Effect of Follow-Up Intervals on Breastfeeding Rates 5–6 Months Postpartum: A Randomized Controlled Trial

Abstract

Objective:

To compare the effect of early versus traditional postpartum follow-up intervals on breastfeeding continuation rates 6 months postpartum.

Methods:

This randomized controlled trial enrolled primiparous women planning to breastfeed to a postpartum appointment either 2–3 weeks or 6–8 weeks after delivery. The primary outcome was the breastfeeding rate in each group 5–6 months after delivery. The study was powered to detect a 50% difference between groups assuming a 34% rate of breastfeeding at 6 months. Participants were contacted by phone 5–6 months after delivery to assess outcomes.

Results:

From March 2014 to July 2016, 649 women were screened and 344 enrolled as follows: 172 to 2–3 week and 172 to 6–8 week follow-up. Demographic, delivery, and support characteristics were similar between groups; however, average infant birth weight and the distribution of gestational ages at the time of delivery were different between groups (p < 0.05). Participants in the 2–3 week group had a breastfeeding rate of 57.7% 6 months following delivery and participants in the 6–8 week group had a rate of 59.3%. Early follow-up was associated with a relative risk of 0.97 (95% CI 0.79–1.19, p = 0.80) and an adjusted relative risk of 1.45 (95% CI 0.71–2.95, p = 0.31), when adjusted for confounding variables, for breastfeeding continuation at 5–6 months.

Conclusions:

Breastfeeding rates at 5–6 months postpartum were comparable between both groups; early follow-up was not associated with an increased rate of breastfeeding. At 6 months postpartum, the breastfeeding rate in both groups approximated the Surgeon General's Healthy People 2020 goal of 60.6%.

Clinical Trial Registration:

ClinicalTrials.gov (Identifier NCT02221895).

Introduction

The need for obstetric care, education, and assessment does not end with delivery of a pregnancy. Following childbirth, women are generally advised to return for a postpartum appointment with an obstetric care provider. Historically, the interval for this follow-up in the United States is conventionally 6–8 weeks following delivery. Evidence to support this timing may be based on antiquated guidance, such as outdated guidelines for conventional cervical cancer screening or the projected timeline for uterine involution.¹

Cervical cancer screening remains an important consideration at the postpartum visit, but other aspects of a woman's health and postpartum recovery should be considered. Tailoring follow-up guidance based on evidence may be difficult; however, little contemporary objective evidence exists to dictate an ideal follow-up interval and some women may benefit from more than one scheduled visit in the postpartum period. Until 2018, Committee Opinion from the American College of Obstetricians and Gynecologists (ACOG) posited that scheduling of the postpartum visit at 6 weeks after delivery is likely based upon the 40 days of convalescent leave typically provided to mothers.² The lack of evidence-based guidelines leaves one to question whether conventional recommendations meet the needs of most postpartum women or provide efficient means to target health concerns and optimize outcomes. The Academy of Breastfeeding Medicine outlines recommendations for infant and maternal assessment and breastfeeding support following childbirth,³ and ACOG only recently revised its guidance on postpartum care timelines to advise for a more continual assessment of maternal status, to include follow-up with a maternal health care provider within 3 weeks of delivery.⁴

There may be drawbacks to delaying the postpartum appointment until 6–8 weeks after childbirth. Among women who initiate breastfeeding, the steepest decline in breastfeeding occurs between 2 and 6 weeks postpartum.⁵ By scheduling the appointment beyond the window when women are most prone to breastfeeding discontinuation, opportunities to intervene and provide support to women who encounter breastfeeding challenges may be missed. In addition, patients' access to contraception and depression screening and treatment may be delayed. Recent expert opinion suggested that earlier postpartum follow-up may be more practical. An earlier appointment may provide an opportunity to provide care and education to women in the postpartum period,⁶ intervene with postpartum depression, and provide family planning counseling before ovulation resumes.

We hypothesized that an early postpartum follow-up interval is associated with a higher rate of breastfeeding in primiparas 5–6 months after delivery. The objective of the study was to compare the effect of an early follow-up interval with a more traditional interval upon postpartum follow-up on breastfeeding continuation. The study's primary outcome was comparison of breastfeeding rates between the two groups 6 months after delivery. The secondary outcomes included a comparative evaluation of each group's success rate in meeting the maternally stated breastfeeding goal as well as assessment of contraceptive provision.

Materials and Methods

This study was a nonblinded randomized controlled trial with intent-to-treat analysis designed to determine whether early interval follow-up at 2–3 weeks postpartum (early arm) was associated with a higher rate of breastfeeding 6 months postpartum compared to the traditional postpartum follow-up interval of 6–8 weeks (traditional arm). The study was approved by the Madigan Army Medical Center Institutional Review Board and registered as a clinical trial on ClinicalTrials.gov (identifier NCT02221895).

Eligible participants were primiparous women 18 years of age or older at the time of delivery, who received care through the Department of Obstetrics and Gynecology, delivered a live-born infant at or beyond 37 weeks 0 days gestation, and expressed a desire to breastfeed their infant. Patients were assessed for eligibility and enrolled within the first 48 hours after delivery; patients who were not available for screener assessment within the first 48 hours after delivery were excluded. Patients who were unable to speak English or read and complete the study surveys, and those who planned to leave the military health system within the 6 month follow-up window were excluded from participation. Patients with any condition deemed by the patient's provider to be an absolute contraindication to breastfeeding (such as maternal HIV/AIDS or planned use of radioactive or chemotherapeutic medications or other medication contraindicated with breastfeeding), mothers of infants with an identified fetal factor that would impair breastfeeding (such as fetal mid-facial defects, known fetal chromosomal abnormality, or known fetal conditioning resulting in fetal hypotonia), and women who were continually separated from their infant for the first 48 hours after delivery (such as maternal ICU admission or infant NICU admission) were also excluded. Participation in the study was voluntary and informed consent was obtained for eligible women who agreed to enroll.

