Abstract
Background:
Implementation of exclusive human milk (EHM) feeding defined as mother's own milk or donor human milk fortified with human milk–derived fortifiers can place an economic burden on institutions.
Methods:
Retrospective study of very low birth weight (VLBW) infants before and after the implementation of EHM feedings. Neonatal demographics and clinical outcomes including necrotizing enterocolitis, severe retinopathy of prematurity, bronchopulmonary dysplasia, late-onset sepsis, days on parenteral nutrition (PN), and length-of-stay were collected. The net cost to the institution was estimated using published data.
Results:
Sixty-four infants in the pre-EHM period and 57 infants in the post-EHM period were enrolled. Net product acquisition cost in 2020 and 2021 was $884,823. The EHM feeding guideline led to a reduction in the mean length of stay and mean days of PN use by 6.3 and 6.8 days per infant, respectively. This led to a cost saving of $1,813,444 ($31,815 per infant). No significant difference in incidence of short-term morbidities was observed. Combining the cost avoidance from clinical outcomes, the estimated financial impact over 2 years excluding insurance reimbursement was an estimated $ 913,840 ($16,032 per infant).
Conclusion:
Implementation of EHM-based feeding in VLBW infants is a cost-effective option for neonatal intensive care units that can result in reduced length of stay and days on PN without adversely impacting short-term morbidities.
Introduction
The American Academy of Pediatrics (AAP) has recommended that preterm infants should receive human milk appropriately fortified with proteins, minerals, and vitamins to ensure optimal nutrient intake and growth. In the setting where mother's own milk (MOM) is either unavailable or contraindicated, appropriately fortified donor human milk (DHM) is recommended.1,2 The short- and long-term benefits of an exclusive human milk (EHM) diet defined as MOM or DHM fortified with human milk–derived fortifier (HMDF) are well documented and include (1) shorter time to full feeds, (2) decreased incidence of necrotizing enterocolitis (NEC), (3) decreased incidence of sepsis, (4) reduction in neonatal intensive care unit (NICU) admission length of stay, (5) decreased rates of readmission after NICU discharge, and (6) improved neurodevelopmental outcomes.3–9
Human milk fortifiers (HMFs) historically have been derived from cow's milk and the addition of HMFs to human milk feeding has led to improvement in infant anthropometric measurements as well as clinical and neurodevelopmental outcomes. However, when compared with human milk feeding alone, the introduction of cow's milk–derived fortifiers (CMDFs) has been shown to increase rates of medical and surgical NEC and is associated with worse morbidity and mortality.3–5 Furthermore, the addition of CMDF has been demonstrated to negatively impact the anti-bacterial activity in MOM. 6 Over the past decade, with the introduction of HMDF, there has been a push for gradual transition away from CMDF especially for extremely preterm and very low birth weight (VLBW) infants who are inherently predisposed to worse clinical outcomes compared with late preterm or term newborns.
Neonatal admissions are associated with high economic and social burden with extreme preterm and VLBW infants disproportionately making up to 36% of these costs.6,10 Comorbidities such as NEC can increase costs by up to 15–30%. 7 Despite the protective benefits conferred by an EHM diet, DHM and HMDF can cost up to 15 times more than formula per milliliter making the cumulative cost a prohibitive factor when institutions consider transitioning to EHM diet. 11 In the era of ever-increasing health care costs, it is imperative that institutions allocate their resources in a cost-effective manner. This had led several institutions to evaluate the economics of implementing an EHM diet. Despite differences in unit protocols and feeding guidelines, studies have demonstrated an estimated cost-saving of $8,167–$31,514 per infant after the implementation of an EHM diet.7,11–14 In 2020, our unit successfully implemented the use of an EHM-derived feeding regimen. The objective of this study was to assess the clinical impact of implementing an EHM feeding regimen in VLBW infants, and to describe a cost–benefit analysis of it use.
Materials and Methods
Study design
This is a retrospective observational study before and after the implementation of EHM feeding defined as the use of MOM or DHM fortified with HMDF for all VLBW infants defined as birthweight (BW) <1,500 g at birth, born and admitted at Los Angeles General Medical Center Level III NICU. Before the availability of HMDF (January 2015 to December 2016), preterm infants were fed DHM and/or MOM fortified with CMDF. There was inconsistent availability of HMDF from 2017 to 2019; therefore, data were collected from January 2020 to December 2021, after implementation of EHM-based nutrition. Inclusion criteria included VLBW infants who were exclusively fed either MOM or DHM fortified with CMDF (preintervention) and HMDF (postintervention) until 34 weeks postmenstrual age (PMA). Exclusion criteria included infants who expired during their NICU admission or with incomplete outcome data. The study was approved by Institutional Review Board at Los Angeles General Medical Center.
