BLOSSoM: Improving Human Milk Provision in Preterm Infants Through Texting Support.

Mother's own milk (MOM) reduces complications of preterm birth. Despite high initiation rates of expression, half of preterm infants do not receive MOM at discharge. Frequent outreach and a short message service (SMS) have improved MOM provision in term dyads. We aimed to improve MOM provision rate from 61% to >80% by implementing standardized lactation education and Breastfeeding & Lactation Outreach via SMS Supporting Mothers (BLOSSoM).
Methods: The baseline period was June 2019 to April 2020. A multidisciplinary team implemented PDSA cycles: education/documentation (standardized lactation education and education documentation, May 2020-April 2021), and BLOSSoM (SMS program providing educational texts/videos, reminders, 2-way communication with neonatal intensive care unit (NICU) lactation, May 2021-December 2021). The primary outcome was MOM provision at NICU discharge/transfer for infants younger than 34 weeks, as analyzed on the SPC chart. BLOSSoM participants evaluated the program using a 5-point Likert scale.
Results: Demographic and clinical characteristics were unchanged among the three periods. However, the monthly MOM provision rate improved from 61% to 81%. Eighty-seven percent of BLOSSoM participants completed the evaluation with 83% rating the program most supportive, 78% rating the videos as the most helpful, followed by team check-ins (54%) and 2-way texting (24%). Conclusions: Using a multidisciplinary approach, we improved the monthly MOM provision rate at discharge/transfer for preterm infants. SMS providing educational texts/media and 2-way communication supporting lactating NICU mothers was critical to our success. Providing another method of communication through SMS was well accepted and valued by the majority.

INTRODUCTION

Mother’s own milk (MOM) reduces preterm birth complications, including necrotizing enterocolitis, sepsis, retinopathy of prematurity, and poor neurodevelopmental outcomes.[1-4] Unfortunately, despite increasing initiation rates of MOM expression,[5] approximately half of preterm infants do not receive MOM at neonatal intensive care unit (NICU) discharge/transfer.[6,7] Furthermore, because the benefit of MOM is dose-dependent, the MOM amount and duration the infant receives are critical.[8-10]

Lower MOM provision is associated with Hispanic and Black race and lower socioeconomic status.[11,12] These populations face additional challenges in providing MOM, such as suboptimal access to the hospital, lack of support at home, need to return to work or school sooner, and caring for other children.[13,14] Therefore, targeting various barriers to initiation and, more importantly, maintaining MOM through interventions embraced by families is essential.

In our mostly Hispanic and Medicaid-insured population, the baseline monthly MOM provision rate within 24 hours of initial discharge/transfer for mothers of infants younger than 34 weeks of gestational age (GA) was 61%. Our SMART aim was to improve MOM provision percentage at discharge/transfer to more than 80% over 19 months.

METHODS

Setting

Our NICU is a 25-bed level III nursery in an urban academic safety-net hospital with 3500 deliveries and 400 NICU admissions annually. Most infants are inborn, discharged home, and Medicaid-insured. Approximately a quarter of infants are born younger than 34 weeks of GA. The unit is staffed by four neonatologists, five advanced practice providers (APPs), and pediatric residents rotating monthly. NICU-dedicated International Board Certified Lactation Consultants (FTE 1.2) provide lactation support. Currently, the unit is not Baby-Friendly designated, nor does it have other lactation support such as peer-counselor programs.

Inclusion and Exclusion Criteria

Mothers with infants born younger than 34 weeks of GA admitted to the NICU were eligible. Exclusion criteria included maternal substance use per unit guideline, mothers who declined to provide MOM at NICU admission, and mothers whose infants died within 24 hours of birth. Exclusion criteria also included late preterm and term infants due to short length of stay (LOS) and their already high MOM provision rates.