Our primary outcome was the rate of breastfeeding continuation 5–6 months after delivery. Breastfeeding was defined, for purposes of this study, as any breast milk produced by the mother and fed to the infant by placing the infant to the breast or by pumping or manually expressing and feeding breast milk to the infant by other means. Nominal assessment of breastfeeding, or subject response of “yes” or “no” to questions regarding breastfeeding, was used to determine the primary outcome. The primary outcome was measured using a standardized, scripted telephone survey administered by one of the study investigators 5–6 months postpartum.

The secondary outcome variables included contraception provision at enrollment and at the time of the follow-up visit, contraceptive use 5–6 months postpartum, and maternal-stated goal attainment for breastfeeding duration.

To calculate the sample size, we assumed the rate of breastfeeding 6 months postpartum in our population to be 34% (Ortolano RM, Geneman B, Fairchok M. Breastfeeding trends of active duty and military dependents at a large military treatment facility. Abstracts from the America Academics of Pediatrics National Conference & Exhibition, 2013. Available at https://aap.confex.com/aap/2013/webprogram/Paper 21977.html. Retrieved March 14, 2018). Using a two-sided confidence interval with α = 0.05 and β = 0.20 powered to assess a 50% difference between the group and with a 1:1 ratio of exposed and unexposed in the sample, 133 enrollees were needed for each arm (www.openepi.com/v37/SampleSize/SSCohort.htm; Kelsey). A 50% difference between the groups was chosen because it represents a clinically significant increase that approaches the Surgeon General's 2020 goal of a 60.6% breastfeeding rate at 6 months.⁷ It also represents a sample size the investigators felt they could reasonably recruit and follow with limited resources. We projected a 30% rate of subjects lost to follow-up (based on an estimated postpartum appointment no-show rate of 15–20% in our clinic). Therefore, we sought to recruit a total sample size of 346.

Study enrollment was carried out at Madigan Army Medical Center in Tacoma, Washington. Study enrollment began in March 2014 and was completed in July 2016.

Patients were assessed for study eligibility before their hospital discharge on the postpartum ward of the hospital. Those who were eligible and agreed to enroll were given information about the study and signed a written consent form to participate. After informed consent was obtained, demographic information was collected and randomization to either the early or traditional follow-up arm occurred. Participants randomized to the early arm were scheduled for a postpartum appointment 2–3 weeks after delivery (14–27 days). Participants randomized to the traditional arm were scheduled for a postpartum appointment 6–8 weeks after delivery (42–63 days). Randomization was performed by computer-generated block randomization with block sizes of six using the RAND function in Excel^®. The randomized assignments denoted as either “Treatment A: 2–3 week follow-up” or “Treatment B: 6–8 week follow-up” were placed on a card within sequentially numbered, sealed opaque envelopes. The randomization process, card, and envelope preparation were completed by the collaborating staff, who had no direct contact with any potential enrollee. These sequentially numbered and sealed envelopes were secured within a locked box until the study investigator allocated an envelope to the enrollee. Before opening the envelope with the treatment assignment, the study investigator denoted on an enrollment log the date of subject enrollment and the sequence number of the envelope assigned to the subject. The envelope number assigned to the participant became their study number, under which all data were collected.

Following randomization to either treatment arm, postpartum appointments were scheduled for participants according to their assigned follow-up interval. The participants completed a questionnaire to delineate their plan for contraception in the postpartum period, their intended duration and exclusivity of breastfeeding, and to categorize any contraceptive or breastfeeding education that occurred during pregnancy or the inpatient postpartum stay. In addition, participants were asked to complete the Breastfeeding Attrition Prediction Tool^© (BAPT^©),⁸ a questionnaire composed of a 94-item, 6-point Likert-type scale with 4 subscales and 16 additional questions that address birth details, demographic information, and current feeding practices.⁸ The BAPT's 4 subscales are positive breastfeeding sentiment, negative breastfeeding sentiment (NBS), social and professional support, and perceived behavioral control.⁹ Administered during the early postpartum period, the tool has shown potential for identifying women at risk for breastfeeding discontinuation.^10,11 When administered postpartum, the reliability coefficients for the scales exceed 0.80 and the BAPT predicts 56.5% of women who will discontinue breastfeeding in the first 8 weeks.¹²

Discharge records were reviewed to determine if each participant received lactation consultant evaluation and whether a contraceptive method was prescribed or provided before hospital discharge. The total length of the hospital admission and the number of days the participant remained in the hospital after delivery were also recorded.

The demographic information, BAPT scores, and characteristics of prenatal and inpatient postpartum care and delivery were included in a logistic regression to calculate an adjusted relative risk for the primary outcome.

