Systematic Review ajog.

org

1 56
2 Long-acting reversible contraception 57
3 58
4 in adolescents: a systematic review 59
5 60
6 and meta-analysis 61
7 Q16 Q1 Justin T. Diedrich, MD, MSCI; MAJ David A. Klein, MD, MPH; Jeffrey F. Peipert, MD, PhD 62
8 63
9 64
Introduction
10 65
Almost 1 in 5 female adolescents and OBJECTIVE: We sought to perform a systematic review of the medical literature to assess Q4
11 66
young women will give birth before age the continuation of long-acting reversible contraceptives among adolescents.
12 67
20 years.1 Of the approximately 574,000 STUDY DESIGN: Ovid-MEDLINE, Cochrane databases, and Embase databases were
13 68
adolescent pregnancies that occur each searched using key words relevant to the provision of long-acting contraception to
14 69
year in the United States, 75% are un- adolescents. Articles published from January 2002 through August 2016 were selected
15 70
intended.2 Although the United States for inclusion based on specific key word searches and detailed review of bibliographies.
16 71
has experienced a recent decline in teen For inclusion, articles must have provided data on method continuation, effectiveness, or
17 72
pregnancy,3 the rate remains higher than satisfaction of at least 1 long-acting reversible contraceptive method in participants <25
18 73
the rates in many other comparable years of age. Duration of follow-up had to be
6 months. Long-acting reversible con-
19 74
developed nations.4 Rates of unintended traceptive methods included intrauterine devices and the etonogestrel implant. Only
20 75
pregnancy in young women in poverty studies in the English language were included. Guidelines, systematic reviews, and
21 76
have increased while rates in more clinical reviews were examined for additional citations and relevant points for discussion.
22 77
afuent women have declined. Racial Of 1677 articles initially identified, 90 were selected for full review. Of these, 12 articles
23 78
and ethnic disparities also exist. The met criteria for inclusion. All studies selected for full review were extracted by multiple
24 79
pregnancy rates among black and Latina reviewers; inclusion was determined by consensus among authors. For studies with
25 80
teens are over twice that of white teens.1 similar outcomes, forest plots of combined effect estimates were created using the
26 81
Adolescents who become pregnant, and random effects model. The meta-analysis of observational studies in epidemiology Q5
27 82
especially those pregnant again within 1 guidelines were followed. Primary outcomes measured were continuation of method at
28 83
year of the previous pregnancy, are more 12 months, and expulsion rates for intrauterine devices.
29 84
RESULTS: This review included 12 studies, including 6 retrospective cohort studies, 5
30 85
prospective observational studies, and 1 randomized controlled trial. The 12 studies
31 86
Q2 From the Department of Obstetrics and included 4886 women age <25 years: 4131 intrauterine device users and 755 implant
32 87
Gynecology, University of California, Riverside, users. The 12-month continuation of any long-acting reversible contraceptive device was
33 Riverside, CA (Dr Diedrich); Departments of
88
84.0% (95% confidence interval, 79.0e89.0%). Intrauterine device continuation was
34 Family Medicine and Pediatrics, Uniformed 89
74.0% (95% confidence interval, 61.0e87.0%) and implant continuation was 84%
35 Services University of the Health Sciences, 90
Bethesda, MD (Dr Klein); and Department of (95% confidence interval, 77.0e91.0%). Among postpartum adolescents, the 12-month
36 91
Obstetrics and Gynecology, Indiana University long-acting reversible contraceptive continuation rate was 84.0% (95% confidence
37 92
School of Medicine, Indianapolis, IN (Dr Peipert). interval, 71.0e97.0%). The pooled intrauterine device expulsion rate was 8.0% (95%
38 93
Received Oct. 25, 2016; revised Nov. 24, 2016; confidence interval, 4.0e11.0%).
39 94
accepted Dec. 19, 2016. CONCLUSION: Adolescents have high continuation of long-acting reversible contra- Q6
40 95
Disclosures: Dr Diedrich is a Nexplanon trainer ceptive methods. Intrauterine devices and implants should be offered to all adolescents
41 for Merck, a contraceptive trainer for Upstream
96
as first-line contraceptive options.
42 USA, and serves on the board of directors of 97
43 URGE. Dr Peipert receives research support Key words: adolescents, birth control, contraception, implant, intrauterine device, 98
44 from Teva, Bayer Healthcare Pharmaceuticals, long-acting reversible contraception, meta-analysis, systematic review, teen 99
and Merck & Co Inc, and serves on advisory 100
45
Q3 boards for Teva Pharmaceuticals and Perrigo.
46 Dr Klein reports no conicts of interest. The
101
47 opinions and assertions contained herein are the
likely to subsequently experience serious Greater use of highly effective 102
48 private views of the authors and are not to be negative educational, economic, health, contraception can reduce unintended 103
49 construed as ofcial or as reecting the views of and social events than are adolescent pregnancy rates in this at-risk popula- 104
the US Air Force, the US Army, or the US Military females of the same age, race, and tion. Long-acting reversible contracep- 105
50
51
at large. ethnicity who did not become pregnant.5 tive (LARC) methods include 106
Corresponding author: Justin T. Diedrich, MD, Inconsistent use of contraceptives, use of intrauterine devices (IUDs) and the
52 107
MSCI. diedrich@ucr.edu
53 less-effective methods, and nonuse of etonogestrel (ENG) subdermal implant. 108
0002-9378/$36.00 contraceptives contribute to the high There are 2 general groups of IUDs
54 2016 Elsevier Inc. All rights reserved. 109
55 http://dx.doi.org/10.1016/j.ajog.2016.12.024 rate of unintended pregnancy among US commercially available in the United 110
adolescents.2 States: hormonal and nonhormonal.