Interventions

PDSA Cycle 1: Formulated a Multidisciplinary Team (May 2020)

We formed a multidisciplinary team composed of NICU lactation consultants (LCs), nurses, and neonatologists. The team identified barriers using a key driver diagram (Fig. 1) focusing on 2 primary drivers: (1) inadequate and inconsistent lactation education and (2) inadequate maternal lactation support during hospitalization.

Fig. 1.

Key driver diagram showing aims, key drivers, and change concepts.

PDSA Cycle 2: Standardized Lactation Education for Families (June 2020)

Mothers admitted with impending premature delivery received a prenatal consult by an advanced practice provider or a pediatric resident who discussed the importance of MOM and recommendation for donor breast milk (DBM), although the information provided was inconsistent.

To standardize lactation education, the team created a verbal script for providers that incorporated educational videos from NeoQIC’s Human Milk QI Collaboration (https://www.neoqicma.org/human-milk-educational-videos)[15] with permission from Parker et al. The videos consisted of 5 parental narratives on the importance and benefit of MOM, including use of colostrum, hand expression and pumping, skin-to-skin care, and finding support while in NICU; all were available in Spanish. Families received a handout with a quick response code, and these videos were accessible anytime on their phones. The video link was provided again by LCs on NICU admission.

PDSA Cycle 3: Standardized Documentation of Education Delivery (October 2020)

Providers documented prenatal consults in the electronic medical record (EMR) using a note template which did not include whether lactation education or handout with educational video quick response code were provided. To standardize education delivery, the team added verbal education, video link provision, language preference, and interpreter used to the note template as mandatory discrete fields.

PDSA Cycle 4: Breastfeeding and Lactation Outreach via SMS Supporting Mothers: BLOSSoM (May 2021)

Before Breastfeeding & Lactation Outreach via SMS Supporting Mothers (BLOSSoM), NICU LCs consulted mothers within 48 hours of all NICU infants’ admission. Mothers received a packet of lactation education, a pump log, set up with pump and supplies, instructed on acquiring a pump, and offered a rental pump until their pump was available. LCs were available 5 days per week during the day and called mothers biweekly if not seen in person. However, depending on the unit’s acuity and census, LCs availability was variable.

To provide consistency, the team created text messages to be sent to mothers automatically on a predetermined schedule from NICU admission through discharge/transfer. These messages contained educational tips, hyperlinks to handouts and videos (NeoQIC videos, hand expression, how to latch, orientation to NICU), reminders, and “checking-in” questions to engage mothers and encourage MOM provision. Mothers of infants 23-27 6/7 and 28-33 6/7 weeks’ GA received texts with different messages due to LOS differences. In both series, mothers received 3 messages spread throughout the day on the first 2 days, 2 messages per day on days 3 through 5, a daily message through day 12, a message every other day through day 25, then a message every 3–4 days until the time of discharge/transfer or mothers stopped providing MOM. All contents were available in Spanish. Mothers also could actively text with the team. Providers sent messages through PreventionPays, a Health Insurance Portability and Accountability Act of 1996 (HIPAA)- compliant 2-way messaging web-based platform, while mothers received and sent texts on their phones. Team members could turn on notifications on their phones to be alerted when a mother texted, allowing prompt response on the platform. Two NICU LCs, an Nurse, and a neonatologist managed the program during business hours. The web-based platform recorded all communication allowing ease of cross-coverage between team members.

All mothers who did not meet exclusion criteria were approached for participation. Once the mother agreed to participate, team members initiated texts. Spanish-speaking mothers received texts in Spanish and support from a bilingual WIC (the Special Supplemental Nutrition Program for Women, Infants, and Children) breastfeeding peer counselor. LCs continued to provide in-person support when mothers were present in NICU and additional phone support when needed.

Participants were offered small incentives to view the initial education videos ($10) and provide MOM ($5/wk). After the discharge or transfer of the infant or the mother stopped providing MOM, she was no longer included in the BLOSSoM program and did not receive any additional texts. No mothers opted out of BLOSSoM, although that was an option. All BLOSSoM participants received outpatient lactation resources and a text exit questionnaire for program evaluation. The LCs called mothers if they did not respond to the text questionnaire. A Denver Health Foundation grant (Cayuse # 20-0586) funded the incentives and the annual platform cost ($1500) (Fig. 2).

