Contraception 94 (2016) 81 86

Original research article

Twelve-month discontinuation of etonogestrel implant in an outpatient

pediatric setting
Elise Berlan a, b,, Kelly Mizraji c , Andrea E. Bonny a, b
a
Section of Adolescent Medicine, Nationwide Children's Hospital, Columbus, OH
b
Department of Pediatrics, The Ohio State University College of Medicine, Columbus, OH
c
The Ohio State University College of Medicine, Columbus, OH
Received 28 September 2015; revised 24 February 2016; accepted 28 February 2016

Abstract

Objective: The etonogestrel (ENG) contraceptive implant is the most effective reversible contraceptive method. Uptake remains limited in
adolescents, a population at high risk for unintended pregnancy. The objectives of this study were to determine the 12-month discontinuation
rate of the ENG implant among adolescents in an outpatient setting and to characterize risk factors for discontinuation.
Study design: A retrospective chart review identified adolescent females aged 12 to 22 years who received the ENG implant in one pediatric
institution between January 1, 2011, and April 15, 2014. Patients were categorized into ENG discontinuers (removed prior to 12 months) and
ENG continuers (continued for 12 months). Associations between demographic, clinical and postplacement characteristics with ENG
discontinuation category were assessed with t tests, 2/Fisher's Exact Tests and backwards stepwise logistic regression.
Results: Of the 750 patients who had an ENG implant inserted, 77 (10.3%) had the device removed prior to 12 months of use. The mean
length of implant use for those who discontinued was 7.5 months. Problematic bleeding was the most commonly cited reason for
discontinuation. Older age at time of insertion, history of pregnancy and 1 medical visit for implant concerns (not including removal) were
independently predictive (p b .01) of method discontinuation.
Conclusion: The vast majority of adolescents continued the ENG implant at 12 months, making it an excellent contraceptive choice for
adolescents within the outpatient pediatric setting. Greater efforts should be made to increase its use by pediatric providers.
Implications: The ENG implant is an excellent contraceptive option for adolescents in the outpatient pediatric setting.
2016 Elsevier Inc. All rights reserved.

Keywords: Adolescents; Contraception; Implant; Etonogestrel; Pediatric; Pediatrician

1. Introduction device (IUD) at last intercourse [8]. American teens who do

Despite recent declines, the United States leads the
developed world in teen births [13]. US adolescents have
unacceptably high rates of pregnancy (~ 4/5 unintended) and
birth [4,5]. Low use of highly effective contraceptive
methods contributes to high teenage pregnancy rates [6,7].
In 2013, only 25.3% of sexually active high school students
reported use of a hormonal contraceptive or intrauterine
Funding: This work was supported by Research Data & Computing
Services at Nationwide Children's Hospital and the Center for Clinical and
Translational Science at The Ohio State University (CTSA grant
UL1TR001070) and by the OSU College of Medicine Bennett Research
Scholarship awarded to Ms. Mizraji.
Corresponding author. +1 614 772 2458
E-mail address: elise.berlan@nationwidechildrens.org (E. Berlan).
http://dx.doi.org/10.1016/j.contraception.2016.02.030
0010-7824/ 2016 Elsevier Inc. All rights reserved.
use contraceptives predominantly rely on pills and condoms,
which are relatively less effective but more commonly
prescribed [6,9,10]. Only a small minority of teens use
long-acting reversible contraceptives (LARCs), the most
effective methods available. In 20112013, approximately
5% of adolescent and young adult women aged 1524 using
birth control chose a LARC method [11].
LARC methods demonstrate first-year failure rates of
0.8%, 0.2% and 0.05% for the levonorgestrel intrauterine
system, copper IUD and etonogestrel (ENG) implant,
respectively [9]. Prior studies have established the superior
effectiveness, safety and acceptability of IUDs and the ENG
implant for adolescent females [1214]. The Institute of
Medicine and the Centers for Disease Control and
Prevention identify reducing unintended and teen pregnancy
as a national priority and call for increased utilization of
82 E. Berlan et al. / Contraception 94 (2016) 8186
LARCs [15,16]. In September 2014, the American Academy
of Pediatrics recommended that LARCs be first-line
contraceptive choices for adolescents [17].