MONTH 2017 American Journal of Obstetrics & Gynecology 1

REV 5.4.0 DTD
YMOB11457_proof
1 February 2017
8:10 pm
ce
 Systematic Review ajog.org

111 167
112 continuation of at least 1 LARC 168
FIGURE 1
113 method among participants with at 169
---
114 least 6 months of follow-up. While our 170
115 primary outcome was continuation at 171
116 12 months, we include studies that 172
117 have 6-month continuation as a sec- 173
118 ondary outcome when assessing 174
119 expulsion. Included studies must have 175
120 provided actual continuation of par- 176
121 ticipants, not estimated continuation. 177
122 Adolescent is not consistently dened 178
123 by specic ages in the medical litera- 179
124 ture, therefore we included women 180
125 24 years of age. When a study 181
126 included age groups extending >24 182
127 years of age, the published article must 183
128 have stratied the results by age group 184
129Q12 and must have included at least 1 185
Studies included in systematic review of adolescent use of long-acting reversible contraceptives. cohort of at least 20 participants
130 186
131
Diedrich. LARCs and adolescents. Am J Obstet Gynecol 2017.
exclusively 24 years of age. When 187
132 data were not reported for such a 188
133 cohort, the study was excluded. In 189
134 The primary mechanism of the implants among adolescents and young addition, studies with
30% loss to 190
135 levonorgestrel-containing IUD (LNG- women. The objective of this system- 12-month follow-up were excluded. 191
136 IUD) is the release of the progestin atic review is to provide an assessment Studies were also excluded if they 192
137 levonorgestrel, which thickens cervical of the ndings of the medical literature described LARC among special pop- 193
138 mucus, thereby preventing fertilization. of the use of LARC methods in young ulations of adolescents (eg, those with 194
139 The mechanism of the nonhormonal women age <25 years. Our hypothesis chronic disease such as HIV). Studies 195
140 copper-containing IUD (Cu-IUD) is was that continuation rates for examining postabortion and post- 196
141 the release of copper ions that inhibit adolescents using the IUD or implant partum adolescents were included. 197
142 sperm function, preventing fertiliza- are high (>75%) at 1 year from Two investigators (J.T.D. and D.A.K.) 198
143 tion. The reversible method of initiation. independently assessed titles and ab- 199
144 contraception most commonly used by stracts for inclusion. Articles that both 200
145 US women is the oral contraceptive Materials and Methods of them deemed to meet inclusion 201
146 pill.6 The failure rate of combined Search strategies and data sources criteria were included. In cases of 202
147 hormonal contraceptive methods (oral We included both randomized disagreement, the senior author (J.F.P.) 203
148 contraceptive pill, ring, or patch) is controlled trials (RCT) and observa- determined whether inclusion criteria 204
149 >20-fold higher than that of LARC tional trials in our review. Meta-analysis were met. 205
150 methods.7 The safety of LARC of observational studies in epidemiology 206
151 methods is well established and has led guidelines were followed.13 A literature Data extraction 207
152 to their endorsement as rst-line con- search was performed of the Ovid- Data were extracted by 2 investigators 208
153 traceptive methods by the American MEDLINE, Cochrane databases, and (J.T.D. and D.A.K.) for all included 209
154 Congress of Obstetricians and Gyne- Embase databases using key words rele- studies. Data extracted included the 210
155 cologists (ACOG)8 and the American vant to the provision of long-acting study methodology, number of partici- 211
156 Academy of Pediatrics (AAP).9 In the contraception to adolescents. Because pants, age range, type of LARC used, 212
157 Selected Practice Recommendations for the goal was to look at contemporary and insertion setting (postpartum, 213
158 Contraceptive Use, the Centers for LARC methods, the search was limited to postabortion, or interval). Interval 214
159 Disease Control and Prevention (CDC) articles published in 2002 or later. The insertion was dened as not during the 215
160 acknowledge that LARCs are the most search was limited to English-language initial postpartum period. The time of Q7
216
161 effective reversible methods and are articles. The full search terms and strat- follow-up, primary and secondary out- 217
162 appropriate for adolescents and egy are shown in online supplementary comes measured, and attrition. We 218
163 nulliparous women.10-12 Although material. noted the number (and ages) of ado- 219
164 multiple studies discuss continuation lescents included, and their specic 220
165 of LARC methods in the adolescent Inclusion and exclusion criteria subgroup outcomes. We also recorded 221
166 population, there is no consensus of To be included in our nal analysis, the number of reported IUD 222
continuation rates for IUDs and articles must have reported data on expulsions.

2 American Journal of Obstetrics & Gynecology MONTH 2017

REV 5.4.0 DTD
YMOB11457_proof
1 February 2017
8:10 pm
ce
 278
277
276
275
274
273
272
271
270
269
268
267
266
265
264
263
262
261
260
259
258
257
256
255
254
253
252
251
250
249
248
247
246
245
244
243
242
241
240
239
238
237
236
235
234
233
232
231
230
229
228
227
226
225
224
223

ajog.org
TABLE 1
Included studies evaluating use of long-acting reversible contraception among adolescent patients
Age Follow-up,
Author Year Study type range, y n LARC mo Outcome Insertion timing Country Nulliparous Attrition Q14

18
Godfrey et al 2010 RCT 14e18 23 Cu-IUD 6 6 mo Continuation Interval US 52% 2% Q15
LNG-IUD Expulsion
REV 5.4.0 DTD
YMOB11457_proof
1 February 2017
8:10 pm
ce

Guazzelli et al19 2010 Prospective <20 44 Implant 12 12 mo Continuation Postpartum Brazil 0 6%

cohort
Alton et al15 2012 Retrospective 11e21 233 Cu-IUD 96 12 mo Continuation Interval US 30% NR
cohort LNG-IUD
Rosenstock et al20 2012 Prospective 14e19 763 Cu-IUD 12 12 mo Continuation Interval, postpartum, US 77% 6%
cohort LNG-IUD postabortion
Implant
Teal and Sheeder21 2012 Retrospective 14e23 136 Cu-IUD 12 12 mo Continuation Postpartum US 0 14%
cohort LNG-IUD Expulsion
Tocce et al 2012 Prospective 13e23 171 Implant 12 12 mo Continuation Postpartum US 0 5%
cohort 6 mo Continuation
Garbers et al17 2013 Retrospective 14e19 73 Cu-IUD 6 6 mo Continuation NR US NR 15%
cohort
Aoun et al16 2014 Retrospective 13e24 999 Cu-IUD 36 12 mo Continuation NR US 16% 13%
cohort LNG-IUD Expulsion
MONTH 2017 American Journal of Obstetrics & Gynecology