Participants' follow-up appointments occurred with a certified nurse midwife, nurse practitioner, or physician provider in the Department of Obstetrics and Gynecology at Madigan Army Medical Center or its satellite Women's Health Clinic on McChord Air Force Base. At the postpartum visit, participants completed a brief questionnaire addressing their current infant feeding method and method of breast milk expression, lactation support engaged since delivery, intended duration and exclusivity of breastfeeding, and contraception plan. Participants also completed the Edinburgh Postnatal Depression Scale (EPDS),¹³ which is a standard screening offered to all postpartum patients seen within the department. Subjects demonstrating an elevated depression screen (EPDS scores of ≥12) were assessed according to a previously approved departmental protocol and referred for behavioral health evaluation, as indicated. The documentation from the postpartum appointment was reviewed by study investigators to record the results of the EPDS, whether the participant was referred for behavioral health evaluation or to a lactation consultant, if a cervical cancer screening was indicated and performed, and if the participant received a method of contraception at the visit or had an additional visit scheduled to address their contraception plan.

At 5–6 months after childbirth, participants were contacted through telephone by a study investigator to complete a standardized, scripted phone survey that addressed their infant feeding practices. If the participant was not reached by telephone on the initial attempt, additional attempts were made weekly. If investigators were unable to contact a participant by the sixth month after childbirth, she was considered lost to follow-up.

Participants who disenrolled from the study were not included in the data evaluation. Subjects who failed to show up for any postpartum appointment within 8 weeks of delivery or who returned for a follow-up appointment outside their assigned interval were contacted for the 6-month follow-up and included in their originally assigned group for data analysis.

Data analysis was performed using IBM^® SPSS^® Statistics (version 22) software and OpenEpi (version 3.01). Maternal characteristics were compared with descriptive statistics using t test for continuous variables and χ² for categorical variables with two-sided p values. p values <0.05 were considered significant. The primary outcome analysis was evaluated using a 2 × 2 table to determine the relative risk (RR) of the exposure with a 95% confidence interval (CI) and exact p value with a χ² test to assess for difference between the two groups. Logistic regression was performed to account for confounding variables and calculate an adjusted relative risk for the primary outcome and potential confounders. An interval analysis to compare breastfeeding rate between intervention groups was performed after the enrollment and follow-up of 182 participants. At the interval analysis, the 2–3 week group had a breastfeeding of 54.8% at 6 months and the 6–8 week group had a breastfeeding rate of 62.2%. The early follow-up intervention was associated with a relative risk of 1.2 (95% CI 0.82–1.74, p 0.357) for breastfeeding discontinuation by 6 months. No significant difference to suggest harm from the intervention was identified, so the study was continued until completion.

Results

Of the 649 patients who were assessed for study eligibility, 305 were excluded and 344 were enrolled (Fig. 1). The most common indications for study exclusion were nonavailability for screener assessment in the first 48 hours after delivery and maternal intent to formula feed. Study enrollment spanned from March 2014 to July 2016; final follow-ups were completed in January 2017. Two of the 346 enrollment envelopes were not used because the integrity of the envelope could not be ascertained; they were discarded. Consequently, the final patient enrollment was 344 instead of the 346 calculated by the power analysis. Of the 344 enrolled study participants, 172 were allocated to a postpartum appointment 2–3 weeks after their delivery and 172 were allocated to a postpartum appointment 6–8 weeks after delivery. Patients who did not receive their intended intervention either failed to show up for a postpartum visit or rescheduled their visit for a time outside the assigned interval. For the primary outcome, 39 (23.1%) participants in the 2–3 week follow-up group and 36 (21.1%) participants in the 6–8 week follow-up group were lost to follow-up. In the 6–8 week follow-up group, six of the subjects who did not complete the 5–6 month follow-up survey were no longer breastfeeding at the time of their postpartum appointment; the same was true for one participant in the 2–3 week follow-up group. It was inferred that these enrollees were not breastfeeding at the 5–6 month postpartum interval and they were included for purposes of determining the primary outcome. Therefore, in the 2–3 week follow-up group, 130 participants were analyzed for the primary outcome. In the 6–8 week follow-up group, 135 participants were analyzed for the primary outcome.

FIG. 1.

Flow diagram of enrollment and follow-up.

The baseline demographic and patient characteristic variables were compared between groups. There was no significant difference between the groups in patient age, total hospital admission time, length of the postpartum stay, number of pregnancies before the current delivery, race, marital status, education level, reported income, infant sex, mode of delivery, military duty status, and the source of prenatal care. There was a difference between the groups in average infant birth weight and the distribution of gestational ages at the time of delivery. The early follow-up group had more deliveries after 41 weeks' gestational age and there were two patients determined to have an estimated gestational age between 36 and 37 weeks after enrollment in the traditional follow-up group (Table 1).

Table 1.

