Table 2 Summary of overall discontinuation rates and rates of discontinuation due to bleeding

LNG-IUD

Shimoni et al. [25]

Prospective comparative observational study

LNG-IUD, 13.5 mg (Skyla) early vs late menstrual cycle insertion

132

3

–

Removal, 7 (4%)

1 for spotting (< 1%)

14%

Teal et al. [14]

Single-arm phase III study, ACCESS IUS

LNG-IUD, 52 mg (Liletta)

1751^a

 > 7 years

–

Discontinued for an AE, 322 (18.8%)

39 (2.2%)

12%

Darney et al. [11]

Secondary analysis of phase III study, ACCESS IUS

LNG-IUD, 52 mg (Liletta)

1751^a

12

–

–

29 (1.7%)

–

Schreiber et al. [13]

Secondary analysis of phase III study, ACCESS IUS

LNG-IUD, 52 mg (Liletta)

1751^a

36

–

–

35 (2.1%); 20 during months

6–18

–

Eisenberg et al. [12]

Single-arm phase III study, ACCESS IUS

LNG20-IUD, 52 mg (Liletta)

1751^a

36

–

Other AEs leading to discontinuation: expulsion, 3.5%; acne, 1.3%; mood swings, 1.3%

1.5%

–

Neri et al. [26]

Prospective single-arm

LNG-IUD, 6 µg/day (Jaydess)

25

12

–

0

Vaitsiakhovich et al. [27]

Analysis of data from an observational study and RCT

LNG-IUD, 52 mg (Mirena)

1860

12, 24

12 months, 13.2%

24 months, 21.5%

–

NR

–

Korjamo et al. [20]

RCT

LNG-IUD (Mirena) immediate vs late insertion following MTOP

267

12

Immediate: 20 (15.0%)

Late: 43 (32.8%)

Immediate: 10 (7.5%)

Late: 15 (11.5%)

NR

–

Cristobal et al. [29]

Prospective, observational, single-arm

LNG-IUD, 52 mg

201

12

Any discontinuations, 5 (2.5%)

–

1 (< 1%) due to bleeding between periods

20%

Whitaker et al. [30]

RCT

LNG-IUD, immediate vs late insertion following caesarean delivery

42

6, 12

6 months

Immediate: 30.0%

Delayed: 40.9%

12 months

Immediate: 40.0%

Delayed: 59.1%

–

NR

–

Gemzell-Danielsson et al. [15]

Prospective single-arm

LNG-IUD

204

6, 12

–

Any discontinuations due to AEs, 5 (2.5%)

1 (0.5%)

20%

Armitage et al. [31]

Prospective, observational

LNG-IUD

100 (89 at follow-up)

12

14 (15.7%)

Removal, 10 (9%)

2 (2.2%)

14%

Nelson et al. [18]

RCT

LNG-IUD 13.5 mg vs 19.5 mg

1432 vs 1452

36

43% vs 40%

Discontinuation for AEs, 21.9% vs 19.1%

4.7% vs 4.9%

11% vs 12%

Gemzell-Danielsson et al. [17]

Post-hoc analysis of phase III RCT (Nelson et al. 2013)

LNG-IUD 13.5 mg vs 19.5 mg

1432 vs 1452

12, 36

1 year

Nulliparous: 21.2% vs 20.2%

Parous: 16.9% vs 14.9%

3 years,

Nulliparous: 45.7% vs 41.9%

Parous: 41.0% vs 38.9%

Discontinuation due to AEs,

3 year, nulliparous, 26.1% vs 20.6%

3 year, parous, 19.2% vs 18.2%

3-year discontinuation

Nulliparous: 5.2% vs 5.6%

Parous: 4.5% vs 4.4%

26% vs 27%

23% vs 24%

Apter et al. [54]

RCT

LNG-IUD (Jaydess, 13.5 mg) vs ENG implant

382 vs 381

12

74 (19.6%) vs 102 (26.8%)