Cohen et al22 2016 Prospective 13e22 244 Cu-IUD 12 12 mo Continuation Postpartum US 0 17%
cohort LNG-IUD 6 mo Continuation
Implant Expulsion
Teal et al23 2015 Retrospective 13e24 1146 Cu-IUD 6 6 mo Continuation Interval US 59% 30%
cohort LNG-IUD Expulsion
Berlan et al25 2016 Retrospective 12e22 750 Implant 12 12 mo Continuation Interval US 85% NR
cohort

Systematic Review
Gemzell-Danielsson et al26 2016 Prospective 12e17 304 LNG-IUD 12 12 mo Continuation Interval Multi 98% 1%
cohort AE
Total N: 4886
AE, adverse events; Cu, copper-containing; IUD, intrauterine device; LARC, long-acting reversible contraceptive; LNG, levonorgestrel-containing; NR, not reported; RCT, randomized controlled trial.
Diedrich. LARCs and adolescents. Am J Obstet Gynecol 2017.
3

334
333
332
331
330
329
328
327
326
325
324
323
322
321
320
319
318
317
316
315
314
313
312
311
310
309
308
307
306
305
304
303
302
301
300
299
298
297
296
295
294
293
292
291
290
289
288
287
286
285
284
283
282
281
280
279
 Systematic Review ajog.org

335 391
336 Assessment of risk of bias 392
Risk of bias was assessed using the

points
337 393

Total
checklist described by Downs and

17
15
22
14
18
20
16
23
18
18
24
7
338 394
339 Black.14 Studies received points for their 395
27
low risks of bias in several categories:

0
1
0
0
0
1
0
0
1
0
0
1
340 396
341 reporting, external validity, bias, and 397
26
0
0
0
1
0
0
0
0
1
0
0
0
342 confounding. There were a total of 27 398
points assigned in the following cate-
25

343 399
0
0
1
1
0
0
0
0
1

1
1
.
344 gories: reporting (10 points possible), 400
external validity (3 points possible), bias
24

345 401
0
0
0
0
0
0
1
0
0
0
0
0
346 (7 points possible), and confounding (7 402
23

points possible). Studies were grouped

0
0
0
0
0
0
1
0
0
0
0
0
Confounding

347 403
348 according to their score, with high scores 404
22

1
1
1

1
0
1
1
1
1
indicating lower risk bias: excellent (25-
.

.
.
349 405
27), good (19-24), fair (14-18), and poor
21

350 406
1
1
1

1
0
1
1
1
1
.

.
.

351 (<14). 407

20

0
1
1
0
1
1
1
1
1
1
1
352 408
.

353 Data synthesis 409

19

The proportion of women continuing

0
1
1
0
1
1
1
1
1

1
.

354 . 410
355 LARC methods were pooled for contin- 411
18

1
1
1

1
1
1
1
1
1
1
.

356 uation rates of 6 and 12 months using a 412

random effects model. Individual esti-
17

357 413
0
1
1
1
1
1
1
1
1
1
1
1

358 mates were weighted by their SE. The 414

16

same technique was used for pro-

0
0
0
1

1
1
1
0
1
0
1

359 415
.

360 portions of women with expulsion of 416

15
0
0
0
0
0
0
0
0
0
0
0
1

361 their IUDs. Heterogeneity of studies was 417

Bias

assessed by using I2 and further charac-

14

362 418
0
0
0
0
0
0
1
0
1
0
0
1
Quality scores assessing risk of bias using Downs and Black14 methodology

363 terized using Egger test of publication 419

bias.
13

364 420
1

1
1
1
1
1
1
1
1
1
.

365 421
12
External

Results
validity

1
1
1

1
1
1
1
.

.
.

366 422
367 Study selection 423
11

1
1
1

1
1
1
1
1

Using our search strategy, 1677 citations

.

368 424
were identied. From these titles and
10

369 425
1
1
1
1
1
1
1
1
1
1
1
1

370 abstracts, 90 articles appeared to meet 426

our inclusion criteria. Of these, 39 were
9
0
0
0
1
1
0
0
0
1
0
0
1

371 427
excluded because they did not provide
8
1
1
1
1
1
1
1
1
1
1
1
1

372 428
373 data on the primary endpoint; 24 were 429
7
0
1
0
1
1
1
1
1
1
0
1
1

374 excluded because the primary endpoint 430

was not stated for the adolescent sub-
6
1
1
1
1
1
1
1
1
1
1
1
1

375 431
376 group; 8 studies were separate analyses of 432
5
0
1
1
1
1
1
1
1
1
1
1
1

other included studies; 4 were excluded

Diedrich. LARCs and adolescents. Am J Obstet Gynecol 2017.

377 433
because follow-up was <70%; and an
4
1
1
1
1
1
1
1
1
1
1
1
1

378 434
additional 3 studies were excluded
3
1
1
1
1
1
1
1
1
1
1
1
1

379 435
Reporting

380 because they were cross-sectional 436

2
1
1
1
1
1
1
1
1
1
1
1
1

381 studies. After exclusions, 12 articles 437

1
1
1
1
1
1
1
1
1
1
1
1
1

382 that met all criteria and were included 438

for analysis.15-26 Figure 1 shows the F1 439
26

383
Teal and Sheeder, 2012
Gemzell-Danielsson et al

384 selection of included articles. 440

21

385 441
20

Teal et al, 2015

Study characteristics
Rosenstock et al
19

386 442
17

Godfrey et al18
Guazzelli et al

Characteristics of individual studies

Berlan et al25
22

Garbers et al

387 443
Alton et al15
Aoun et al16

23
Cohen et al

Tocce et al

are presented in Table 1. A total of T1 444

TABLE 2

388
Author

389 4886 adolescent and young adult 445

390 women (<25 years of age) were 446
included from all studies. Sample sizes

4 American Journal of Obstetrics & Gynecology MONTH 2017

REV 5.4.0 DTD
YMOB11457_proof
1 February 2017
8:10 pm
ce
 ajog.org Systematic Review
447 503
448 FIGURE 2 504
449 505
print & web 4C=FPO