Improving adolescents' access to LARCs in the primary
care pediatric office setting could increase uptake of LARC
methods by adolescents and lead to significant reductions in
teen pregnancy [18,19]. The ENG implant (Nexplanon)
may appeal to pediatricians due to ease and simplicity of
placement, perceived noninvasiveness and lack of need for
pelvic examination. However, pediatrician concern about
appropriateness of LARCs for adolescents may limit
adoption of new contraceptive practices. Few large studies
have evaluated adolescents' experiences with the ENG
implant, leaving pediatricians and adolescent health care
providers in need of more adolescent-specific evidence.
The objectives of this study were to (a) determine
12-month discontinuation and (b) characterize risk factors
for discontinuation of the ENG implant in a sample of
adolescents seen in a variety of outpatient clinical settings at
an urban children's hospital.
2. Methods
2.1. Participants and setting
The Nationwide Children's Hospital (NCH) Institutional
Review Board approved the study protocol. A retrospective
chart review was conducted of all adolescent females ages
1222 who received the ENG implant between January 1,
2011, and April 15, 2014, at any outpatient clinic within the
NCH system. Eligible patients were identified via Interna-
tional Classification of Diseases, Ninth Revision, codes
(V25.5, V25.43), Current Procedural Terminology codes
(11981, 11983) and medication orders for ENG implant.
Patient lists were compared for redundant patients.
2.2. Data collection
Data collection occurred in June 2015 so that all eligible
patients could have continued the ENG implant for a minimum
of 1 year. We reviewed each patient's ENG placement visit
and all subsequent medical visits and telephone contacts in the
year following placement. Study data were collected and
managed using REDCap hosted at NCH [20].
At time of ENG implant placement, information was
collected regarding patients' age, race/ethnicity, insurance
status, weight (kg), height (cm) and reason for ENG implant
placement. If height and weight were not recorded at the
placement visit, the closest recorded height and weight to
time of insertion were used. The type of medical provider
who placed the ENG implant was also noted. Additional
clinical factors that were collected on patients at time of ENG
placement included history of pregnancy, history of
gynecological problems, contraceptive method at time of
placement, contraceptive use in the 2 years prior and
sexually transmitted infection (STI) diagnosis in the 1 year
prior to placement. STI diagnosis was confirmed by
reviewing patient notes and laboratory results.
All telephone consultations and medical visits within the
year following placement were reviewed for relevance to ENG
implant. The total numbers of telephone consultations and
medical visits for ENG implant concerns, not including ENG
removal, were tabulated. Additional data collected included
prescription of temporizing measures [e.g., nonsteroidal
anti-inflammatory drugs (NSAIDs), hormonal medication]
and STI diagnoses in the year following ENG placement.
For those who underwent ENG implant removal, date of
removal was recorded. Patients were then categorized into ENG
discontinuers (i.e., removed the device prior to 12 months) and
ENG continuers (continued the device for 12 months or longer).
For ENG discontinuers, mean months of use was calculated.
Additional clinical information collected on ENG discontinuers
included reason for removal, type of provider performing
removal and patient contraceptive choice after removal.
2.3. Statistical analysis
The main outcome measure was ENG implant discontin-
uation at 12 months. The unadjusted associations between
subject, clinical and postplacement characteristics with ENG
continuation category were assessed with t tests for
continuous variables and 2 /Fisher's Exact Tests for
categorical variables.
All demographic, clinical and postplacement characteris-
tics associated with ENG continuation at p value b 0.1 were
eligible for multivariable, stepwise logistic regression
modeling predicting ENG implant discontinuation prior to
12 months. The age cutoff of 16.8 years was chosen via
receiver operating characteristic curve to maximize the
positive predictive value for discontinuation and to minimize
the false-positive predictive value. The final adjusted logistic
regression model retained all variables associated with ENG
implant discontinuation at a p value b .05. Odds ratios were
calculated for retained variables.