Fig. 2.

Flowchart of BLOSSoM.

Study of the Interventions

Data were extracted from EMR on all infants younger than 34 weeks GA admitted to our NICU retrospectively during the baseline period (June 2019–April 2020) and prospectively during intervention periods (May 2020–December 2021). Mothers with twin pregnancies were included once in the maternal analysis, but data for the infant characteristics included data from all infants. In addition, 4 authors (M.H., K.H., C.R., B.B.) manually collected data for the completion of viewing initial education videos, BLOSSoM participation, and maternal feedback.

Measures

The primary outcome measure was the monthly percentage of any MOM provision within 24 hours of initial NICU discharge/transfer by mothers of infants younger than 34 weeks of GA plotted on a Statistical Process Control (SPC) chart. Evaluation of maternal satisfaction for the support received through BLOSSoM used a 5-point Likert scale (1 = not supportive, 5 = very supportive). BLOSSoM participants rated the frequency of the texts (1 = too little, 2 = just right, 3 = too much, 4 = way too much). Participants identified which interventions were helpful and were allowed to choose more than 1 intervention.

The process measures were (1) delivery of standardized verbal education and video link during the prenatal consult, (2) use of interpreters for non-English speaking mothers during the prenatal consult, (3) completion of educational video viewing, and (4) initial willingness to participate in BLOSSoM. As donor milk use may increase as an unintentional consequence of education around providing breast milk to preterm infants, the balancing measure was the amount of DBM in milliliters used per hospital day.

Analysis

The percentage of mothers providing MOM at NICU discharge/transfer per month was plotted on an SPC p-chart using QI Charts Version 2.0.22 (Scoville Associates, TX). The upper and lower control limits were set to 3 sigmas. Standard SPC charting rules for determining special cause were used as evidence of improvement.[16] Continuous variables were analyzed with either an Analysis of variance or Kruskal-Wallis test depending on the normality of the distribution. Categorial variables were analyzed with Chi-square and Fisher’s Exact tests. SPSS version 22 (IBM SPSS, Armonk, NY) was used for statistical analysis.

Ethical Considerations

The hospital Quality Improvement Review Committee determined that this project was not human subjects’ research and did not require Institutional Review Board review.

RESULTS

There were 76 mothers and 87 infants in the baseline period (June 2019–April 2020); 72 mothers and 77 infants in the education/documentation period (May 2020–April 2021), and 47 mothers who agreed to participate out of 48 mothers who qualified and 51 infants in the BLOSSoM period (May 2021–December 2021). Race/ethnicity, primary language, insurance, and maternal age did not differ. More mothers in the BLOSSoM period met exclusion criteria (11 of 62, 18%) as compared to the baseline period (4 of 91, 4%) and education/documentation period (8 of 87, 9%) (P = 0.03). Still, there was no difference in reason for exclusion between the periods. The majority of exclusions were due to substance use (Table 1). There was no difference in the infants’ GA, birth weight, mode of delivery, LOS, or disposition. More male infants were in the education/documentation period (P = 0.02), and fewer singleton deliveries were in the baseline period (P = 0.03) (Table 2). Although not statistically significant, there was an increase in the percentage of Spanish-speaking and Black mothers who provided MOM at discharge/transfer in the BLOSSoM period (Table 3).

Table 1.