Clinical variables
Data collected included neonatal demographics, BW and gestational age (GA), anthropometric measurements at birth and at discharge and their corresponding z scores based on Fenton 2013 growth calculator for preterm infants. Clinical outcomes collected included NEC ≥ Bell's stage 2, bronchopulmonary dysplasia (BPD) defined as the need for supplemental oxygen at ≥36 weeks PMA, severe retinopathy of prematurity (ROP) requiring intervention (laser photocoagulation, intravitreal administration of antivascular endothelial growth factor), late-onset sepsis (LOS), days of parenteral nutrition (PN) per infant, and length of stay.
Feeding guidelines
Before 2017, enteral feeding of all VLBW infants were initiated on MOM or DHM as soon as the infant displayed clinical stability as determined by the attending neonatologist on clinical service. Upon reaching a total enteral feeding of 80 mL/kg/day, MOM/DHM was fortified with CMDF. In 2017, the feeding guideline was updated to provide an EHM-based diet to VLBW infants. The updated guidelines introduced HMDF in place of CMDF for infants born with a BW of <1,250 g and was eventually extended to all VLBW infants. In addition, human-milk caloric fortifier (2 kcal/oz) was added when infants tolerated 120 mL/kg/day of enteral feeds and were off PN. At 34 weeks PMA, infants fed fortified MOM were transitioned to MOM with CMDF, whereas infants fed fortified DHM were transitioned to preterm formula diet.
To standardize preparation of nutrition for all infants admitted to the NICU, a dedicated milk laboratory was established and was staffed with trained personnel 12 hours/day, 365 days/year.
Economic analysis
To perform an economic analysis, the estimated average cost saving for each clinical outcome measure was obtained from published literature (Table 1).12,15–18
Estimated Cost for the Different Outcomes in Very Low Birth Weight Infants
Statistical analysis
Continuous variables were reported as mean ± standard deviation and categorical variables were reported as number and percent. Student t-test and chi-square tests or Fisher exact tests were used to compare continuous and categorical variables between the two cohorts as appropriate, respectively. Logistic regression analysis was performed to determine the relationship between medical NEC and use of EHM. Box–Cox transformation regression model was used to determine association of length of stay and days on PN with EHM. Both models were adjusted for GA. STATA 14 (College Station, TX) was used to perform data analysis. A value of p < 0.05 was considered statistically significant.
Results
Patient demographics
A total of 92 VLBW infants were admitted to the NICU in 2015–2016. After excluding infants who expired (six died within the first 48 hours and one died on postnatal day 13 because of hypoxic respiratory failure and was never fed enterally) and those with incomplete data (n = 21), a total of 64 infants were included in the pre-EHM analysis. A total of 69 VLBW infants were admitted to the NICU in 2020–2021. Of those, nine infants expired (six never received enteral nutrition, one died because of multiple congenital anomalies, one had volvulus complicated by abdominal compartment syndrome, and one infant died owing to NEC totalis at ∼35 weeks PMA during the transition off an EHM diet) and three had incomplete data resulting in post-EHM cohort of 57 infants (Fig. 1). The mean GA (28.3 vs. 29.2 weeks), BW (1,064.6 vs. 1,092.9 g) in the pre- and postintervention cohort were not significantly different as well as the BW z-scores and discharge weight and z scores in both cohorts (Table 2).

Flow diagram for study population.
Demographics of the Study Population
Values are given as mean ± standard deviation.
EHM, exclusive human milk; SD, standard deviation.
Clinical outcomes
After the implementation of HMDF, we observed a net reduction in cases of medical NEC (14% vs. 5%, p: 0.11) and an increase in surgical NEC cases (0% vs. 2%, p = 0.28), although neither outcomes was statistically significant. We found no difference in the incidence of BPD, LOS, and severe ROP. After the implementation of an EHM diet, the total average days of total parenteral nutrition (TPN) significantly decreased from 31.2 to 24.4 days per infant (p = 0.01). There was a reduction in mean length of stay from 74.3 to 68 days; however, it was not statistically significant (p = 0.15). Regression analysis after adjusting for GA indicated that the use of EHM did not result in a significant reduction of medical NEC (odds ratio 0.33, 95% confidence interval [CI]: 0.08–1.3, p = 0.12) or lower TPN days (Coefficient: −0.08, 95% CI: −0.33 to 0.16, p = 0.51) and length of stay (Coef: −0.03, 95% CI: −0.31 to 0.25, p = 0.86).