–

16 (4.2%) vs 44 (11.5%)

22%

Short et al. [56]

Prospective, comparative, observational

LNG-IUD (Mirena) vs ENG

247 vs 116

24

32 (13%) vs 20 (17%)

–

9 (4%) vs 13 (11%)

28%

Weisberg et al. [57]

Prospective, comparative, observational

LNG-IUD (Mirena) vs ENG

179 vs 132

36

84 (47%) vs 36 (27%)

–

9 (23%) vs 27 (54%)

11%

Short et al. 2012[58]

Prospective, comparative, observational

LNG-IUD (Mirena) vs ENG

211 vs 100

12

12 (6%) vs 11 (11%)

–

6 (3%) vs 9 (9%)

50%

Cu-IUD

Yaron et al. [33]

Retrospective, observational

Cu-IUD, Ballerine MIDI

207

 ≥ 12

–

Any removal, 56 (27.1%)

Any removal excluding for pregnancy, 22.7%

33 (15.9%)

59%

Sanders et al. [34]

Prospective, longitudinal, observational

Cu-IUD, CuT380A

77 (72 at follow-up)

6

–

Any removals, 8 (11%)

NR

–

Bateson et al. [35]

Prospective, observational

Cu-IUD (TT380 short or long, or a Multiload device)

211

12, 36

Any discontinuation

1 year: 20.1%

3 years: 80, 38.7%

For AEs at 3 years, 59 (27.9%)

3-years, 28 (13.3%)

35%

Jagroep et al. [36]

Retrospective, observational

Cu-IUD, Cu-T380A or Cu­T375

1047

5 years

–

Any removal, 188 (18%)

23 (2.2%) due to complications such as pelvic pain or bleeding

12%

Scavuzzi et al. [37]

Cross-sectional, nulligravida vs parous women

Cu-IUD, CuT380A

157

NR

Any discontinuation

Nulligravida: 24.1%

Parous: 13.4%

–

Nulligravida: 6.0%

Parous: 1.4%

25% vs 12%

Wiebe and Trussell [38]

Prospective case series

Cu-IUD, SCu380A

51

12

–

Any removal, 9 (17.6%)

8 (16%) removed for symptoms

–

Garbers et al. [39]

Retrospective cohort analysis

Cu-IUD, CuT380A

283

6, 18

–

Any removal,

6 months, 31 (11%)

18 months, 78 (28%)

18 months, 24 (8.5%)

31%

Shimoni et al. [40]

RCT

Cu-IUD, immediate vs late insertion following MTOP

156

6

–

Any removal

Immediate, 10 (14%)

Delayed, 5 (8%)

Bleeding and pain cited as main reasons for removal

–

Reeves et al. [41]

RCT

Two Cu-IUDs: VeraCept175 vs CuT380S

198 vs 100

12, 24

Any discontinuation

12 months: 16% vs 32%

24 months: 31% vs 40%

–

For pain/bleeding

At 12 months: 3.5% vs 17.0%

At 24 months, 3.0% vs 15.1%

22% vs 53%

10% vs 38%

Akintomide et al. [42]

Retrospective, comparative, case control review

Two Cu-IUDs: Mini TT380 Slimline vs standard-sized TT380 Slimline

63 vs 67

12

10 (15%) vs 20 (32%)

–

For pain and bleeding, 3 (4.5%) vs 14 (22%)

30% vs 70%

LNG-IUD vs Cu-IUD

Bachofner et al. [43]

Retrospective comparative chart review

LNG-IUD, 52 mg vs Cu-T IUD (3rd generation) vs GyneFix 300 Cu-IUD

3rd generation Cu-IUDs (Multiload Cu375, Nova-T 380 and Mona Lisa Cu375)

419 vs 296

12, 36

–

Removal

12 months: 77 (18.4%) vs 61 (20.6%)

36 months, 116 (27.7%) vs 98 (33.1%)

12 months: 8 (1.9%) vs 9 (3.0%)

10% vs 15%

Phillips et al. [44]