3. Results
Of the 750 patients who had an ENG implant inserted
during the study period, 77 (10.3%) had the device removed
prior to 12 months. Mean age was approximately 1 year
older (p b .001) for those who discontinued as compared to
those who continued the method (Table 1). ENG disconti-
nuers were significantly more likely to have had a prior
pregnancy (p = .004) and to have used one or more
contraceptive methods in the preceding 2 years (p = .048).
Postplacement, ENG discontinuers were more likely to
contact the clinic or have medical visits for ENG implant
concerns. Temporizing measures, particularly hormonal
 E. Berlan et al. / Contraception 94 (2016) 8186 83

Table 1
Unadjusted associations of subject characteristics with ENG implant continuation
Subject characteristic ENG Implant Continuation a (n =673) ENG Implant Discontinuation b (n= 77) p value c
Demographic characteristics
Age (years) at insertion, mean (SD) 16.9 (1.9) 17.7 (1.6) b .001
Race/ethnicity, n (%) .596
Caucasian 213 (31.6) 23 (29.9)
African-American 382 (56.8) 47 (61.0)
Asian 4 (0.6) 0 (0.0)
Hawaiian/Pacific Islander 4 (0.6) 0 (0.0)
Biracial 40 (5.9) 6 (7.8)
Hispanic/Latino 10 (1.5) 1 (1.3)
Other 8 (1.2) 0 (0.0)
Unknown 12 (1.8) 0 (0.0)
Insurance status, n (%) .196
Public 490 (72.8) 56 (72.7)
Private 158 (23.5) 15 (19.5)
None 25 (3.7) 6 (7.8)
Clinical characteristics
Provider type, n (%) .364
Adolescent Medicine 490 (72.8) 52 (67.5)
Pediatric Primary Care 176 (26.2) 23 (29.9)
Other specialty clinic 7 (1.0) 2 (2.6)
BMI categorization, n (%) .209
Normal 280 (41.6) 37 (48.1)
Overweight 162 (24.1) 20 (26.0)
Obese 228 (33.9) 18 (23.4)
Missing 3 (0.4) 2 (2.6)
Reason for device, n (%) .960
Noncontraceptive 13 (1.9) 1 (1.3)
Contraceptive 627 (93.2) 73 (94.8)
Both 32 (4.8) 3 (3.9)
Missing 1 (0.1) 0 (0.0)
History of pregnancy, n (%) .004
No 581 (86.3) 57 (74.0)
Yes 92 (13.7) 20 (26.0)
History of STI d 1 year prior, n (%) .276
No 558 (82.9) 60 (77.9)
Yes 115 (17.1) 17 (22.1)
Gynecological problem, e n (%) .877
No 495 (73.6) 56 (72.7)
Yes 178 (26.4) 21 (27.3)
Past 2-year contraceptive use, n (%) .048
None 143 (21.2) 9 (11.7)
1 or more 530 (78.8) 68 (88.3)
Method at time of placement, n (%) .979
None 212 (31.5) 26 (33.8)
Progestin-only pill 6 (0.9) 1 (1.3)
Combined hormonal contraception 108 (16.0) 11 (14.3)
Depot medroxyprogesterone acetate 337 (50.1) 38 (49.4)
ENG implant 7 (1.0) 1 (1.3)
Hormonal IUD 3 (0.4) 0 (0.0)
Postplacement characteristics
STI d during 1st year of use, n (%) .542
No 601 (89.3) 67 (87.0)
Yes 72 (10.7) 10 (13.0)
Telephone consultations, f n (%) .006
None 638 (94.8) 67 (87.0)
1 or more 35 (5.2) 10 (13.0)
Medical visits, f n (%) b .001
None 546 (81.1) 23 (29.9)
1 or more 127 (18.9) 54 (70.1)
(continued on next page)
84 E. Berlan et al. / Contraception 94 (2016) 8186
Table 1 (continued)
Subject characteristic ENG Implant Continuation a (n= 673) ENG Implant Discontinuation b (n= 77) p value c
Temporizing measures, n (%)
None 629 (93.5) 61 (79.2) b .001
NSAIDs 4 (0.6) 2 (2.6) .119
Hormonal medication 40 (5.9) 14 (18.2) b .001
Other 1 (0.1) 0 (0.0) 1.000
a
Continued 12 months or longer.