Subject Demographic and Delivery Characteristics

Characteristic	Early interval group (n = 169)	Traditional interval group (n = 171)	p
Age (y)	25.0 ± 4.3	25.3 ± 4.5	0.46
Total length of hospital stay (days)	3.6 ± 0.7	3.6 ± 0.6	0.42
Length of hospital stay after delivery (days)	2.0 ± 0.3	2.0 ± 0.3	1.0
Number of prior pregnancies	1.3 ± 0.7	1.3 ± 0.7	1.0
Infant delivery weight	3464.7 ± 453.1	3302.7 ± 437.7	<0.05
Race			0.35
White	99 (58.6)	103 (60.2)
Black	21 (12.4)	16 (9.4)
Hispanic	16 (9.5)	11 (6.4)
Asian/Pacific Islander	4 (2.4)	12 (7.0)
Native American	1 (0.6)	2 (1.2)
Mixed race	2 (1.2)	3 (1.8)
Other	23 (13.6)	18 (10.5)
Not stated	3 (1.8)	6 (3.5)
Marital status			0.61
Married	156 (92.3)	74 (90.2)
Single	8 (4.7)	6 (7.3)
Separated/Divorced	2 (1.2)	1 (0.6)
Not stated	3 (1.8)	3 (1.8)
Education			0.36
Grade school	0 (0.0)	1 (0.6)
High school	46 (27.2)	49 (28.7)
College	83 (49.1)	93 (54.4)
Graduate school	12 (7.1)	11 (6.4)
Not stated	28 (16.6)	17 (9.9)
Reported income			0.79
$10,000 or less	3 (1.8)	4 (2.3)
$10,001–25,000	12 (7.1)	13 (7.6)
$25,001–40,000	34 (20.1)	35 (20.5)
$40,001–55,000	14 (8.3)	24 (14.0)
$55,001–70,000	15 (8.9)	18 (10.5)
$70,001–85,000	11 (6.5)	8 (4.7)
$85,001–100,000	14 (8.3)	13 (7.6)
$100,000 or greater	8 (4.7)	5 (2.9)
Not stated	58 (34.3)	51 (29.8)
Gestational age at delivery (weeks)			<0.05
36 + 0–36 + 6	0 (0.0)	2 (1.2)
37 + 0–37 + 6	10 (5.9)	15 (8.8)
38 + 0–38 + 6	36 (21.3)	42 (24.6)
39 + 0–39 + 6	48 (28.4)	62 (36.3)
40 + 0–40 + 6	48 (28.4)	42 (24.6)
41 + 0–41 + 6	27 (16.0)	8 (4.7)
Infant sex			0.75
Male	86 (50.9)	84 (49.1)
Female	83 (49.1)	87 (50.6)
Mode of delivery			0.54
Vaginal	133 (78.7)	141 (82.5)
Operative vaginal	7 (4.1)	8 (4.7)
Cesarean birth	29 (17.2)	22 (12.9)
Military duty status			0.10
Active duty	55 (32.5)	42 (24.6)
Nonactive duty/dependent	114 (67.5)	129 (75.4)
Source of prenatal care			0.20
OB	106 (62.7)	94 (55.0)
CNM	26 (15.4)	35 (20.5)
Centering™ Pregnancy	35 (20.7)	42 (24.6)
Other	2 (1.2)	0 (0.0)

Enrollment information characterizing maternal intent for breastfeeding exclusivity and planned duration of breastfeeding, mean scores on the four subscales of the BAPT, report of breastfeeding education before delivery, and the rate of lactation consultant evaluation before hospital discharge was not statistically different between the two groups. Provision of contraception at the time of discharge and the intended methods of contraception at the time of discharge was also similar between groups (Table 2).

Table 2.

Characteristics of Lactation and Contraception Support and Planning at Enrollment

Variable	Early interval group (n = 169)	Traditional interval group (n = 171)	p
Seen by lactation consultant before discharge			0.42
Yes	98 (58.0)	111 (64.9)
No	60 (35.5)	51 (29.8)
No information	11 (6.5)	9 (5.3)
Intended breastfeeding exclusivity			0.60
Breast milk only	113 (66.9)	117 (68.4)
Breast milk and formula	14 (8.3)	18 (10.5)
No response	42 (24.9)	36 (21.1)
Intended breastfeeding duration			0.98
Unsure	11 (6.5)	12 (7.0)
1–3 months	1 (0.6)	2 (1.2)
3–6 months	12 (7.1)	13 (7.6)
6–9 months	26 (15.4)	27 (15.8)
9–12 months	48 (28.4)	52 (30.4)
>12 months	28 (16.6)	28 (16.4)
No response	43 (25.4)	37 (21.6)
BAPT^©scores
NBS	183.7 ± 69.1	174.9 ± 82.2	0.36
PBS	375.2 ± 56.4	370.7 ± 56.0	0.53
SPS	253.9 ± 109.0	244.4 ± 108.4	0.49
BFC	47.6 ± 5.2	47.8 ± 9.3	0.81
Report of breastfeeding education before delivery			0.70
Yes	67 (39.6)	73 (42.7)
No	59 (34.9)	61 (35.7)
No response	43 (25.4)	37 (21.6)
Prescription for contraception method provided at discharge			0.13
Yes	96 (56.8)	83 (48.5)
No	73 (43.2)	88 (51.5)
Intended method of contraception at discharge			0.56
None	2 (1.8)	5 (2.9)
Unsure	11 (6.5)	10 (5.8)
Abstinence	0 (0.0)	0 (0.0)
Natural family planning	0 (0.0)	3 (1.8)
Lactational amenorrhea	1 (0.6)	2 (1.2)
Progestin contraceptive	32 (19.2)	31 (18.1)
Combined contraceptive	15 (8.9)	12 (7.0)
Medroxyprogesterone acetate	2 (1.2)	5 (2.9)
Condoms/barrier method	14 (8.3)	17 (9.9)
Intrauterine device	21 (12.5)	24 (14.0)
Contraceptive implant	13 (7.7)	19 (11.1)
Female sterilization	2 (1.2)	0 (0.0)
Male sterilization	0 (0.0)	1 (0.6)
Multiple methods	15 (8.9)	11 (6.4)
No response	39 (23.2)	31 (18.1)

Within the early and traditional follow-up groups, 91.7% and 88.9% of participants, respectively, presented for a postpartum visit (p = 0.38). Within the early follow-up group, 83.4% were seen during their assigned interval; 86.5% of participants in the traditional interval group were seen at their assigned interval (p = 0.42).