---
450 506
451 507
452 508
453 509
454 510
455 511
456 512
457 513
458 514
459 515
460 516
461 517
462 518
463 519
464 520
465 521
466 522
467 523
468 524
469 525
470 526
471 527
472 528
473 529
474 530
475 531
476 532
477 533
478 534
479 535
480 536
481 Q13 Pooled 12-month continuation rates of long-acting reversible contraceptive (LARC) methods among adolescents. 537
482 CI, confidence interval. 538
483 Diedrich. LARCs and adolescents. Am J Obstet Gynecol 2017. 539
484 540
485 541
486 542
487 from the included studies ranged from patients. LARCs were placed post- Median follow-up was the same for 543
488 23-1146. Among the included studies, partum in 4 studies,19,21,22,24 and in- both prospective and retrospective 544
489 755 subjects used the subdermal terval placement in 3 studies.15,18,23 studies. Ten of the included studies 545
490 implant and 4131 used the IUD. There There were 2 studies16,17 that did not were performed in the United 546
491 were 8 studies that included the Cu- specify the timing of LARC placement. States,15-18,20,22,23,25,27,28 1 study was 547
492 IUD,15-18,20-23 9 studies that included One study20 allowed placement post- performed in Brazil,19 and 1 study was a 548
493 the LNG-IUD,15-18,20,22,23,26,27 and 4 partum, postabortion, or interval. multinational study.26 549
494 studies that included the ENG Of the 12 studies, 1 study was a RCT,18 550
495 implant.19,22,25,28 Many of the studies and the remaining 11 were obser- Randomized controlled trial 551
496 compared LARC methods. Compari- vational studies. Five studies were pro- Godfrey and colleagues18 performed a 552
497 sons between Cu-IUD and LNG-IUD spective cohort studies19,20,26-28 and 6 pilot RCT that randomized 23 adoles- 553
498 were performed in 8 studies15-18,20-23; studies were retrospective cohort cents and young women between age 554
499 2 studies included cohorts of both studies.15-17,21,23,25 Overall, approxi- 13-18 years to either Cu-IUD (n 11) 555
500 IUD and the ENG implant users.20,22 mately 34% of adolescents in the or LNG-IUD (n 12). Subjects had an 556
501 Three studies included only data for included studies were nulliparous. interval IUD placement or placement at 557
502 6 months of continuation,17,18,23 Follow-up ranged between 6-96 months, least 7 weeks postpartum. Continuation 558
which account for a total of 1242 with median follow-up of 12 months. at 6 months was 75% for LNG-IUD

MONTH 2017 American Journal of Obstetrics & Gynecology 5

REV 5.4.0 DTD
YMOB11457_proof
1 February 2017
8:10 pm
ce
 Systematic Review ajog.org

559 615
560 FIGURE 3 616
561 617
print & web 4C=FPO

---
562 618
563 619
564 620
565 621
566 622
567 623
568 624
569 625
570 626
571 627
572 628
573 629
574 630
575 631
576 632
577 633
578 634
579 635
580 636
581 637
582 638
583 639
584 640
585 641
586 642
587 643
588 644
589 645
590 646
591 647
Pooled 12-month continuation rates of intrauterine devices (IUD) among adolescents.
592 648
CI, confidence interval.
593 649
Diedrich. LARCs and adolescents. Am J Obstet Gynecol 2017.
594 650
595 651
596 652
597 653
598 and 45% for Cu-IUD (P .15). Despite their method of choice. Of the 763 ad- Paolo, Brazil. The cohort had an average 654
599 high discontinuation, the majority of olescents and young women (14-19 age of 17 years; 91% had 1 child and the 655
600 subjects reported being satised with years) who started a LARC method at remainder had
2. All women included 656
601 their IUD at 6 months (70% of LNG- baseline, continuation at 12 months was had a subdermal implant placed and 657
602 IUD and 80% of Cu-IUD users). Two 81% among LNG-IUD users, 76% were followed prospectively for 1 year; 658
603 Cu-IUD expulsions, but no LNG-IUD among Cu-IUD users, and 82% among 6% were lost to follow-up. Continuation 659
604 expulsions, were reported. implant users. By 12 months, <6% of was 94% at 12 months, and the rate of 660
605 adolescent participants had been lost to amenorrhea was 38% by 12 months. 661
606 Observational studies follow-up. Expulsion of IUDs was not Another prospective study was per- 662
607 In 2012, Rosenstock and colleagues20 reported in the article. However, another formed by Cohen and colleagues.22 Ad- 663
608 published a subanalysis of the adoles- article from the same study population olescents and young women (ages 13-22 664
609 cents participating in the Contraceptive estimated the risk of expulsion at 10.5 years) who chose postplacental IUDs 665
610 CHOICE Project. CHOICE was a pro- per 100 IUD users per 12 months (95% (n 82) or subdermal implant (n 666
611 spective observational study of women condence interval [CI], 8.0e13.5) 162) to be placed prior to discharge were 667
612 in the St Louis, MO, area who were among women <20 years of age.29 included. At 12 months, IUD continua- 668
613 provided with no-cost contraception for Guazzelli and colleagues19 included 44 tion was 62% and implant continuation 669
614 2-3 years. All participants received tier- adolescents who presented <6 months was 72%. The observed IUD expulsion 670
based contraceptive counseling and postpartum for LARC at a clinic in So rate reported was 21%.

