates of major complications during surgical
abortions in relation to obesity class. Complications included hemorrhage,
need for repeat evacuation, uterine perforation, cervical laceration,
medication reaction, unexpected surgery, or unplanned admission to the
hospital. Chi-squared and analysis of variance were used for analysis.
RESULTS: We included 2468 procedures: 1475 procedures
(59.8%) in the first trimester and 993 procedures (40.2%) in the
second trimester. The overall complications rate was 2.2%. Second-
trimester procedures were more likely than those in the first trimester
to have complications (3.1% vs 1.6%; P¼.009). Overall, 39.6% of
the women were obese, and 9.6% of them met criteria for class 3
obesity (body mass index, >40 kg/m2). Women who underwent
second-trimester abortions with class 3 obesity had a rate of
complication of 8.7%, which was significantly more than normal
weight women (odds ratio, 5.90; 95% confidence interval,
1.93e8.07; P<.001).
COMMENT: Surgical abortions are overall safe procedures, but class
3 obesity increases the rate of complication in second-trimester
procedures.
Key words: complication, obesity, surgical abortion

I n 2014, 19% of all pregnancies in the

United States ended in abortion.1
With such wide scale prominence, the
safety of abortion procedures has been a
major topic of focus. Over the years,
studies repeatedly have found induced
abortion procedures to be very safe.1-8
Major complication rates in most
studies are 0.5e2%2,3,8; the mortality
rate is 0.6 per 100,000.9 To put this into
perspective, the risk of death from
childbirth in the United States is esti-
mated to be 14 times higher than the risk
from induced abortion.9
There are known risk factors that
increase the mortality rate of surgical
terminations, the most studied of which
is gestational age. According to an anal-
ysis of the Abortion Mortality Surveil-
lance System, the risk of death increases
by 38% with each additional week
gestation.3 Similarly, the risk of major
morbidity also increases with increasing
Cite this article as: Mark KS, Bragg B, Talaie T, et al.
Risk of complication during surgical abortion in obese
women. Am J Obstet Gynecol 2018;218:238.e1-5.
0002-9378/$36.00
ª 2017 Elsevier Inc. All rights reserved.
https://doi.org/10.1016/j.ajog.2017.10.018
gestational age.4,5,10 Fortunately, the vast
majority of women who undergo
termination are at a gestational age of
13 weeks; only 8.4% of abortions occur
in the second trimester.11
Obesity is a known risk factor for
many poor pregnancy outcomes as well
as for complications of surgical proced-
ures.12-14 Because approximately one-
third of adult women are obese15 and
obese women have a higher rate of
unplanned pregnancy,16 it is important
to understand the effects of obesity on
the risk of complications from abortion.
Two recent studies evaluated obesity as a
risk factor for surgical abortions. One
study focused on second-trimester dila-
tion and evacuation procedures and
found no difference in complications
that were related to increasing body mass
index (BMI) with procedures that were
performed in a hospital-based setting,
many of which were performed in the late
second trimester.10 The second study
included first- and second-trimester
surgical abortions and also found no in-
crease in complications with increasing
BMI, but included mostly first-trimester
terminations and had a low rate of class
III obesity.2 To date, there is limited evi-
dence on the safety of second-trimester
terminations in obese women in ambu-
latory procedure centers.
The aim of this study was to evaluate
the relationship of obesity to procedural
complications for surgical terminations
in the first and second trimester in an
ambulatory center.
Methods
This study is a retrospective review of
women who underwent procedures for
termination of pregnancy at a single site
between October 2009 and June 2014.
The analysis was performed with the use
of a quality improvement database that
was kept prospectively by the nurse
practitioner who managed the clinic.
The study was determined by the Uni-
versity of Maryland Institutional Review
Board to be exempt.
Patients at the clinic undergo a pre-
procedure visit with a trained nurse
practitioner or physician for counseling
and evaluation. The clinic has a protocol
for preoperative evaluation in which the
nurse practitioner and physician deter-
mine which patients are appropriate for
in-office procedures with moderate
sedation and which patients require
anesthesiologist participation and an
operating room setting. For second-