At the 5–6 month follow-up, participants in the 2–3 week group had a breastfeeding rate of 57.7% and participants in the 6–8 week group had a breastfeeding rate of 59.3% (p = 0.80). The early follow-up intervention was associated with a relative risk of 0.97 (95% CI 0.79–1.19, p = 0.80) for breastfeeding continuation at the time of the 5–6 month follow-up. Using logistic regression to account for potential confounding variables, the adjusted relative risk of breastfeeding continuation in the early follow-up group was 1.45 (95% CI 0.71–2.95, p = 0.31). Planned breastfeeding exclusivity at the time of enrollment, the BAPT NBS score, and nonoperative vaginal delivery were protective for continued breastfeeding 5–6 months after delivery (Table 3).

Table 3.

Binary Logistic Regression for Primary Outcome to Adjust for Confounding Variables at the Time of Enrollment

Variable	Adjusted RR (95% CI)	p
Early postpartum follow-up interval	1.45 (0.71–2.95)	0.31
Seen for postpartum examination	2.01 (0.58–6.93)	0.27
Prenatal care by midwife	0.74 (0.28–1.95)	0.54
Prenatal care in group care model	0.83 (0.35–1.97)	0.67
Total number of prenatal visits	1.03 (0.98–1.09)	0.27
Breastfeeding education reported	1.26 (0.62–2.55)	0.52
Active duty at time of delivery	0.66 (0.28–1.59)	0.36
Total length of hospital stay	1.41 (0.84–2.40)	0.19
Length of postpartum stay	0.71 (0.25–2.07)	0.53
Hormonal contraception prescribed	0.61 (0.30–1.25)	0.18
Inpatient lactation consultant visit	0.78 (0.38–1.60)	0.49
Planned exclusive breastfeeding	3.01 (1.01–8.91)	< 0.05
Planned breastfeeding duration >6 months	1.63 (0.67–3.96)	0.28
BAPT NBS	1.01 (1.00–1.01)	<0.05
BAPT PBS	0.99 (0.99–1.00)	0.08
BAPT SPS	1.00 (1.00–1.00)	0.48
BAPT BFC	1.00 (0.95–1.06)	0.96
Age	0.92 (0.84–1.01)	0.12
Vaginal delivery	2.37 (1.04–5.46)	<0.05
Gestational age at delivery <39 weeks	0.79 (0.34–1.81)	0.57
Infant birth weight	1.00 (1.00–1.00)	0.74
Infant sex, male	0.95 (0.48–1.88)	0.88
White ethnicity	0.76 (0.36–1.62)	0.48
Annual income > $40,000	1.09 (0.49–2.41)	0.84
Married	0.89 (0.26–3.01)	0.85
Education level, some college or greater	1.51 (0.67–3.41)	0.32

NBS, negative breastfeeding sentiment; PBS, positive breastfeeding sentiment; SPS, social and professional support.

At the time of final follow-up, there was no significant difference in the percentage of women who met their initial stated goal for breastfeeding duration. When reported intent for breastfeeding duration was reassessed at the 5–6 month follow-up, there was no difference between the groups. There was also no difference in the percentage of women who reported using infant formula or the percentage of women who reported infant supplementation with foods. Both groups demonstrated the largest initiation in formula use in the first month after delivery, with another peak in the second month. The uptake in formula use coincided with the reported peaks for breastfeeding cessation in the first month after delivery and again at the third postpartum month, which was not statistically different between groups (Table 4).

Table 4.

Breastfeeding, Infant Feeding, and Contraceptive Characteristics at 5–6 Months

Variable	Early interval group	Traditional interval group	p
Initial maternal goal for breastfeeding duration met	(n = 169)	(n = 171)	0.98
Yes	67 (39.6)	65 (38.0)
No	30 (17.8)	32 (18.7)
Initial maternal goal not stated	39 (23.1)	37 (21.6)
Unsure intent	8 (4.7)	8 (4.7)
Lost to follow-up	25 (14.8)	29 (17.0)
Interval between delivery and breastfeeding discontinuation	(n = 53)	(n = 50)	0.85
<1 month	5 (9.4)	6 (12.0)
1 month	10 (18.9)	9 (18.0)
2 months	0 (0.0)	0 (0.0)
3 months	21 (39.6)	24 (48.0)
4 months	11 (20.8)	8 (16.0)
5 months	5 (9.4)	2 (4.0)
6 months	1 (1.9)	1 (2.0)
Stated initial intent for breastfeeding duration at 5–6 month follow-up	(n = 129)	(n = 130)	0.44
1 month	0 (0.0)	1 (0.8)
3 months	7 (5.4)	6 (4.6)
6 months	47 (36.4)	40 (30.8)
1 year	69 (52.7)	72 (55.4)
1–2 years	7 (5.4)	8 (6.2)
>2 years	0 (0.0)	3 (2.3)
Infant eating solid foods	(n = 127)	(n = 130)	0.43
Yes	81 (63.8)	89 (68.5)
No	46 (36.2)	41 (31.5)
Formula supplementation	(n = 128)	(n = 129)	0.19
Yes	69 (53.9)	59 (46.1)
No	59 (46.1)	70 (54.3)
Infant age when formula supplementation started	(n = 69)	(n = 59)	0.51
<1 month	41 (20.3)	14 (23.7)
1 month	12 (17.4)	18 (30.5)
2 months	27 (39.1)	18 (20.5)
3 months	10 (14.5)	6 (10.2)
4 months	2 (2.9)	2 (3.4)
5 months	1 (1.4)	0 (0.0)
not stated	3 (4.3)	1 (1.7)
Current contraceptive practice	(n = 128)	(n = 130)
None	10 (7.0)	8 (6.2)	0.52
Abstinence	9 (7.0)	9 (6.9)
Natural family planning	15 (11.7)	7 (5.4)
Lactational amenorrhea	0 (0.0)	2 (1.5)
Hormonal contraceptive (oral, ring, and patch)	40 (31.3)	43 (43.1)
Medroxyprogesterone acetate	4 (3.1)	2 (1.5)
Condoms/barrier method	13 (10.2)	15 (11.5)
LARC, not specified	36 (28.1)	43 (33.1)
Female sterilization	1 (0.8)	0 (0.0)
Male sterilization	0 (0.0)	1 (0.8)

LARC, long-acting reversible contraceptive.