Retrospective, comparative, observational

LNG-IUD vs Cu-IUD

770 vs 186

24, 36, 48, 60

Any discontinuations 24 months: 35.1% vs 42.3%

At any time: 554 (71.9%) vs 100 (53.8%)

–

At any time: 31 (4.0%) vs 18 (9.7%)

6% vs 18%

Hall and Kutler [45]

Prospective, comparative, survey

LNG-IUD (Mirena) vs CuT380A

88 vs 21

12

Any discontinuations, 4 (4.5%) vs 3 (14.3%)

–

0 (0%) vs 2 (9.5%)

0% vs 67%

Maguire et al. [46]

Secondary analysis of RCT assessing lidocaine for insertion pain

LNG-IUD vs CuT380A

62 vs 137

12

–

Removals: 6 (9.7%) vs 15 (10.9%)

–

–

Wildemeersch et al. [47]

Analysis of data collected from studies of FibroPlant and GyneFix

LNG-IUD (FibroPlant) vs Cu-IUD (GyneFix)

50 vs 104

12

Any discontinuation: 2 (4.3%) vs 4 (3.3%)

–

NR

–

Flamant et al. [48]

Prospective, comparative, observational

LNG-IUD vs Cu-IUD

43 vs 94

6

Any discontinuation: 15 (20%) vs 34 (22.1%)

–

1 (2.3%) vs 9 (9.6%)

12% vs 26%

McNicholas et al. [49]

Retrospective, comparative, observational

LNG-IUD vs Cu-IUD

53 vs 24

Median of 9 months

Any discontinuation: 20.8% vs 16.7%

–

NR

–

Lara-Torre et al. [50]

Retrospective, comparative, chart review

LNG-IUD vs Cu-IUD

77 vs 12

36

Removal, 25 (32.6%) vs 7 (58.3%)

For AEs, 17 (22.1%) vs 5 (41.7%)

LNG-IUD vs Cu-IUD vs implant

Agostini et al. [52]

Retrospective, comparative, cross-sectional

LNG-IUD vs Cu-IUD vs ENG implant

5405 vs 3896 vs 1482

12, 24

12 months: 5.0% vs 5.9% vs 10.6%

24 months: 8.9% vs 11.9% vs 16.4%

–

NR

–

Sanders et al. [53]

Prospective, comparative, observational

LNG-IUD (52 mg) vs Cu-IUD (T380) vs ENG implant

82 vs 33 vs 65

12

10% vs 12% vs 9%

–

NR

–

Grunloh et al. [22]

Prospective, comparative, cohort study, Contraceptive CHOICE Project

LNG-IUD vs Cu-IUD vs ENG

3610 vs 952 vs 1366

6

263 (7.3%) vs 76 (8.0%) vs 94 (6.9%)

–

Heavy bleeding: 3 (0.1%) vs 9 (0.9%) vs 0

Irregular/frequent bleeding: 14 (0.4%) vs 10 (1.1%) vs 50 (3.7%)

1% vs 11% vs 0%

O'Neil-Callahan et al. [23]

Prospective, comparative, cohort study, Contraceptive CHOICE Project

LNG-IUD vs Cu-IUD vs ENG

4423 (LARC)

12, 24

12 months: 12% vs 15% vs 17%

24 months: 21% vs 23% vs 31%

–

NR

–

Peipert et al. [24]

Prospective, comparative, cohort study, Contraceptive CHOICE Project

LNG-IUD vs Cu-IUD vs implant (vs non-LARC)

1890 vs 434 vs 522

12

12.5% vs 16.0% vs 16.7%

–

For bleeding or cramps, 5% vs 14% vs 10%

–

Modesto et al. [55]

RCT of routine vs intensive counselling

LNG-IUD vs CuT380A IUD vs ENG

99 vs 100 vs 98

12

19% vs 26.8% vs 17.4%

–

2.7% vs 4.0% vs 2.1%

14% vs 15% vs 12%