Maternal Characteristics

	Baseline: June 2019–April 2020 (n = 76)	Education and Documentation Period: May 2020–April 2021 (n = 72)	BLOSSoM Period: May 2021–December 2021 (n = 48)	P
Race/ethnicity				0.17
Hispanic (%)	42 (56)	35 (49)	30 (62)
White (%)	16 (21)	24 (33)	7 (15)
African American (%)	13 (17)	9 (13)	9 (19)
Asian (%)	4 (5)	1 (1)	0
Unknown (%)	1 (1)	3 (4)	2 (4)
Primary language				0.24
English (%)	54 (71)	58 (81)	29 (61)
Spanish (%)	18 (24)	13 (18)	15 (31)
Other (%)	4 (5)	1 (1)	4 (8)
Insurance				0.24
Medicaid (%)	61 (80)	53 (74)	38 (79)
Commercial (%)	6 (8)	14 (19)	4 (8)
Self-pay (%)	7 (9)	4 (6)	6 (13)
Financial assistance (%)	3 (3)	1 (1)	0
Excluded	n = 4	n = 8	n = 11	0.91
Substance use (%)	3 (75)	5 (62)	8 (73)
Declined to provide MOM (%)	1 (25)	1 (13)	2 (18)
Infant death within 24 hours (%)	0	2 (25)	1 (9)

Table 2.

Infant Characteristics

	Baseline: June 2019–April 2020 (n = 87)	Education and Documentation Period: May 2020–April 2021 (n = 77)	BLOSSoM Period: May 2021–December 2021 (n = 51)	P
Median GA in weeks (IQR)	31 (28,32)	31 (28, 33)	31 (28, 32)	0.84
Mean birthweight in grams (SD)	1456 (473)	1457 (494)	1474 (458)	0.97
Male gender (%)	52 (60)	32 (42)	32 (63)	0.02
Singleton delivery (%)	64 (74)	67 (87)	45 (88)	0.03
Vaginal delivery (%)	37 (43)	40 (52)	21 (41)	0.70
Median LOS in days (IQR)	33 (21, 52)	31 (21, 56)	33 (21, 51)	0.99
Disposition (%)				0.96
Home	77 (89)	70 (91)	46 (90)
Transferred	8 (9)	5 (6)	5 (10)
Expired	2 (2)	2 (3)	0

Table 3.

Maternal Characteristics of Mothers Providing MOM at Discharge/Transfer

	Baseline: June 2019–April 2020 (n = 46)	Education and Documentation Period: May 2020–April 2021 (n = 44)	BLOSSoM Period: May 2021–Dec 2021 (n = 40)	P
Race/ethnicity				0.80
Hispanic (%)	28 (61)	22 (50)	24 (60)
White (%)	8 (18)	13 (30)	6 (15)
African American (%)	7 (15)	6 (14)	8 (20)
Asian (%)	2 (4)	1 (2)	0
Unknown (%)	1 (2)	2 (4)	2 (5)
Primary language				0.62
English (%)	31 (68)	32 (73)	25 (63)
Spanish (%)	14 (20)	11 (25)	12 (30)
Other (%)	1 (2)	1 (2)	3 (7)
Insurance				0.25
Medicaid (%)	33 (72)	29 (66)	32 (80)
Commercial (%)	5 (11)	12 (27)	4 (10)
Self-pay (%)	6 (13)	2 (5)	4 (10)
Financial Assistance (%)	2 (4)	1 (2)	0

Outcome Measures

MOM Provision Percentage at NICU Discharge/Transfer (SPC Chart)

The mean monthly percentage of mothers providing any MOM at their infant’s NICU discharge/transfer in younger than 34 weeks of GA increased from a baseline mean of 61% to 81% after implementation of standardized education/documentation and BLOSSoM (Fig. 3). There were 8 consecutive points above the centerline during PDSA cycle 4. Updated mean and control limits were plotted in the SPC charts after the detection of special cause.

Fig. 3.

Percentage of mothers providing mother’s own milk at NICU discharge/transfer in infants <34 weeks of gestation by month.