At the 5–6 month follow-up, the contraceptive methods in use were comparable between the two groups. Hormonal contraception (progestin or combined pill, transdermal patch, or vaginal ring) was the most common reported means of contraception, followed by a long-acting reversible contraceptive (LARC) method (intrauterine device or subdermal implant). The relative risk of LARC use was 0.85 (95% CI 0.59–1.23, p = 0.40) in the 2–3 week follow-up group (Table 4).

Data collected at the time of postpartum examination indicated that exclusive formula feeding was significantly less likely in the 2–3 week follow-up group (OR 0.18, 95% CI 0.06–0.50, p < 0.05). Based upon the information collected about the self-reported timing of breastfeeding discontinuation at the 5–6 month follow-up, this suggests that the period between 2–3 and 6–8 weeks postpartum is a peak time for breastfeeding attrition.

At the postpartum visit, the groups demonstrated no significant difference in intended breastfeeding duration or contraception use, planned LARC use, self-referral for lactation assistance, or breast pump use. There was also no significant difference in the EPDS scores and the rate of referral for Behavioral Health services. Women in the early follow-up group were less likely to have cervical cancer screening performed and more likely to be referred to a lactation consultant by their provider (Table 5).

Table 5.

Characteristics at the Postpartum Examination

Variable	Early interval group (n = 155)	Traditional interval group (n = 152)	p
Interval between delivery and postpartum appointment (days)	22.3 ± 8.26	44.74 ± 4.77
Exclusivity of breastfeeding			<0.05
Breast milk only	90 (58.1)	79 (52.0)
Formula only	4 (2.6)	22 (14.5)
Breast milk and formula	41 (26.5)	35 (23.0)
Breast milk, unknown exclusivity	16 (10.3)	11 (7.2)
Not stated	4 (2.6)	5 (3.3)
Intended breastfeeding duration			0.74
Unsure	11 (7.1)	9 (5.9)
<1 month	2 (1.3)	0 (0.0)
1–3 months	3 (1.9)	3 (2.0)
3–6 months	18 (11.6)	11 (7.2)
6–9 months	25 (16.1)	16 (10.5)
9–12 months	34 (21.9)	29 (19.1)
>12 months	21 (13.5)	26 (17.1)
Breastfeeding discontinued since enrollment	3 (1.9)	14 (9.2)
No response	38 (24.5)	44 (28.9)
Intended method of contraception at the time of postpartum visit			0.07
None	4 (2.6)	9 (5.9)
Unsure	3 (1.9)	10 (6.6)
Abstinence	1 (0.6)	1 (0.7)
Natural family planning	0 (0.0)	0 (0.0)
Lactational amenorrhea	4 (2.6)	0 (0.0)
Hormonal contraceptive	64 (41.3)	43 (28.3)
Progestin	51 (32.9)	29 (19.1)
Combined	13 (8.4)	14 (9.2)
Medroxyprogesterone acetate	5 (3.2)	7 (4.6)
Condoms/barrier method	11 (7.1)	18 (11.8)
Intrauterine device	17 (11.0)	20 (13.2)
Contraceptive implant	21 (13.5)	19 (12.5)
Female sterilization	0 (0.0)	0 (0.0)
Male sterilization	1 (0.6)	1 (0.7)
Multiple methods	20 (12.9)	15 (9.9)
Include IUD	3 (1.9)	3 (2.0)
Include implant	5 (3.2)	2 (1.3)
Other	0 (0.0)	1 (0.7)
Not stated	4 (2.6)	8 (5.3)
LARC requested at postpartum visit	46 (30.3)	53 (35.1)	0.37
Sought breastfeeding assistance since discharge
Yes	65 (41.9)	62 (40.8)	0.89
No	51 (32.9)	48 (31.6)
No response	39 (25.2)	42 (27.6)
Referred to lactation consultant
Yes	12 (7.7)	3 (2.0)
No	105 (67.7)	108 (71.1)	<0.05
Declined referral	7 (4.5)	0 (0.0)
Already engaged with lactation consultant	31 (20.0)	41 (27.0)
Use of breast pump or hand expression since hospital discharge
Yes	112 (72.3)	108 (71.1)	0.19
No	7 (4.5)	2 (1.3)
No response	36 (23.2)	42 (27.6)
EPDS score
0–11	123 (79.4)	119 (78.3)	0.66
12–13	5 (3.2)	6 (3.9)
≥14	2 (1.3)	5 (3.3)
Not completed	25 (16.1)	22 (14.5)
Referred to behavioral health
Yes	6 (3.9)	5 (3.3)	0.55
No	140 (90.3)	136 (89.5)
Patient declined referral	6 (3.9)	4 (2.6)
Already engaged with behavioral health	3 (1.9)	7 (4.6)
Cervical cytology indicated
Yes	35 (22.6)	36 (23.7)	0.82
No	120 (77.3)	116 (76.3)
Cervical cytology performed
Yes	20 (12.9)	33 (21.7)	<0.05
No	135 (87.1)	119 (78.3)

EPDS, Edinburgh Postnatal Depression Scale.

