Methods

A randomised controlled trial (RCT) conducted in large hospital in North Carolina, USA
Study duration: August 2012 to April 2014
Sample size: 96 participants were recruited to detect 30% difference in contraceptive implant use at 1 year postpartum among peripartum women assigned to receive immediate postpartum insertion (prior to hospital discharge) and delayed postpartum insertion (4 to 6 weeks) of etonogestrel contraceptive implant with a study power of 80% and alpha of 0.05

Participants

Adolescents and young women aged 14 to 24 years, gave birth to a healthy infant, spoke English or Spanish, and desired to use the contraceptive implant.
Exclusion criteria included one or more of the following.

• Current or past history of thrombosis or thromboembolic disorders.

• Hepatic tumours.

• Active liver disease.

• Undiagnosed abnormal genital bleeding.

• Known, suspected or history of carcinoma of the breast.

• Hypersensitivity to the implant.

• Use of hepatic enzyme inducers.

• Maternal hospital intensive care unit (ICU) admission after delivery.

• Postpartum haemorrhage requiring blood transfusion.

• Hospital stay > 7 days postpartum.

• Coagulopathy.

• Hemolysis, elevated liver enzymes, and low platelet levels (HELLP) syndrome.

Most study participants were Hispanic/Latino (40% of women assigned to both immediate and delayed postpartum insertion groups).

Interventions

Intervention arm: participants received etonogestrel contraceptive implant within 48 hours of delivery.
Control arm: participants received the same contraceptive method but had been administered at 4 to 6 weeks after delivery.

Outcomes

Primary outcome: contraceptive implant use at 12 months postpartum.
Other outcome measures: satisfaction, bleeding patterns, breastfeeding

Notes

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

Participants were randomly assigned using computer‐generated random numbers in blocks of 4 and 6.

Allocation concealment (selection bias)

Low risk

The assignments were enclosed in sequentially numbered, opaque, sealed envelopes.

Blinding of participants and personnel (performance bias)
primary outcome

Low risk

Although blinding of the participants was not possible, initiation rate was unlikely to be affect by the performance bias.

Blinding of participants and personnel (performance bias)
secondary outcomes

High risk

Blinding of the participants was not possible which may affect the self‐reported side‐effects.

Blinding of outcome assessment (detection bias)
All outcomes

Unclear risk

No information available.

Incomplete outcome data (attrition bias)
All outcomes

High risk

There was a high rate of loss to follow‐up in both groups (23% (11/48) and 44% (21/48) in immediate insertion and delayed insertion groups respectively.

Selective reporting (reporting bias)

Low risk

The trial authors reported all potentially relevant outcomes.

Other bias

Unclear risk

Many outcomes in this study were not analysed following an intention‐to‐treat basis.

Methods

A RCT with non‐inferiority design conducted in a university hospital in Utah, USA
Study duration: January to October 2009
Sample size: a total sample size of 34 participants was required in each group when the immediate postpartum insertion of the etonogestrel implant compared to delayed postpartum insertion was expected to delay time to lactogenesis stage II with a non‐inferiority margin for the mean difference of 8 hours at the study power of 80% and alpha of 0.05.

Participants

Healthy peripartum women with healthy, term newborns who desired the etonogestrel implant for contraception.

Exclusion criteria were onset of lactogenesis before randomisation, haemorrhage requiring transfusion, severe pregnancy‐induced hypertension, prolonged hospitalisation, coagulopathy, liver disease, undiagnosed genital bleeding, or other relative contraindication to etonogestrel implant insertion (known or suspected pregnancy; known, suspected, or history of breast cancer;or hypersensitivity to any of the components in the etonogestrel implant). Women taking inducers of hepatic enzymes were also excluded, including barbiturates, griseofulvin, rifampin, phenylbutazone, phenytoin, carbamazepine, felbamate, oxcarbazepine, topiramate, modafinil, protease inhibitors, and herbal products

Most study participants were Hispanic white (73% and 60% of women assigned to immediate and delayed postpartum insertion groups, respectively).

Interventions

Intervention arm: participants received etonogestrel contraceptive implant within 1 to 3 days after delivery.
Control arm: participants received delayed insertion of etonogestrel contraceptive implant at 4 to 8 weeks postpartum visit

Outcomes

Primary outcome: time to lactogenesis stage II documented by maternal perception, lactational failure
Other outcome measures: lactational failure, postpartum contraceptive prevalence, participant‐reported vaginal bleeding, and mean creamatocrit value of breast milk at first postpartum visit

Notes

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

Participants were randomly assigned using computer‐generated random numbers in blocks of varying sizes of 2, 4, and 6.