238.e1 American Journal of Obstetrics & Gynecology FEBRUARY 2018

ajog.org GYNECOLOGY Original Research

genetic if they were done for a fetal

TABLE 1
anomaly or because of an embryonic
Characteristics of women undergoing abortion procedures
gestation. All other procedures, including
Total No complications Complications those performed for maternal indications,
Variable (n¼2468) (n¼2414) (n¼54) P value were classified as “nongenetic.” Trimesters
Body mass index, kg/m2a 29.3  7.8 29.3  7.8 29.8  8.4 .642 were categorized as first or second
trimester where second trimester was
Age, y a
25.9  6.5 25.9  6.5 25.8  6.0 .877
considered >13 completed weeks of
Graviditya 4.3  2.8 4.3  2.8 4.4  2.5 .937 gestation. The primary exposure for anal-
Paritya 1.3  1.4 1.4  1.5 1.3  1.3 .727 ysis was BMI. Because BMI was not
Previous cesarean delivery b
.355 included in the initial dataset, a chart
review was performed to determine the
Yes 637 (25.8) 620 (25.7) 17 (31.5)
BMI for all participants in the cohort. This
No 1831 (74.2) 1794 (74.3) 37 (68.5) was calculated by the weight and height
Previous cesarean delivery, n .405c from the patient’s chart on the day of the
1 378 367 11 procedure and classified into categories
based on the World Health Organization
2 172 167 5
classification17 of underweight (BMI, <
3 61 61 0 18.5 kg/m2), normal weight (BMI,
4 21 21 1 18.5e24.9 kg/m2), overweight (BMI,
5 2 2 0 25.0e29.9 kg/m2), obese class I (BMI,
30e34.9 kg/m2), class II (BMI, 35e39.9
6 1 1 0
kg/m2) and class III (BMI, 40 kg/m2).
All 651 634 17 Our primary outcome was any
Previous therapeutic 1.3  1.7 1.3  1.7 1.1  1.6 .474 complication that included hemorrhage
abortion, na (defined as >500 mL estimated blood
Gestational age, moa 12 4/7  4.3 12 3/7  4.3 13 5/7  4.7 .031 loss), need for repeat evacuation, uterine
perforation, cervical laceration, medica-
Geneticb .059
tion reaction, unexpected surgery, or
Yes 103 (4.2) 98 (95.1) 5 (4.9) unplanned admission to the hospital.
No 2365 (95.8) 2316 (97.9) 49 (2.1) The outcome was double recorded in the
Surgeon b
.791 dataset: (1) recording “yes/no” for
complication and (2) recording the type
Resident 1078 (43.7) 1054 (97.8) 24 (2.2)
of complication. If any discrepancy was
Attending 1390 (56.3) 1360 (97.8) 30 (2.2) found between the yes/no variable and
a b c
Data are given as meanstandard deviation; Data are given as number (percent); Analysis of variance. the information recorded for type of
Mark et al. Surgical abortion risk and obesity. Am J Obstet Gynecol 2018.
complication, a chart review was per-
formed to determine whether a compli-
cation had occurred. Additionally, chart
trimester procedures (those at >13 Procedures were performed by either reviews were performed for those par-
completed weeks gestation), laminaria an attending or resident obstetrician/ ticipants who did have a complication to
placement was performed routinely gynecologist under the supervision of an confirm the complication.
18e24 hours before the procedure. attending physician. First-trimester The initial data set was maintained in
Second-trimester procedures at >20 procedures were performed by resi- an Excel (Microsoft Corporation, Red-
weeks of gestation were performed only dents of all levels, and second-trimester mond, WA) spreadsheet. Bivariate anal-
after review and approval by an ethics procedures were performed by third- ysis was performed with chi-squared
board whose members agreed that there and fourth-year residents. analysis, Fisher’s exact test, and analysis
was a maternal or fetal indication for The following demographic and pro- of variance with MedCalc statistical
termination. All procedures in the cedural data were included in the data set: software (version 17.9.7; MedCalc Soft-
outpatient clinic were performed with maternal age; gestational age; obstetric ware, Mariakerke, Belgium). Data were
moderate sedation without intubation history that included gravidity, parity, exported into Stata software (version 14;
with the use of midazolam and fentanyl; number of previous cesarean deliveries, Stata Corporation, College Station, TX)
if the patient requests or has contrain- and history of abortions; operator who for logistic regression. Variables that
dications, procedures were done without performed the procedure, and whether the were associated with both the exposure
sedation. All patients receive appropriate termination was done for genetic indi- and the outcome were deemed eligible
antibiotic prophylaxis. cations. Procedures were considered for inclusion into the adjusted model.