6 American Journal of Obstetrics & Gynecology MONTH 2017

REV 5.4.0 DTD
YMOB11457_proof
1 February 2017
8:10 pm
ce
 ajog.org Systematic Review
671 727
672 FIGURE 4 728
673 729
print & web 4C=FPO

---
674 730
675 731
676 732
677 733
678 734
679 735
680 736
681 737
682 738
683 739
684 740
685 741
686 742
687 743
688 744
689 745
690 746
691 747
692 748
693 749
694 750
695 751
696 752
697 753
698 754
699 755
700 756
701 757
702 758
703 759
704 Pooled 12-month continuation rates of etonogestrel (ENG) subdermal implants among adolescents. 760
705 CI, confidence interval. 761
Diedrich. LARCs and adolescents. Am J Obstet Gynecol 2017.
706 762
707 763
708 764
709 Tocce and colleagues24 performed a Gemzell-Danielsson and colleagues26 population, 70% were parous and the 765
710 prospective cohort study of 171 post- evaluated the use of a new IUD among median age at insertion was 16 years. At 766
711 partum adolescents and young women girls and adolescents (12-17 years of 12 months, continuation was 70% 767
712 (ages 13-24 years) who had subdermal age). The IUD evaluated was a LNG-IUD among the youngest group of adoles- 768
713 implant placed prior to discharge. This containing 13.5 mg of levonorgestrel cents (age <18 years) and 89% among 769
714 group was compared to a control group released at a rate of 8 mg/d. There were those age 18-21 years. The number of 770
715 of adolescents who chose any other 304 adolescents who had the LNG-IUD IUD expulsions was not reported. 771
716 method. The primary outcomes were inserted, and all were followed up for Teal and Sheeder21 performed a 772
717 contraceptive continuation and repeat 12 months. Continuation at 12 months retrospective cohort study of parous 773
718 pregnancy rates. Continuation of the was 83%. There were 10 expulsions (3%) adolescents and young women (14-23 774
719 implant was 97% at 6 months and 86% during 12 months. years of age) who had each received a 775
720 at 12 months. The odds of pregnancy In a retrospective cohort study, Alton LNG-IUD or Cu-IUD. The average 776
721 were 8 times higher for those who did and colleagues15 identied 233 adoles- insertion time was 8 months postpartum 777
722 not choose immediate postpartum cents age <21 years who had each (none were placed immediately after 778
723 implant (odds ratio, 8.0; 95% CI, received a Cu-IUD (n 11) or LNG- placental delivery). Median continuation 779
724 2.8e23.0) compared to women who did IUD (n 222) during an 8-year of IUD use was 14 months; range was not 780
725 choose insertion. period. The IUDs had been placed at a reported. There was no difference in 781
726 In a subgroup analysis of a large private faculty clinic or at a hospital- continuation based on type of IUD. 782
multinational prospective phase III trial, based Title X clinic. Of their study Continuation rates were censored at 60

MONTH 2017 American Journal of Obstetrics & Gynecology 7

REV 5.4.0 DTD
YMOB11457_proof
1 February 2017
8:10 pm
ce
 Systematic Review ajog.org

783 839
784 FIGURE 5 840
785 841
print & web 4C=FPO

---
786 842
787 843
788 844
789 845
790 846
791 847
792 848
793 849
794 850
795 851
796 852
797 853
798 854
799 855
800 856
801 857
802 858
803 859
804 860
805 861
806 862
807 863
808 864
809 865
810 866
811 867
812 868
813 869
814 870
815 Pooled postpartum 12-month continuation rates of long-acting reversible contraceptive (LARC) devices. 871
816 CI, confidence interval. 872
817 Diedrich. LARCs and adolescents. Am J Obstet Gynecol 2017. 873
818 874
819 875
820 876
821 months. Twelve-month continuation lost to follow-up. An expulsion rate of placed, only 10% had discontinued by 12 877
822 was 55%, and an expulsion rate of 15% 4% was observed. months (90% continuation). 878
823 was observed. Teal and colleagues23 performed a 879
824 Garbers and colleagues17 retro- retrospective study of adolescents and Assessment of risk of bias 880
825 spectively reviewed charts of 73 ado- young women (13-24 years) who desired The majority of the included studies 881
826 lescents and young women (ages an IUD. The goal of this study was to were of fair or good quality under their Q8 882
827 14-19 years) who had sought family quantify complications and unsuccessful criteria. Overall, the average score for 883
828 planning services and had Cu-IUDs insertions among 1177 who had an reporting results was 9 of 10 points; 884
829 placed. According to chart review, attempted IUD placement. Among the average scores for external validity were 885
830 6-month continuation of the Cu-IUD 1146 who had a successful insertion, 2.25 of 3 points; average scores for bias 886
831 was 88%. IUD expulsions were not continuation of the IUD was 95% at 6 were 4 of 7 points; and the average score 887
832 reported. months. A 2% IUD expulsion rate was for confounding was 2.4 of 7 points. See 888
833 Aoun and colleagues16 reviewed observed. Table 2 for results. The majority of T2 889
834 charts of 999 adolescents and young A retrospective study was performed studies had low scores for confounding 890
835 women (age 14-24 years) who received a by Berlan and colleagues25 evaluating 12- and bias, which mainly is due to study 891
836 Cu-IUD or LNG-IUD. At the time of month continuation of the subdermal design. Because only 1 study included 892
837 insertion, approximately 16% of partic- implant by adolescents 12-22 years of was a RCT, there is a higher risk of bias 893
838 ipants were nulliparous. At 12 months, age. The majority (85%) were nullipa- among the remaining studies. However, 894
continuation was 80%. Only 13% were rous. Of 750 patients who had the device 1 advantage of a meta-analysis of

8 American Journal of Obstetrics & Gynecology MONTH 2017

REV 5.4.0 DTD
YMOB11457_proof
1 February 2017
8:10 pm
ce
 ajog.org Systematic Review
895 951
896 FIGURE 6 952
897 953
print & web 4C=FPO