Parental Satisfaction of Support Received Through BLOSSoM

Of the 47 BLOSSoM participants, 41 (87%, 95% confidence interval [CI], 74–95) completed an evaluation. The program was rated “most supportive” by 34 of 41 (83%, 95% CI, 68–93) mothers with a median score of 5 (interquartile range [IQR]: 5–5). The majority of mothers found the videos helpful (32 of 41, 78%, 95% CI, 62–89) followed by team checking in (22 of 41, 54%, 95% CI, 37–69) and 2-way texting with LCs (10 of 41, 24%, 95% CI, 12–40). Most mothers (39 of 41, 95%, 95% CI, 83–99) stated the number of texts received was just right, with only 2 of 41 (5%, 95% CI, 0–17) stating there were too many texts.

Process Measures

Delivery of Standardized Verbal Education and Video Link During Prenatal Consult (PDSA Cycles 2 and 3)

After standardization of lactation education (June 2020–September 2020), 21 of 21 (100%, 95% CI, 84–100) of mothers received verbal education, and 8 of 21 (38%, 95% CI, 18–62) received the educational video link. After documentation of education delivery was incorporated into note template (October 2020–December 2021), delivery of verbal education remained consistent (57 of 57, 100%, 95% CI, 94–100, P = 0.47) and receipt of video link increased to 43 of 57 (75%, 95% CI, 62–86) (P < 0.001).

Use of Interpreters for Non-English Speaking Mothers During Prenatal Consult (PDSA Cycle 3)

After preferred language and use of interpreter were added to the note template (October 2020–December 2021), 13 of 15 non-English speaking mothers (87%, 95% CI, 60–98) had documented interpreter use. However, 1 mother declined to use an interpreter.

Completion of Educational Video Viewing (PDSA Cycle 2–4)

During June 2020–September 2020 (standardized education), 15 of 29 mothers (52%, 95% CI, 33–71) completed watching the educational videos. After providers consistently provided video links from October 2020 to April 2021 (education documentation), 22 of 34 mothers (65%, 95% CI, 46–80) watched the videos. After BLOSSoM implementation, which included a video link in a text message (May 2021–December 2021), 45 of 48 mothers (94%, 95% CI, 83–99) viewed the video in its entirety. A higher percentage of mothers in the BLOSSoM period watched the videos as compared to standardized education (P < 0.0001) and education documentation periods (P < 0.001).

Initial Willingness to Participate in BLOSSoM (PDSA 4)

Only 1 of 48 mothers (2%, 95% CI, 0–11%) in the BLOSSoM period declined participation as she was not interested in breastfeeding support in English as her primary language was Italian.

Balancing Measure

Amount of DBM Used Per Hospital Day (mL/hd)

There was no increase in the amount of DBM used between the baseline period (18.8 mL/hd, IQR 1.5–56.9), education/documentation period (21.5 mL/hd, IQR 3.6–71.5), and BLOSSoM period (15.5 mL/hd, IQR 1.7–71.7) (P = 0.74).

DISCUSSION

By focusing on education standardization and providing consistent, accessible lactation support to NICU mothers, we achieved our SMART aim and improved MOM provision percentage at NICU discharge/transfer in mothers of infants younger than 34 weeks of GA from 61% to 81% over 19 months. Key lessons learned included the importance of a multidisciplinary approach and the success of incorporating short message service (SMS) to support and communicate with mothers during their infant’s hospitalization. Although a statewide collaborative did not see improvement in MOM provision,[17] other single-centers have similarly demonstrated MOM provision improvement using a “care bundle” where multiple evidence-based interventions were implemented together.[18,19]

We did not see a significant change in MOM provision after implementing standardized lactation education. Education increases MOM initiation and provision[20,21] and often is a key focus as it is easily accomplishable. We updated the EMR note template that functioned as a “check off” reminder for providers to deliver education as Kalluri et al[17] did. EMR functions are frequently used to improve quality and efficiency.[22] However, education alone did not lead to significant change, consistent with a large review stating no conclusive benefit from antenatal education.[23]