There were no reported harms or unintended effects seen in either group as a result of the intervention.

Discussion

Our study demonstrates no significant difference in breastfeeding rates 6 months after delivery among women who were scheduled for an early postpartum appointment 2–3 weeks after delivery when compared to the traditional timing of the postpartum appointment at 6–8 weeks after delivery. The observed rate of breastfeeding within the study population approximated the 2020 goal of 60.6% set forth by the United States Surgeon General and was higher than the estimated breastfeeding rate used to inform the power analysis. An intent for breastfeeding exclusivity at the time of enrollment, nonoperative vaginal delivery, and the BAPT NBS score were associated with breastfeeding continuation at 5–6 months. The data suggest that the period between 2–3 and 6–8 weeks after delivery is a time of significant breastfeeding attrition. Contraceptive use (including LARC use), self-referral for lactation support services, depression screening results, and referrals for behavioral health support were similar between groups. Cervical cancer screening was less likely to be performed and women were more likely to be referred to a lactation consultant at the postpartum encounter in the 2–3 week follow-up group.

Although 6–8 weeks after delivery is common interval for the postpartum appointment in the United States, expert opinion recently published by ACOG and the Academy of Breastfeeding Medicine's “Going Home Protocol” is working to transform this standard.^2–4 This study provides objective evidence to support that the timing of the postpartum appointment and lactation support offered at this visit may not significantly impact long-term breastfeeding outcome. It is likely that multiple other variables, such as intent for breastfeeding exclusivity and women's perceptions of breastfeeding support, contribute more significantly to breastfeeding success. Therefore, some flexibility and individualization of the timing for any postpartum encounters may better address this population's unique health considerations.

The postpartum appointment provides an opportunity for patient education, referral for additional care, contraception provision, preventative health care, and depression screening. Given that different patients have different needs, there may be a role for individualizing the timing of the postpartum appointment for different populations and patient characteristics. Targeted or enhanced breastfeeding support and education directed at women identified to be at risk for early breastfeeding cessation or lower rates of breastfeeding initiation may be a key strategy. In a large prospective cohort, Hobbs et al. identified that women undergoing planned Cesarean section were less likely to plan to breastfeed, less likely to initiate breastfeeding, and more likely to discontinue breastfeeding by 12 weeks postpartum. Furthermore, women who underwent unplanned or emergency Cesarean reported more difficulty with breastfeeding and required more resources for lactation support while inpatient and after discharge.¹⁴ Another study identified that women with symptoms of depression or anxiety are less likely to initiate breastfeeding and show reduced durations of breastfeeding. Women who had a mental health visit for depression or anxiety before pregnancy were much less likely to initiate breastfeeding than those who had no mental health visits for that purpose. Women who exhibited symptoms of postpartum depression were also less likely to breastfeed or exclusively breastfeed at 3 months postpartum.¹⁵ These women may benefit from one or more strategically timed follow-up intervals and antenatal interventions to assess infant feeding practices and plans; these interactions could serve as additional opportunities to provide evaluation and resources for behavioral health concerns, contraceptive planning, and preventative health care and education.

Lactation consultants, nurses, and health care providers all contribute to meet the needs of women in the postpartum period. ACOG advises that infant feeding plans should be addressed during a postpartum encounter and lactation support staff should be a part of the postpartum care team.² This inclusion highlights the priority that professional organizations place on breastfeeding support as part of a comprehensive postpartum visit.

Contraceptive planning is another component of the postpartum visit.² Contraception provision in the postpartum period may reduce the rate of unintended and close-interval pregnancies. In the United States, approximately half of women rely on no contraception or less-effective methods of contraception in the 18 months following childbirth.¹⁶ While use of LARC methods is in high demand among postpartum women, many are unable to access their method of choice.¹⁷ Facilitating follow-up in the postpartum period may provide additional access points for women desiring LARC methods. An earlier postpartum visit could provide an opportunity to meet patients' contraceptive needs before resumption of ovulation.

This study has several limitations. The population was located in the state of Washington, which was identified in the CDC 2016 Breastfeeding Report Card as one of only a few states that met the Healthy People 2020 goal rates for both breastfeeding initiation and 6-month duration of breastfeeding.¹⁸ Breastfeeding rates show a geographic trend in the United States, so our population's rates may be influenced by regional breastfeeding practices and trends. Our study population is unique, in that our beneficiaries either serve in the military or are dependents of military Servicemembers. Subjects had access to prenatal care, obstetric education, and contraception as a component of their medical benefits. Access to services such as sterilization, LARC, support from a lactation consultant and behavioral health specialists, and the choice between physician, midwife, and group obstetric care are included in our open access and collaborative care model. In addition, patients have no direct financial burden when accessing medical care and can obtain their contraceptive selection at no cost; there is no financial disincentive to provide LARC methods or sterilization immediately postpartum, and this was a practice available to patients during the study period. These factors may not mirror the availability and practices outside the setting of military medicine. Due to the eligibility requirements for care, our population excluded single mothers other than those serving in the military. Women serving in the Armed Forces have a guaranteed, paid maternity benefit. When the study began, Servicemembers were guaranteed 6 weeks of paid maternity leave. In January 2016, a Department of Defense policy change occurred, which guaranteed all female active duty Servicemembers 12 weeks of paid maternity leave.¹⁹ These factors may have made our population, specifically the single/unmarried population, different from the general population. However, our population included a broad range of ages, backgrounds, and socioeconomic statuses, which is congruent with the general population.