---
898 954
899 955
900 956
901 957
902 958
903 959
904 960
905 961
906 962
907 963
908 964
909 965
910 966
911 967
912 968
913 969
914 970
915 971
916 972
917 973
918 974
919 975
920 976
921 977
922 978
923 979
924 980
925 981
926 982
927 983
Pooled intrauterine device (IUD) expulsion rates among adolescents.
928 984
CI, confidence interval.
929 985
Diedrich. LARCs and adolescents. Am J Obstet Gynecol 2017.
930 986
931 987
932 988
933 observational studies is generalizability continuation when LARC devices were adolescents. These ndings support the 989
934 and obtaining estimates that are closer to placed in the postpartum setting. At recommendations of the ACOG30 as well 990
935 real-life continuation. In practice, 12 months, continuation was 84.0% as the AAP,9 which encourages adoles- 991
936 women are able to choose their contra- (95% CI, 71.0e97.0%). IUD expulsion cents to consider LARC methods. The 992
ceptive method and are not randomly rates were pooled in Figure 6, and the CDC also recommends using the F6
937 993
938 assigned one. overall effect was an expulsion rate of methods most effective and medically 994
939 8.0% (95% CI, 4.0e11.0%). There is appropriate for adolescents, including 995
940 Data synthesis signicant heterogeneity among studies LARC methods.31 996
941 The 12-month continuation rates of all in all the comparisons above (I2 > 92%, 997
942F2 LARC methods are provided in Figure 2. P < .001). Strengths and limitations 998
943 Continuation was 84.0% (95% CI, A large strength of this review is its 999
944 79.0e89.0%) for all LARC methods Comment estimates of continuation among the 1000
945 combined. The 12-month continuation Main ndings different studies reviewed. Additionally, 1001
946F3 rate for IUDs is shown in Figure 3. At 12 This systematic review and meta- the pooled continuation rate includes a 1002
947 months, continuation was 74.0% (95% analysis demonstrates that adolescent large sample. By including observational 1003
948 CI, 61.0e87.0%). The 12-month continuation of LARC methods is high at trials in addition to RCT, we are able to 1004
949 continuation of ENG implant 84% 12 months. This includes parous ado- estimate real-world continuation of 1005
950 (95% CI, 77.0e91.0%) is shown in lescents who had devices placed post- LARCs more accurately (increased 1006
F4 Figure 4. Figure 5 shows 12-month partum as well as nulliparous generalizability). In actual practice,
F5
MONTH 2017 American Journal of Obstetrics & Gynecology 9
REV 5.4.0 DTD
YMOB11457_proof
1 February 2017
8:10 pm
ce
 Systematic Review ajog.org

1007 1063
patients are given a choice of contra- women in TRICARE. The study in- 2. Finer LB, Zolna MR. Declines in unintended
1008 pregnancy in the United States, 2008-2011. 1064
1009 ceptive method and must choose what cludes women of all ages, but in a 1065
N Engl J Med 2016;374:843-52.
1010 they think best ts their priorities and multivariable analysis found that girls 3. Hamilton BE, Mathews TJ. Continued de- 1066
1011 lifestyles. Meta-analyses that include and women 14-19 years of age had clines in teen births in the United States, 2015. 1067
1012 only RCT may actually introduce selec- higher continuation of LARC methods NCHS Data Brief 2016;(259):1-8.
1068
tion bias by eliminating patient choice of when compared to women 35-40 years 4. United Nations. Demographic yearbook 2013.
1013 Sixty-four. New York; 2014. Available at: http:// 1069
1014 method. of age: 1.34 (95% CI, 1.27e1.42). Data 1070
unstats.un.org/unsd/demographic/products/dyb/
1015 A weakness of this review is the sig- were not listed for continuation by age dyb2009-2010.htm. Accessed March 2, 2015. 1071
1016 nicant heterogeneity among the indi- group at 12 months.37 Another large 5. Jeha D, Usta I, Ghulmiyyah L, Nassar A. 1072
1017 vidual studies examined, which limits retrospective study evaluated insurance A review of the risks and consequences of
1073
1018 the ability to combine continuation claims data of approximately 90,000 adolescent pregnancy. J Neonatal Perinatal
1074
Med 2015. Q9
1019 outcomes. However, these studies are women who had an IUD placed from 1075
6. Daniels K, Daugherty J, Jones J. Current
1020 representative of the variety of settings of 2002 through 2009. Within this cohort contraceptive status among women aged 15- 1076
1021 LARC provision. When assessing expul- there were 1528 girls and women age 44: United States, 2011-2013. NCHS Data Brief 1077
1022 sions, there may be signicant differ- 15-19 years who used the LNG-IUD and 2014;(173):):1-8.
1078
ences between Cu-IUD and LNG-IUD. 307 in the same age group who used the 7. Winner B, Peipert JF, Zhao Q, et al. Effec-
1023 1079
1024 Additionally, 1 study26 assessed a lower- Cu-IUD. Among those 20-24 years of tiveness of long-acting reversible contraception.
1080
N Engl J Med 2012;366:1998-2007.
1025 dose LNG-IUD (13.5 mg reservoir, age, 7860 and 2027 used the LNG-IUD 8. Committee on Adolescent Health Care Long- 1081
1026 releasing 8 mg/d), which should be and Cu-IUD, respectively. In the Acting Reversible Contraception Working 1082
1027 distinguished from the other LNG-IUDs younger cohort the continuation rates Group, the American College of Obstetricians
1083
1028 on the market in the United States. for LNG-IUD and Cu-IUD were 88.2% and Gynecologists. Adolescents and long-
1084
Liletta (Medicines 360) and Mirena and 79.8% at 12 months, respectively. In acting reversible contraception: implants and
1029 intrauterine devices. Committee opinion no. 1085
1030 (Bayer) contain 52 mg of levonorgestrel the cohort aged 20-24 years, continua- 539. Obstet Gynecol 2012;120:983-8. 1086
1031 and release 20 mg/d, Kyleena (Bayer) tion was 87.7% and 84.1% for the LNG- 9. Committee on Adolescence. Contraception 1087
1032 contains 19.5 mg and releases 9 mg/d, IUD and Cu-IUDs at 12 months, for adolescents. Pediatrics 2014;134:
1088
1033 and Skyla (Bayer) contains 13 mg and respectively.38 The main limitation of e1244-56.
1089
1034 releases 8 mg/d. Another limitation may this type of study is the potential inac- 10. Curtis KM, Tepper NK, Jatlaoui TC, et al. US
1090
medical eligibility criteria for contraceptive use,
1035 be the inclusion of women up to 25 years curacy of claims data. 2016. MMWR Recomm Rep 2016;65:1-103. 1091
1036 of age. This was done intentionally as 11. Curtis KM, Jatlaoui TC, Tepper NK, et al. US 1092
1037 there is not an agreed-upon denition of Conclusion with implications selected practice recommendations for contra-
1093
1038 the exact years of adolescence. Addi- LARC continuation in adolescents and ceptive use, 2016. MMWR Recomm Rep
1094
tionally, not all studies presented data on young women is high. IUDs and im- 2016;65:1-66.
1039 12. Gavin L, Moskosky S, Carter M, et al. 1095
1040 barriers to removal. When LARC plants should be offered to all adoles- Providing quality family planning services: rec- 1096
1041 removal is not free or easily attainable for cents as rst-line contraceptive options ommendations of CDC and the US Ofce of 1097
1042 participants, this may not only falsely as long as device removal is readily Population Affairs. MMWR Recomm Rep
1098
1043 increase continuation, but in fact be available. Efforts to increase the use of 2014;63:1-54.
1099
construed as coercive. LARC will help decrease the rates of 13. Stroup DF, Berlin JA, Morton SC, et al.
1044 Meta-analysis of observational studies in 1100
1045 unintended adolescent pregnancy, epidemiology: a proposal for reporting. Meta- 1101
1046 Comparison with existing literature abortion, and unintended childbearing, analysis of observational studies in epidemi- 1102
1047 A Cochrane review by Krashin et al32 and will thereby also help lower the ology (MOOSE) group. JAMA 2000;283:
1103
1048 and systematic review by Deans and disparities in those rates among 2008-12.
1104
Grimes33 and Usinger et al34 showed different sociodemographic adolescent 14. Downs SH, Black N. The feasibility of
1049 creating a checklist for the assessment of the 1105
adolescent continuation of 75-86% with populations. -
1050 methodological quality both of randomized and 1106
1051 IUDs. However, these reviews included non-randomized studies of health care in- 1107
1052 only RCT, which are different from ACKNOWLEDGMENT terventions. J Epidemiol Community Health
1108
1053 most prospective observational trials in The authors would like to acknowledge the work
1998;52:377-84.
1109
that patients in observational trials 15. Alton TM, Brock GN, Yang D, Wilking DA,
1054 of Ms Wendy Larson, medical librarian at Fort 1110
Hertweck SP, Loveless MB. Retrospective re-
1055 ordinarily are allowed choice of con- Belvoir Community Hospital, Fort Belvoir, VA,
view of intrauterine device in adolescent and 1111
traceptive method. Observational who made this systematic review possible.
1056 young women. J Pediatr Adolesc Gynecol 1112
1057 studies may provide greater generaliz- 2012;25:195-200.
1113
ability and a more realistic estimate of 16. Aoun J, Dines VA, Stovall DW, Mete M,
1058 REFERENCES Nelson CB, Gomez-Lobo V. Effects of age, 1114
1059 continuation rates. RCT can lead to an 1115
increased discontinuation because the 1. Martinez G, Copen CE, Abma JC. Teenagers parity, and device type on complications and
1060 in the United States: sexual activity, contracep- discontinuation of intrauterine devices. Obstet 1116
1061 methods are assigned randomly.35,36 tive use, and childbearing, 2006-2010 National Gynecol 2014;123:585-92. 1117
1062 This study is also consistent with a Survey of Family Growth. Vital Health Stat 23 17. Garbers S, Haines-Stephan J, Lipton Y,
1118
large retrospective study including 2011;23:1-35. Meserve A, Spieler L, Chiasson MA. Continuation