FEBRUARY 2018 American Journal of Obstetrics & Gynecology 238.e2

Original Research GYNECOLOGY ajog.org

transfusions. The remaining complica-

FIGURE
tions included 5 additional reevacua-
Rates of procedure complication of body mass index category
tions, a cervical laceration that required
10 repair, and a uterine perforation that
9 required laparotomy, blood transfusion,
8 and admission to the hospital.
7 There was no difference between
6 those patients with and without com-
Complicaon plications by any of the cofactors, with
5
Rate (%) All
the exception of mean gestational age.
4
3
First trimester Whereas the mean gestational age
2 Second trimester among those patients without compli-
1 cations was 12 3/7, it was 13 5/7among
0 those patients with complications
(P¼.031). More women who underwent
first-trimester procedures were obese
(44.4% vs 32.0% in the second trimester;
Rates of all complications in all, first trimester, and second trimester procedures compared to P<.001), which included those patients
increasing body mass index category. with class III obesity (12.1% vs 8.0%,
Mark et al. Surgical abortion risk and obesity. Am J Obstet Gynecol 2018. respectively; P<.001).
Residents performed 1078 procedures
(43.7%), and attending physicians per-
Results were reported as odds ratio (OR) Overall there were 54 complications, formed 1390 procedures (56.3%).
with 95% confidence interval (CI). for a complication rate of 2.2%. Second- Attending physicians performed more of
trimester procedures had a higher the second-trimester procedures (62.1%
Results complication rate than first-trimester vs 37.2%). However, there was no sig-
A total of 2502 procedures were per- procedures (3.1% vs 1.6%; P¼.009). A nificant difference in the complication
formed during the timeframe of the logistic regression was performed, and rates when we compared procedures that
study. Thirty-four of them were there was an increased risk for every in- were performed by attending physicians
excluded because BMI data was not cremental increase in gestational age and residents in the first (1.43% vs
available, which resulted in 2468 pro- (OR, 1.08; 95% CI, 1.02e1.16). 1.69%; P¼.696) or second trimesters
cedures available for analysis. Seven Of the 23 complications that occurred (3.04% vs 3.25%; P¼.847).
participants had a discrepancy in the in first-trimester procedures, 18 of them The Figure shows complications by
dataset regarding presence or absence of (78.3%) were immediate reevacuations, BMI category with a linear increase in
a complication: 5 participants had “yes” only 2 of which were complicated by complications with increasing BMI
recorded for complication but no type hemorrhage. The additional 5 first- category in second-trimester procedures.
recorded, and 2 participants had “no” trimester procedures included 1 hem- Table 2 shows the complication rate by
recorded for complication but subse- orrhage without reevacuation, 1 patient BMI category overall and stratified by
quently had a complication type recor- with an inability to tolerate the proced- trimester. There was no statistical differ-
ded. On review of the medical records, ure in the outpatient setting who had to ence found between BMI categories
we determined that none of these be taken to the operating room for when first- and second-trimester pro-
patients had complications and subse- general anesthesia, 1 patient whose cedures were analyzed together (Table 2).
quently were all included and analyzed cervix was unable to be dilated When stratified by trimester, overall there
in the “no complication” group. adequately, 1 anesthesia complication does not appear to be a linear increase in
Overall 31.5% of the women were that required prolonged observation, complication with increasing BMI cate-
normal weight; 2.3% of them were un- and 1 patient with uterine perforation gory among first-trimester procedures
derweight; 26.6% of them were over- that was managed laparoscopically and (Figure). Complications increased by
weight, and 39.6% of them were obese, who was discharged the same day. Of the BMI category in second-trimester pro-
which included 9.6% with a BMI >40 31 complications in second-trimester cedures. Women with class III obesity
kg/m2. First-trimester terminations procedures, 24 complications (77.4%) had the highest stratified complication
accounted for 59.8% (1475) of all pro- were hemorrhages, of which 6 compli- rate of 8.7%. This difference was found to
cedures; 40.2% of the women (993) cations required reevacuation, 6 patients be statistically significant when we
underwent second-trimester dilation were admitted to the hospital, and 1 compared it with that of normal weight
and evacuations, which included 0.8% of complication was related to a cervical women, with an odds ratio of 5.90 (95%
procedures (19) that occurred 20 laceration that required repair. None of CI, 1.93e8.07; P<.001). We performed a
weeks gestation. (Table 1). the hemorrhages required blood logistic regression that compared