b
Discontinued prior to 12 months.
c
p values for numeric data from independent-samples Student's t test and for categorical comparisons from Pearson 2 tests or Fisher's Exact Test as
appropriate.
d
Chlamydia trachomatis, Neisseria gonorrhea and Trichomonas vaginalis.
e
Dysmenorrhea, menorrhagia, oligomenorrhea, amenorrhea, polycystic ovary syndrome, ovarian cysts and endometriosis.
f
Specifically for ENG implant concerns not including device removal.

medications, were more frequently (p b .001) prescribed to as likely, patients with prior pregnancy were 2.8 times as
ENG discontinuers. likely and patients who sought medical care for device
The mean length of implant use for those who concerns were 10.7 times as likely to have the device
discontinued prior to 12 months was 7.5 months (Table 2). removed within 12 months.
The majority of implants were removed by Adolescent
Medicine providers, and problematic bleeding was the most
commonly cited reason for device removal. Seventy-four 4. Discussion
(96.1%) discontinuers picked less effective contraceptive
methods after implant removal. ENG implant discontinuation prior to 12 months was low
In multivariable logistic regression (Table 3), older age at among adolescents in an outpatient clinical setting. Prob-
insertion, history of pregnancy and 1 medical visit for lematic bleeding was the most common reason for removal.
implant concerns (not including removal) were indepen- Younger age at placement was not a risk factor for early
dently predictive of method discontinuation prior to discontinuation, whereas prior pregnancy was associated
12 months. Patients older than 16.8 years were 2.4 times with removal prior to 12 months. To our knowledge, ours
represents the largest study of adolescent ENG users in a
Table 2 pediatric office setting to date.
Descriptive summary of ENG implant discontinuers a Our finding of 10.3% discontinuation at 12 months is
Summary characteristic (n= 77) within the range of prior published rates (9%28%) [21,22].
Months of use, mean (SD) 7.5 (3.0) The finding that younger age is not a risk factor for
Implant removed by, n (%) discontinuation is consistent with 2 recent US studies and
Adolescent Medicine 57 (74.0) studies from the United Kingdom and Australia [14,21,2325].
Pediatric Primary Care 15 (19.5) A US study of insurance claims found that 12.4% of adolescent
Interventional Radiology 2 (2.6)
ENG implant users had their device removed within 1 year, and
Other 3 (3.9)
Reason for removal,b n (%) adolescent removal rates were lower than those among older
No reason stated 2 (2.6) women [14]. The Contraceptive CHOICE Project included 477
Bleeding problems 47 (61.0)
Headaches 5 (6.5)
Weight gain 10 (13.0) Table 3
Acne 3 (3.9) Logistic regression modeling predicting ENG implant discontinuation a
Emotional lability 7 (9.1)
Predictor Adjusted odds Adjusted
Abdominal pain 3 (3.9)
Ratio (95% CI) p value
Palpable foreign body intolerance 11 (14.3)
Other reason 14 (18.2) Constant 0.02 b.001
Contraceptive choice after removal, n (%) Age at insertion N16.8 years 2.4 (1.344.18) .003
None 21 (27.3) History of pregnancy 2.8 (1.475.29) .002
Progestin-only pill 1 (1.3) 1 Medical visit b 10.7 (6.1818.64) b.001
Combined hormonal contraception 31 (40.3) a
Backwards elimination stepwise logistic regression procedure
DMPA 21 (27.3)
employed. p b.1 required to qualify for modeling. Variables eligible but
Hormonal IUD 3 (3.9)
not retained: 1 telephone consultations, 1 contraceptive in past 2 years
a
Discontinued prior to 12 months. and temporizing measures.
b b
Multiple responses could be reported per subject. Specifically for ENG implant concerns not including removal.