10 American Journal of Obstetrics & Gynecology MONTH 2017

REV 5.4.0 DTD
YMOB11457_proof
1 February 2017
8:10 pm
ce
 ajog.org Systematic Review
1119 1175
of copper-containing intrauterine devices at 6 difference? Am J Obstet Gynecol 2012;206: Editorial and Production Staff (Serials) MMWR
1120 months. Contraception 2013;87:101-6. 481.e1-7. Editorial Board. Recomm Reports 2014;63(4). Q11 1176
1121 18. Godfrey EM, Memmel LM, Neustadt A, et al. 25. Berlan E, Mizraji K, Bonny AE. Twelve- 32. Krashin J, Tang JH, Mody S, Lopez LM. 1177
1122 Intrauterine contraception for adolescents aged month discontinuation of etonogestrel implant in Hormonal and intrauterine methods for contra- 1178
1123 14-18 years: a multicenter randomized pilot an outpatient pediatric setting. Contraception ception for women aged 25 years and younger. 1179
study of levonorgestrel-releasing intrauterine 2016;94:81-6. Cochrane Database Syst Rev 2015;8:
1124 1180
system compared to the copper T 380A. 26. Gemzell-Danielsson K, Buhling KJ, CD009805.
1125 Contraception 2010;81:123-7. Dermout SM, Lukkari-Lax E, Montegriffo E, 33. Deans EI, Grimes DA. Intrauterine devices 1181
1126 19. Guazzelli CAF, de Queiroz FT, Barbieri M, Apter D. A phase III, single-arm study of LNG- for adolescents: a systematic review. Contra- 1182
1127 Torloni MR, de Araujo FF. Etonogestrel implant IUS 8, a low-dose levonorgestrel intrauterine ception 2009;79:418-23. 1183
1128 in postpartum adolescents: bleeding pattern, contraceptive system (total content 13.5 mg) in 34. Usinger KM, Gola SB, Salas M, 1184
efcacy and discontinuation rate. Contraception postmenarcheal adolescents. Contraception Smaldone A. Intrauterine contraception contin-
1129 1185
2010;82:256-9. 2016;93:507-12. uation in adolescents and young women: a
1130 20. Rosenstock JR, Peipert JF, Madden T, 27. Teal SB, Sheeder J. IUD use in adolescent systematic review. J Pediatr Adolesc Gynecol 1186
1131 Zhao Q, Secura GM. Continuation of reversible mothers: retention, failure and reasons for 2016;29:659-67. 1187
1132 contraception in teenagers and young women. discontinuation. Contraception 2012;85:270-4. 35. Sivin I, el Mahgoub S, McCarthy T, et al. 1188
1133 Obstet Gynecol 2012;120:1298-305. 28. Tocce K, Sheeder J, Python J, Teal SB. Long Long-term contraception with the levonorges- 1189
21. Teal SB, Sheeder J. IUD use in adolescent acting reversible contraception in postpartum ad- trel 20 mcg/day (LNg 20) and the copper T
1134 1190
mothers: retention, failure and reasons for olescents: early initiation of etonogestrel implant is 380Ag intrauterine devices: a ve-year ran-
1135 discontinuation. Contraception 2012;85:270-4. superior to IUDs in the outpatient setting. J Pediatr domized study. Contraception 1990;42: 1191
1136 22. Cohen R, Sheeder J, Arango N, Teal SB, Adolesc Gynecol 2012;25:59-63. 361-78. 1192
1137 Tocce K. Twelve-month contraceptive continu- 29. Madden T, McNicholas C, Zhao Q, 36. Goldman JA, Dekel A, Reichman J. Imme- 1193
1138 ation and repeat pregnancy among young Secura GM, Eisenberg DL, Peipert JF. Associ- diate postabortion intrauterine contraception in 1194
mothers choosing post-delivery contraceptive ation of age and parity with intrauterine device nulliparous adolescents. Isr J Med Sci 1979;15:
1139 1195
implants or postplacental intrauterine devices. expulsion. Obstet Gynecol 2014;124:718-26. 522-5.
1140 Contraception 2016;93:178-83. 30. Committee on Adolescent Health Care 37. Chiles DP, Roberts TA, Klein DA. Initiation 1196
1141 23. Teal SB, Romer SE, Goldthwaite LM, Long-Acting Reversible Contraception Working and continuation of long-acting reversible 1197
1142 Peters MG, Kaplan DW, Sheeder J. Insertion Group, the American College of Obstetricians contraception in the United States military 1198
1143 characteristics of intrauterine devices in adoles- and Gynecologists. Adolescents and long- healthcare system. Am J Obstet Gynecol 1199
cents and young women: success, ancillary acting reversible contraception: implants and 2016;215:328.e1-9.
1144 1200
measures, and complications. Am J Obstet intrauterine devices. Committee opinion no. 38. Berenson AB, Tan A, Hirth JM,
1145 Gynecol 2015;213:515.e1-5. 539. Obstet Gynecol 2012;120:983-8. Wilkinson GS. Complications and continuation Q10 1201
1146 24. Tocce KM, Sheeder JL, Teal SB. Rapid 31. Frieden TR, Jaffe HW, Richards CL, et al. of intrauterine device use among commercially 1202
1147 repeat pregnancy in adolescents: do immediate Providing quality family planning services. Cen- insured teenagers. Obstet Gynecol 2013;121: 1203
1148 postpartum contraceptive implants make a ters for Disease Control and Prevention; MMWR 951-8. 1204
1149 1205
1150 1206
1151 1207
1152 1208
1153 1209
1154 1210
1155 1211
1156 1212
1157 1213
1158 1214
1159 1215
1160 1216
1161 1217
1162 1218
1163 1219
1164 1220
1165 1221
1166 1222
1167 1223
1168 1224
1169 1225
1170 1226
1171 1227
1172 1228
1173 1229
1174 1230