238.e3 American Journal of Obstetrics & Gynecology FEBRUARY 2018

ajog.org GYNECOLOGY Original Research

therefore, we were able to identify com-

TABLE 2
plications in this category more easily.
Abortion complication rates by body mass index and procedure trimester
Additionally, their procedures were all
Rate of complications, Odds ratioa performed in a hospital setting, many of
Body mass index, kg/m2 n/N (%) (95% confidence interval) the patients were under deep sedation
All procedures that was administered by an anesthesi-
ologist. We excluded women in this
Underweight 0/57 (0) 0.40 (0.02e6.78)
analysis who were deemed inappropriate
18.5e24.9 (normal) 16/777 (2.1) N/A for an ambulatory procedure and/or
25e29.9 (overweight) 16/657 (2.4) 1.19 (0.59e2.40) who required more than nurse-
30e34.9 (class 1) 9/449 (2.0) 0.97 (0.43e2.22) administered conscious sedation
because our intention was to determine
35e39.9 (class 2) 4/270 (1.5) 0.72 (0.24e2.16)
the safety of a truly ambulatory setting.
>40 (class 3) 9/258 (3.5) 1.71 (0.75e3.93) In addition, more than one-half of their
Total 54/2468 (2.2) N/A procedures were done at a gestational age
Firstetrimester procedures past 20 weeks, which likely represents a
different baseline risk level.
Underweight 0/37 (0) 0.51 (0.03e8.79)
Benson et al,2 who evaluated 4968
18.5e24.9 (normal) 10/407 (2.5) N/A women who underwent first- and
25e29.9 (overweight) 7/375 (1.9) 0.76 (0.29e2.00) second-trimester surgical abortions, also
30e34.9 (class 1) 4/301 (1.3) 0.53 (0.17e1.72) found no difference in complication rate
based on BMI in either trimester. They
35e39.9 (class 2) 0/176 (0) 0.11 (0.01e1.84)
had a lower rate of obesity in their
>40 (class 3) 2/179 (1.1) 0.45 (0.10e2.07) cohort; 24.6% of their patients were
Total 23/1475 (1.6) N/A considered obese, and 3.7% had a BMI
Secondetrimester procedures >40 kg/m2 compared with 39.6% and
10.5% in our study, respectively. Their
Underweight 0/20 (0) 1.37 (0.07e25.12)
definition of complication was slightly
18.5e24.9 (normal) 6/370 (1.6) N/A different, which may account for the
25e29.9 (overweight) 9/281 (3.2) 2.01 (0.71e5.71) difference in conclusions. Although in
30e34.9 (class 1) 5/148 (3.4) 2.12 (0.64e7.06) our study we considered >500 mL of
blood loss to be a hemorrhage, they
35e39.9 (class 2) 4/95 (6.4) 2.66 (0.74e9.65)
defined hemorrhage >100 mL of blood
>40 (class 3) 7/79 (8.7) 5.90 (1.93e18.07) loss. Although >100 mL blood loss may
Total 31/993 (3.1) N/A be considered abnormal in many pro-
N/A, not applicable. cedures, we believed that an estimated
a
In comparison to normal weight women. blood loss of <500 mL was unlikely to be
Mark et al. Surgical abortion risk and obesity. Am J Obstet Gynecol 2018. clinically significant or require more re-
sources than would generally be available
in an ambulatory setting.
complication rates between BMI cate- complications of 2.2% was similar to More than 10% of the patients in our
gories in second-trimester procedures; previously published reports.2,6,10,18 cohort had a BMI >40 kg/m2. With rates
the increase in complications in women The finding of an increased compli- of obesity rising19 and high rates of un-
with class III obesity remained signifi- cation rate in women in the second met family planning needs in obese
cant, even when we controlled for in- trimester with category III obesity is women,20 this is not a trend that is likely
cremental increases in gestational age new. Two recent studies did not find to reverse anytime soon. Because obesity
(OR, 5.04; 95% CI, 1.65e15.39). differences in complication rates at any is a known factor in delaying a diagnosis
BMI cutoff.2,10 Lederle et al10 reviewed of pregnancy and presentation for
Comment 4520 dilation and evacuations that were abortion services,21 second-trimester
Surgical abortions are safe procedures performed over a 4-year period in a procedures may be increasingly more
with infrequent major complications. hospital setting and found no difference common in obese women.
The findings of our study agree with in complication rates related to BMI. Our study has several limitations.
previous evidence that supports the One-quarter of the cohort in the study Given the design of the study, certain
safety of these procedures in ambulatory by Lederle et al was obese compared with information such as complications after
settings and performed by skilled pro- almost 40% in our study. Our propor- discharge was not available. This likely
fessionals. Our overall rate of major tion of class III obesity was higher; underestimates the overall complication