 E. Berlan et al. / Contraception 94 (2016) 8186
participants aged 1419 who selected the ENG implant and
found no statistical difference in ENG removal rates between
adolescents and adults [24]. Generalizability of CHOICE study
findings may be limited by the fact that half of the teen
participants had previously been pregnant and contraception
was offered free of charge. Our current study among adolescents
in an outpatient pediatric setting, however, supports that
adolescents maintain ENG implants at high rates.
There are several plausible explanations why younger
adolescents were less likely to discontinue the ENG implant
within 12 months compared to older adolescents. Younger
users may experience less or be more tolerant of bleeding
irregularities. However, lower removal rates among younger
patients may indicate barriers to health care access (e.g.,
limited transportation, consent barriers and low health
literacy). Follow-up studies are needed to better understand
why younger adolescents have lower rates of removal.
The primary study limitation is our lack of data on ENG
implant removals performed outside our network. Accordingly,
our findings may underestimate actual ENG implant removal
rates. However, because we are the only children's hospital
network in our region, we do not suspect that many patients
sought removal outside of our system. The retrospective nature
of the current study confers additional limitations [26]. Although
a subsample of our charts was double-reviewed for accuracy, we
cannot exclude the possibility of misclassification translating
information from the EHR to the database. Moreover, research
staff members were not blinded to the study purpose, potentially
introducing bias of interpretation. Finally, some clinical
variables, including pregnancy status, history of STI and
contraceptive use, relied on patient report and provider
documentation and may have been underreported.
Our study extends the existing evidence that adolescents
are excellent candidates for the ENG implant. We found that
only a small proportion of adolescent patients discontinued
the method within 12 months. Given the current levels of
unintended teen pregnancy in the United States and the
contraceptive potential of LARC methods, efforts to increase
adolescent access to the ENG implant should be accelerated.
The outpatient pediatric setting is an ideal access point for
teens at risk for unintended pregnancy.
Acknowledgments
The authors would like to acknowledge Hannah L.H. Lange,
MPH, for her thoughtful input and edits of the manuscript.
References
[1] Kearney MS, Levine PB. Why is the teen birth rate in the United States
so high and why does it matter? J Econ Perspect 2012;26:14166.
85
[2] Sedgh G, Finer LB, Bankole A, Eilers MA, Singh S. Adolescent
pregnancy, birth, and abortion rates across countries: levels and recent
trends. J Adolesc Health 2015;56:22330, http://dx.doi.org/10.1016/
j.jadohealth.2014.09.007.
[3] Martin JA, Hamilton BE, Curtin SC, Mathews TJ. Births: final data for
2013. Natl Vital Stat Rep 2015;64.
[4] U.S. Department of Health and Human Services. Healthy People 2020;
2015.
[5] Finer LB, Zolna MR. Unintended pregnancy in the United States:
incidence and disparities, 2006. Contraception 2011;84:47885, http://
dx.doi.org/10.1016/j.contraception.2011.07.013.
[6] Martinez G, Copen CE, Abma JC. Teenagers in the United States:
sexual activity, contraceptive use, and childbearing, 20062010
national survey of family growth. Vital Health Stat 2011;23(31):15.
[7] Darroch JE, Singh S, Frost JJ. Differences in teenage pregnancy rates
among five developed countries: the roles of sexual activity and
contraceptive use. Fam Plan Perspect 2001;33:24450, 281.
[8] Kann L, Kinchen S, Shanklin SL, Flint KH, Kawkins J, Harris WA, et al.
Youth risk behavior surveillanceUnited States, 2013. MMWR Surveill
Summ 2014;63(Suppl 4):168 [doi: ss6304a1 pii].
[9] Trussell J. Contraceptive failure in the United States. Contraception
2011;83:397404, http://dx.doi.org/10.1016/j.contraception.2011.01.021.
[10] Swanson KJ, Gossett DR, Fournier M. Pediatricians' beliefs and prescribing
patterns of adolescent contraception: a provider survey. J Pediatr Adolesc
Gynecol 2013;26:3405, http://dx.doi.org/10.1016/j.jpag.2013.06.012.