MONTH 2017 American Journal of Obstetrics & Gynecology 11

REV 5.4.0 DTD
YMOB11457_proof
1 February 2017
8:10 pm
ce
 Systematic Review ajog.org

1231 1287
1232 Supplemental Information 2 etonogestrel implant: (tw) 3 (continuation or discontinuation or 1288
1233 Ovid MEDLINE 3 levonorgestrel intrauterine system: side effects or compliance or sat- 1289
1234 1 exp contraception/or exp contra- (tw) isfaction).mp. [mptitle, original ti- 1290
1235 ceptive devices/or exp contraceptive 4 LARC.mp. [mptitle, short title, tle, abstract, mesh headings, heading 1291
1236 agents/ abstract, full text, keywords, caption words, keyword] 1292
1237 2 1 and LARC (tw) text] 4 1 and 2 and 3 1293
1238 3 long acting reversible 5 1 or 2 or 3 or 4 5 limit 4 to yr2002 -Current 1294
1239 contraception.tw. 6 from 5 keep 2, 9e11, 14e15, 20, 1295
1240 4 exp intrauterine devices/ 24e25 Embase - 1296
1241 5 exp drug implants/ and 1 7 contracept: (tw) (((contraception OR (contraceptive 1297
1242 6 etonogestrel implant: (tw) 8 (teen: or adolesce:) (tw) device) OR (contraceptive agent)) 1298
1243 7 levonorgestrel intrauterine system: 9 7 and 8 AND LARC) 1299
1244 (tw) 10 from 9 keep 7, 13, 16, 23, 39, 60 OR 1300
1245 8 2 or 3 or 4 or 5 or 6 or 7 11 6 or 10 1301
(long acting reversible contraception)
1246 9 limit 8 to English language OR (intrauterine device) 1302
1247 10 limit 9 to yr2002 -Current Reviews - Cochrane Central Register of 1303
11 limit 10 to adolescent (13e18 years) Controlled Trials OR
1248 1304
1249 12 limit 10 to young adult (19e24 1 (contraception or contraceptive).mp. (((contraception OR (contraceptive 1305
1250 years) [mptitle, original title, abstract, device) OR (contraceptive agent)) 1306
1251 13 11 or 12 mesh headings, heading words, AND (drug implant)) 1307
1252 keyword] OR 1308
1253 Reviews - Cochrane Database of 2 (adolescent or young or nullipar- 1309
(etonogestrel implant)
1254 Systematic Reviews ous).mp. [mptitle, original title, 1310
1 long acting reversible abstract, mesh headings, heading OR
1255 1311
1256 contraception.tw. words, keyword] (levonorgestrel intrauterine system) 1312
1257 1313
1258 1314
1259 1315
1260 1316
1261 1317
1262 1318
1263 1319
1264 1320
1265 1321
1266 1322
1267 1323
1268 1324
1269 1325
1270 1326
1271 1327
1272 1328
1273 1329
1274 1330
1275 1331
1276 1332
1277 1333
1278 1334
1279 1335
1280 1336
1281 1337
1282 1338
1283 1339
1284 1340
1285 1341
1286 1342

11.e1 American Journal of Obstetrics & Gynecology MONTH 2017

REV 5.4.0 DTD
YMOB11457_proof
1 February 2017
8:10 pm
ce