FEBRUARY 2018 American Journal of Obstetrics & Gynecology 238.e4

Original Research GYNECOLOGY
rate in all BMI categories. Also, because
the study excluded women who had
procedures performed in the operating
room with anesthesiologist-administered
deep sedation or general anesthesia, this
particularly high-risk group was not
included in analysis; therefore, the true
rate of complications in obese women
may have been underestimated. However,
given that most ambulatory clinics likely
have a similar decision analysis regarding
appropriateness of a patient for an in-
office procedure, these results are likely
applicable to most practices.
This study confirms the overall safety
of surgical abortions in ambulatory
practices. The increased rate of compli-
cations found in second-trimester pro-
cedures in morbidly obese women
should not deter providers from offering
services to women, but rather encourage
early access to care. n abortion by dilation and evacuation. Obstet
References
1. Guttmacher Institute. Fact sheet 2017:
Induced abortions in the United States. Available
at: https://www.guttmacher.org/fact-sheet/
induced-abortion-united-states?gclid¼CJuvi47
ojdICFYGAfgodnbgDsA. Accessed February
13, 2017.
2. Benson LS, Micks EA, Ingalls C, Prager SW.
Safety of outpatient surgical abortion for obese
patients in the first and second trimester. Obstet
Gynecol 2016;128:1065-70.
3. Bartlett LA, Berg CJ, Shulman HB, et al. Risk
factors for legal induced abortion-related mor-
tality in the United States. Obstet Gynecol
2004;103:729-37.
4. Ferris LE, McMain-Klein M, Colodny N,
Fellows GF, Lamont J. Factors associated with
238.e5 American Journal of Obstetrics & Gynecology FEBRUARY 2018
immediate abortion complications. Can Med
Assoc J 1994;154:1677-85.
5. Beuhler JW, Schulz KF, Grimes DA,
Hogue CJ. The risk of serious complications
from induced abortion: do personal character-
istics make a difference? Am J Obstet Gynecol
1985;153:14-20.
6. White K, Carroll E, Grossman D. Complica-
tions from first-trimester aspiration abortion: a
systematic review of the literature. Contracep-
tion 2015;92:422-38.
7. Cates W, Schulz K, Grimes D, et al. Dilatation
and evacuation procedures and second-
trimester abortions. JAMA 1982;248:559-63.
8. Upadhyay U, Desai S, Zlidar V, et al. Incidence
of emergency department visits and complica-
tions after abortion. Obstet Gynecol 2015;125:
175-83.
9. Raymond EG, Grimes DA. The comparative
safety of legal induced abortion and childbirth in
the United States. Obstet Gynecol 2012;119:
215-9.
10. Lederle L, Steinauer JE, Montgomery A,
Aksel S, Drey EA, Kerns JL. Obesity as a risk
factor for complications after second-trimester
Gynecol 2015;126:585-92.
11. Jatlaoui TC, Ewing A, Mandel MG, et al.
Abortion surveillance: United States 2013.
MMWR Surveill Summ 2016;65:1-44.
12. American College of Obstetricians and
Gynecologists. Obesity in pregnancy. Practice
Bulletin no.: 156. Obstet Gynecol 2015;126:
e112-26.
13. Bouwman F, Smits A, Lopes A, et al. The
impact of BMI on surgical complications and
outcomes in endometrial cancer surgery: an
institutional study and systematic review of the
literature. Gynecol Oncol 2015;139:369-76.
14. Tjeertes EK, Hoeks SE, Beks SB,
Valentijn TM, Hoofwijk AG, Stolker RJ. Obesity: a
risk factor for postoperative complications in
general surgery? BMC Anesthesiol 2015;15:
112.
ajog.org
15. Ogden C, Carroll M, Kit B, Flegal K. Preva-
lence of childhood and adult obesity in the
United States, 2011-2012. JAMA 2014;311:
806-14.
16. McKeating A, O’Higgins A, Turner C,
McMahon L, Sheehan SR, Turner MJ. The rela-
tionship between unplanned pregnancy and
maternal body mass index 2009-2012. Eur J
Contracept Reprod Health Care 2015;20:409-18.
17. World Health Organization. WHO obesity
classification. Available at: http://apps.who.int/
bmi/index.jsp?introPage¼intro_3.html. Accessed
February13, 2017.
18. Peterson WF, Berry FN, Grace MR,
Gulbranson CL. Second-trimester abortion by
dilatation and evacuation: an analysis of 11,747
cases. Obstet Gynecol 1983;62:185-90.
19. National Institute of Diabetes and Digestive
and Kidney Disease. Overweight and obesity
statistics. Available at: https://www.niddk.nih.
gov/health-information/health-statistics/Pages/
overweight-obesity-statistics.aspx. Accessed
February 2, 2017.
20. Ingraham N, Roberts SC, Weitz TA. Prior
family planning experience of obese women
seeking abortion care. Womens Health Issues
2014;1:e125-30.
21. Foster DG, Jackson RA, Cosby K, Weitz TA,
Darney PD, Drey EA. Predictors of delay in each
step leading to abortion. Contraception
2008;77:289-93.
Author and article information
From the University of Maryland School of Medicine
(Drs Mark, Chawla, and Murphy and Ms Talaie) and the
University of Maryland Medical Center (Ms Bragg),
Baltimore, MD; Virginia Commonwealth University School
of Medicine, Richmond VA (Dr Terplan).
Received Aug. 16, 2017; revised Oct. 11, 2017;
accepted Oct. 16, 2017.
The authors report no conflict of interest.
Corresponding author: Katrina S. Mark, MD. kmark@
fpi.umaryland.edu