Allocation concealment (selection bias)

Low risk

Allocation concealment was ensured because the assignments were enclosed in sequentially numbered, opaque, sealed envelopes.

Blinding of participants and personnel (performance bias)
primary outcome

Low risk

Although blinding of the participants was not possible, initiation rate was unlikely to be affect by the performance bias.

Blinding of participants and personnel (performance bias)
secondary outcomes

High risk

Blinding of the participants was not possible which may affect the self‐reported side effects.

Blinding of outcome assessment (detection bias)
All outcomes

Unclear risk

No information was available.

Incomplete outcome data (attrition bias)
All outcomes

Low risk

There was a low rate of loss to follow‐up in both groups; (17% (6/35) and 17% (6/34) in the immediate insertion and delayed insertion groups respectively.

Selective reporting (reporting bias)

Low risk

The trial authors reported all potentially relevant outcomes.

Other bias

Unclear risk

The trial authors did not provide the definitions of abnormal vaginal bleeding patterns applied in this study, which was an outcome of interest in this review.

Methods

A RCT conducted in a large hospital in North Carolina, USA
Study duration: June 1992 to February 1993
Sample size: 250 participants were recruited to detect 10% difference of prevalence of abnormal vaginal bleeding occurring among peripartum women assigned to receive immediate postpartum insertion (1 to 3 days) and delayed postpartum insertion (4 to 6 weeks) of levonogestrel contraceptive implant with a study power of 80% and alpha of 0.05

Participants

Peripartum women were offered inclusion in this trial if they intended to use contraceptive implant for contraception, had vaginal births, had predelivery gestational age of 34 weeks or more, had a spontaneous delivery of a grossly intact placenta, had a predelivery haemoglobin value of 8.0 g/dL or higher, and had no contraindications toNorplant insertion as listed by the manufacturer including acute liver disease, acute thromboembolic disorder, abnormal genital bleeding, pregnancy, breast cancer, or an allergy to silicone or levonorgestrel
Exclusion criteria included women who intended to breastfeed, experienced postpartum haemorrhage which defined as total estimated blood loss > 500 mL or infection, currently received heparin or warfarin sodium, massively obese (postpartum weight > 250 pounds), or had a history of hypertension, seizures, or bleeding disorders

250 non‐breastfeeding postpartum women who met the study inclusion were enrolled. Most women who participated in this study were black (75.2% and 72.5% of women in the immediate and delayed postpartum insertion groups, respectively)

Interventions

Intervention arm: participants received levonorgestrel contraceptive implant within 48 hours of delivery.
Control arm: participants received the same contraceptive method but had been administered at 4 to 6 weeks after delivery.

Outcomes

Primary outcome of interest was postpartum bleeding evaluated by the change of haemoglobin level and participants’ self‐reported of vaginal bleeding patterns
Secondary outcomes included blood pressure and weight changes and participants’ self‐report on other side effects including nausea, hair loss, hirsutism, headaches and acne

All outcomes were evaluated during the visit scheduled at 4 to 6 weeks after delivery

Notes

The trial investigators excluded 9 women after randomisation. The included study did not state about the definitions of abnormal vaginal bleeding patterns which were the primary safety concerns of this study

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

The allocation sequence was computer generated and sequences was numbered sequentially.

Allocation concealment (selection bias)

Low risk

Allocation concealment was ascertained by enclosing assignments in opaque, sealed envelopes.

Blinding of participants and personnel (performance bias)
primary outcome

Low risk

Although blinding of the participants was not possible, initiation rate was unlikely to be affect by the performance bias.

Blinding of participants and personnel (performance bias)
secondary outcomes

High risk

Blinding of the participants was not possible which may affect the self‐reported side effects.

Blinding of outcome assessment (detection bias)
All outcomes

Unclear risk

No information was available.

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Loss to follow‐up rates were 12% and 9% of participants assigned to immediate and delayed insertion groups, respectively.

Selective reporting (reporting bias)

Low risk

The trial authors reported all potentially relevant outcomes.

Other bias

Unclear risk

The trial investigators excluded 9 women after randomisation. In addition, no well‐defined terminology of abnormal vaginal bleeding was applied in this trial.