In addition, the study did not collect information pertaining to patient and provider satisfaction related to the timing of the follow-up visit. In the absence of any decisive, objective finding to argue for a change in clinical practice, this subjective information may have been useful.

The strengths of the study include its randomized controlled design and the multiple variables collected at distinct points from the time of the inpatient postpartum hospitalization until 5–6 months postpartum. Although the study was not powered to look specifically at contraceptive use, depression screening, and treatment for behavioral health concerns, these variables were included in the analysis and can guide the reader when contemplating whether the study finding justifies an alteration in clinical practice.

Conclusion

Understanding how women decide to breastfeed and the factors that support decisions for planned breastfeeding exclusivity may be keys to improving future breastfeeding rates. These decisions are likely multifactorial and may occur well in advance of pregnancy or delivery. Health care support in the postpartum period may have less magnitude than the decisions, education, and social constructs that influence women's infant feeding choices before delivery. Our study suggests that earlier contact with a health care provider in the postpartum period does not affect this outcome. Importantly, an earlier postpartum visit did not negatively impact breastfeeding rates. As certain groups of patients are at risk for low breastfeeding initiation or early breastfeeding discontinuation; providers can utilize this information to create an individual postpartum care plan for their patients and decide on appropriate appointment timing. Tailoring the timing of this visit to the unique considerations of each patient would help to better serve patient needs in this critical period.

Better understanding of the factors that inform infant feeding choices would allow us to better direct efforts for improvements. As postpartum care and routine preventative health care are basic tenets of women's health care, the findings of our study are relevant for general obstetric practice. Disproval of our hypothesis is just as impactful in targeting future elements of clinical practice.

Footnotes

Acknowledgment

We acknowledge the support of the Department of Clinical Investigation, Madigan Army Medical Center, in the completion of this study.

Disclosure Statement

No competing financial interests exist.

References

Tilley

, Mishell

. How to choose a contraceptive for your postpartum patient. OBG Manage March, 2012; 24:36–41.

American College of Obstetricians and Gynecologists. Optimizing postpartum care: Committee Opinion No. 666. Obstet Gynecol, 2016; 127:e187–e192.

ABM Clinic Protocol #2: Guidelines for hospital discharge of the breastfeeding term newborn and mother: “The going home protocol.”. Breastfeed Med, 2014; 9:3–8.

American College of Obstetricians and Gynecologists. Optimizing postpartum care: Committee Opinion No. 736. Obstet Gynecol, 2018; 131:e140–e150.

Bonuck

, Trombley

, Freeman

, et al. Randomized, controlled trial of a prenatal and postnatal lactation consultant intervention on duration and intensity of breastfeeding up to 12 months. Pediatrics, 2005; 116:1413–1426.

Speroff

, Mishell

. The postpartum visit: It's time for a change in order to optimally initiate contraception. Contraception, 2008; 78:90–98.

The Surgeon General's Call to Action to Support Breastfeeding. Available at www.surgeongeneral.gov/library/calls/breastfeeding/calltoactiontosupportbreastfeeding.pdf (accessed January 23, 2015).

Janke

Breastfeeding Attrition Prediction Tool (BAPT). Anchorage, AK: University of Alaska Anchorage, School of Nursing and Health Science, 2008.

Gill

, Reifsnider

, Lucke

, et al. Predicting breast-feeding attrition: Adapting the breast-feeding attrition prediction tool. J Perinat Neonatal Nurs, 2007; 21:216–224.

10.

Janke

. Prediction of breastfeeding attrition: Instrument development. Appl Nurs Res, 1992; 5:48–53.

11.

Dick

, Evans

, Arthurs

, et al. Predicting early breastfeeding attrition. J Hum Lact, 2002; 18:21–28.

12.

Evans

, Dick

, Lewallen

, et al. Modified Breastfeeding Attrition Prediction Tool: Prenatal and postnatal tests. J Perinat Educ, 2004; 13:1–8.

13.

Cox

, Holden

, Sgovksy

. Edinburgh postnatal depression scale. Br J Psychiatry, 1987; 150:782–786.

14.

Hobbs

, Mannion

, McDonald

, et al. The impact of cesarean section on breastfeeding initiation, duration, and difficulties in the first four months postpartum. BMC Pregnancy Childbirth, 2016; 161–169.

15.

Wouk

, Stuebe

, Meltzer-Brody

. Postpartum mental health and breastfeeding practices: An analysis using the 2010–2011 pregnancy risk assessment monitoring system. Matern Child Health J, 2017; 21:636–647.

16.

White

, Teal

, Potter

. Contraception after delivery and short interpregnancy intervals among women in the United States. Obstet Gynecol, 2015; 125:1471–1477.

17.

Potter

, Hopkins

, Aiken

, et al. Unmet demand for highly effective postpartum contraception in Texas. Contraception, 2014; 90:488–495.

18.

Centers for Disease Control and Prevention. Breastfeeding Report Care, United States, 2016. Available at www.cdc.gov/breastfeeding/pdf/2016breastfeedingreportcard.pdf (accessed 15 March, 2018).

19.

Department of Defense. Carter announces 12 weeks paid military maternity leave, other benefits. Available at: www.defense.gov/News/Article/Article/645958/carter-announces-12-weeks-paid-military-maternity-leave-other-benefits/ (accessed 15 March, 2018).