Tang 2014.

Methods	Location: Raleigh, North Carolina (USA) Recruitment time: May 2011 to January 2012 Sample size calculation and outcome of focus: Assumed 20% LARC use in control, based on state survey results. Analysis population of 626 (313 in each arm) required for 80% power using two‐side alpha .05 to detect 10% increase to 30% LARC use in intervention arm. Assuming 20% drop‐out rate, 800 planned.
Participants	General with N: 800 healthy postpartum women Source: WakeMed Hospital Inclusion criteria: Admitted to postpartum unit at hospital Delivery of live infant > 24 weeks gestational age Age 14 to 45 years Ability to speak English or Spanish fluently Willing to be contacted by phone until at least 8 weeks after delivery Exclusion criteria: History of tubal ligation or hysterectomy Partner had a vasectomy Fertility treatment to conceive this pregnancy Previous randomization into this study
Interventions	1) Intervention: routine postpartum counseling plus one‐minute script (LARC script), given to women during postpartum admission. Script informs about long‐acting reversible contraceptive (LARC) methods, i.e., contraceptive implant and IUD 2) Control: routine postpartum counseling (no LARC script and no standard counseling)
Outcomes	Primary: self‐reported LARC use Secondary: self‐reported interest in LARC use; use of any contraceptive method; reasons for not using contraceptive method of choice Assessment by 10‐min phone survey (25 items) after 6‐week postpartum visit
Notes	Due to error in preparing allocation envelopes, 394 allocated to LARC script (15 incorrectly) and 403 to no LARC script (19 incorrectly).
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Simple 1:1 randomization scheme by computer generated blocks of 6, 8 and 10; developed and stored by statistician at North Carolina Translational and Clinical Sciences Institute
Allocation concealment (selection bias)	Low risk	Sequentially numbered, opaque, sealed envelopes
Blinding (performance bias and detection bias) All outcomes	Low risk	Double blind (investigator and outcomes assessor); research assistants who opened allocation envelope were not same as those who completed participant follow‐up survey.
Incomplete outcome data (attrition bias) All outcomes	Low risk	Loss to follow‐up: intervention 31/400 (8%); control 31/400 (8%)