Our second focus was to provide ongoing, easily accessible lactation support. BLOSSoM reiterated initial lactation education, provided readily available LC support, and granted convenient access to supportive educational videos, which most likely explains the improvement in MOM provision. Improvement in the provision of MOM is associated with LCs present in the NICU.[24] Recently, cell phones or audio-visual calls for telelactation are increasingly recognized as a potential new modality to reach vulnerable populations in term dyads. In addition, SMS has emerged as promising in prolonging breastfeeding duration in term dyads, particularly for those of lower socioeconomic status.[25-28] BLOSSoM provided a platform for mothers unable to visit regularly and engage with NICU LCs. Moreover, the educational content was readily available on their phones as a text at any time, which explains the increase in video viewing from 52% to 94%. It is possible BLOSSoM allowed mothers to come to volume earlier and establish a larger milk supply early in their NICU stay. Despite the lapsed time and barriers mothers faced later, they could maintain some or any MOM for longer periods through the time of discharge/transfer.

BLOSSoM was well accepted by our NICU mothers, with the majority rating the program highly. Hispanic mothers, in particular, have high social media/SMS use rates and are more responsive to information received through these modalities.[29] Most of our mothers found the hyperlinks to videos and handouts helpful. 2-way texting is effective in increasing breastfeeding in term infants of low socioeconomic status,[26] but was only found to be helpful to a third of our mothers, likely due to the occasionally delayed responses. A statewide collaborative showed racial/ethnic disparities in MOM provision, especially a decline after the first month among Hispanic mothers.[30] Although the difference was not statistically significant, we saw an increase in MOM provision among Spanish-speaking mothers. By providing consistent outreach, education, videos from other Spanish-speaking mothers, and response to inquiries in a well-accepted communication method, SMS may aid in addressing racial/ethnic disparities.

The protective benefits of DBM are reduced compared to MOM due to the processing, prolonged storage, and lack of mother-specific elements such as prebiotics. However, DBM still has protective benefits compared to formula.[31] While the use of DBM in NICUs has increased significantly,[32] especially in NICUs aiming to achieve exclusive MOM feedings,[33] the cost of DBM remains expensive.[34] In our initiative, the increased attention to MOM provision did not lead to increased DBM use, which would have been a significant cost increase.

LIMITATIONS

This study had several limitations. First, it reflects a single center’s experience in a level III NICU with only 25% of infants younger than 34 weeks of GA. The small sample size can lead to unstable estimates of the outcome variable and limit its reliability. Trialing BLOSSoM in a NICU with a larger population of preterm infants is needed to determine whether this intervention is generalizable.

Second, mothers were required to have a phone with texting capability for the additional lactation support via SMS. In this study, all participants owned a cell phone aligning with prior reports that most Hispanic mothers own cell phones and use social media/SMS frequently.[29] Therefore, if other centers have higher rates of mothers not having personal devices with texting capability, BLOSSoM will not be applicable.

Third, it is possible that providing incentives may have influenced outcomes towards favorable participation and engagement.[35] However, 25% (11 of 44) of mothers who received gift cards did not use them. Also, the incentives were relatively small, and we had a mother decline participation due to incentives being too small. Therefore, we speculate that the impact of incentives on participation in BLOSSoM or MOM provision was minimal.

Lastly, BLOSSoM requires appropriate infrastructure and provider investment. Hence, BLOSSoM will not be feasible in those many NICUs that do not have LCs.[36] Furthermore, data collection after the implementation of BLOSSoM was only 8 months, so future studies should focus on long-term sustainability.

CONCLUDING SUMMARY

Using a multidisciplinary approach, we improved the monthly MOM provision percentage at discharge/transfer for mothers of infants younger than 34 weeks. In addition, SMS providing educational texts/media and 2-way communication was well accepted and valued by the majority.

ACKNOWLEDGMENTS

Assistance with the study: Dr. Margaret G. Parker at Boston Medical Center allowed the use of NeoQIC videos her team created for a component of this project.

DISCLOSURE

The authors have no financial interest to declare in relation to the content of this article.