[11] Daniels K, Daugherty J, Jones J. Current contraceptive status among women
aged 1544: United States, 20112013. NCHS Data Brief 2014;173:18.
[12] Winner B, Peipert JF, Zhao Q, Buckel C, Madden T, Allsworth JE, et al.
Effectiveness of long-acting reversible contraception. N Engl J Med
2012;366:19982007, http://dx.doi.org/10.1056/NEJMoa1110855.
[13] Secura GM, Madden T, McNicholas C, Mullersman J, Buckel CM,
Zhao Q, et al. Provision of no-cost, long-acting contraception and
teenage pregnancy. N Engl J Med 2014;371:131623, http://
dx.doi.org/10.1056/NEJMoa1400506.
[14] Berenson AB, Tan A, Hirth JM. Complications and continuation rates
associated with 2 types of long-acting contraception. Am J Obstet Gynecol
2015;212, http://dx.doi.org/10.1016/j.ajog.2014.12.028 [761.e1,761.e8].
[15] Frieden T. 2010 national center for health statistics keynote address,
2015Hyattsville, MD: Department of Health and Human Services; 2010.
[16] National Research Council. Initial national priorities for comparative
effectiveness research; 2009.
[17] Ott MA, Sucato GS. Committee on Adolescence. Contraception for
adolescents. Pediatrics 2014;134:e125781, http://dx.doi.org/10.1542/
peds.20142300.
[18] Potter J, Koyama A, Coles MS. Addressing the challenges of clinician
training for long-acting reversible contraception. JAMA Pediatr
2015;169:1034, http://dx.doi.org/10.1001/jamapediatrics.2014.2812.
[19] Zuckerman B, Nathan S, Mate K. Preventing unintended pregnancy: a
pediatric opportunity. Pediatrics 2014;133:1813, http://dx.doi.org/
10.1542/peds.20131147.
[20] Harris PA, Taylor R, Thielke R, Payne J, Gonzalez N, Conde JG.
Research electronic data capture (REDCap)a metadata-driven
methodology and workflow process for providing translational
research informatics support. J Biomed Inform 2009;42:37781,
http://dx.doi.org/10.1016/j.jbi.2008.08.010.
[21] Arribas-Mir L, Rueda-Lozano D, Agrela-Cardona M, Cedeno-Benavides
T, Olvera-Porcel C, Bueno-Cavanillas A. Insertion and 3-year follow-up
experience of 372 etonogestrel subdermal contraceptive implants by
family physicians in Granada, Spain. Contraception 2009;80:45762,
http://dx.doi.org/10.1016/j.contraception.2009.04.003.
[22] Teunissen AM, Grimm B, Roumen FJ. Continuation rates of the
subdermal contraceptive Implanon((R)) and associated influencing
factors. Eur J Contracept Reprod Health Care 2014;19:1521, http://
dx.doi.org/10.3109/13625187.2013.862231.
86 E. Berlan et al. / Contraception 94 (2016) 8186
[23] Harvey C, Seib C, Lucke J. Continuation rates and reasons for removal [25] Cea Soriano L, Wallander MA, Andersson S, Filonenko A, Garcia
among Implanon users accessing two family planning clinics in Rodriguez LA. The continuation rates of long-acting reversible
Queensland, Australia. Contraception 2009;80:52732, http:// contraceptives in UK general practice using data from The Health
dx.doi.org/10.1016/j.contraception.2009.05.132. Improvement Network. Pharmacoepidemiol Drug Saf 2015;24:528,
[24] Peipert JF, Zhao Q, Allsworth JE, Petrosky E, Madden T, Eisenberg D, http://dx.doi.org/10.1002/pds.3710.
et al. Continuation and satisfaction of reversible contraception. [26] Vassar M, Holzmann M. The retrospective chart review: important
Obstet Gynecol 2011;117:110513, http://dx.doi.org/10.1097/ methodological considerations. J Educ Eval Health Prof 2013;10:12,
AOG.0b013e31821188ad. http://dx.doi.org/10.3352/jeehp.2013.10.12.