Economic analysis
Our product acquisition cost to acquire HMDF was estimated to be $313,784, and $499,639 in 2020 and 2021, respectively. During the same period, we spent $71,400 on DHM resulting in a net product acquisition of cost of $884,823 over a 2-year period. The cost to establish and maintain a new state of the art milk laboratory was estimated to be $207,690 including staff salaries to run the milk laboratory (Table 3). Although we found no statistical difference in medical NEC between the two cohorts, the overall reduction in incidence from nine to three cases led to a cost avoidance of $444,024. In addition, we observed a decrease of 6.8 days/infant in TPN days and 6.3 days/infant in total length of stay, equated to a cost saving of $556,594 and $1,256,850, respectively. Savings from reduced TPN days and length of stay was estimated to be $1,813,444 ($31,815 per infant). When combined with the cost avoidance from clinical outcomes, over the course of 2 years, we observed a net savings of $913,840 equating to an estimated saving of $16,032 per infant (Table 4).
Cost of Establishment and Maintenance of Nutrition Laboratory
HEPA, high efficiency particulate air.
Cost Avoidance Before and After the Introduction of Human Milk–Derived Fortifier
Values are given as mean ± standard deviation.
BPD, bronchopulmonary dysplasia; NEC, necrotizing enterocolitis; ROP, retinopathy of prematurity.
Discussion
Our study aimed to analyze the clinical and fiscal impact of an EHM diet in an academic level III NICU. Over the course of 2 years, we demonstrated a total savings of $913,840 for the institution, which is an estimated savings of $16,032 per infant. Our savings estimate is comparable with the per infant reduction observed in previous studies.11–14 A large proportion of our cost savings are generated from reduction in days of PN utilization and average length of stay. Over 2 years, reduction in PN use and length of stay led to a cost saving of $1,813,444, which equates to $31,815 per infant. This was achieved by a net reduction of PN utilization by 21% after the implementation of an EHM diet.
We hypothesize that this can partially be attributed to the clinical protections offered by an EHM diet, notably improved feeding tolerance as shown in previous studies, as well as a reduction in comorbidities that would have resulted in the stopping of enteral feeds and requiring infants to go back on PN after the initial TPN utilization period. 7 We also found that our average length of stay declined by ∼9% in the 2-year period postintervention. The mean GA in the post-EHM cohort was ∼1 week older, which may contribute to the decrease in length of stay; however, similar weights, and z-scores at time of birth leads us to believe that at least some of the reduction in LOS is likely owing to improved feeding tolerance and overall improved clinical course.
In line with AAP recommendations, the last decade has seen a push toward an EHM diet that has been successfully implemented in some academic NICUs across the nation. 13 The implementation of EHM in community NICUs has lagged as implementation is a complex process that requires a significant time and financial investment to acquire new products, update feeding guidelines, and hiring/training staff. The upfront investment required to complete the transition may limit hospital administrations from taking part if the project is not deemed to be financially sustainable. 13 Furthermore, the estimated cost of DHM can range from $27 to $590 per infant depending on the willingness or ability of the mother to provide breast milk, which can be challenging for already cash-sensitive institutions. 19
This particularly highlights the need for efficient resource utilization to achieve the highest “return on investment.” In our study, we noted that our DHM acquisition over 2 years was estimated at $71,400. Although the utilization of DHM is not exclusively limited to the patients in the analyzed cohort, it is nonetheless a significant cost that has the potential to be optimized by increasing the percentage of MOMs being utilized for infant nutrition. At present, it is estimated that only 13% of infants exclusively receive MOM and 72% of mothers are unable to produce enough milk necessary for an EHM diet for their newborns. 19 This limitation in MOM is owing to mother's ability and/or willingness to provide milk.
One of goals of the NICU should be to foster an environment that encourages and supports mothers to provide breast milk. This includes the use of community-oriented educational materials during pregnancy and must be supplemented with the necessary collection and storage infrastructure—NICU-specific lactation specialists, breast pumps, freezers, milk warmers—during the course of an infant's NICU admission. 20 Recognizing this need, our institution is currently in the process of hiring a lactation specialist specific for the NICU parents to enhance the utilization of MOM.
After the implementation an EHM diet, we noted a 56% reduction in total NEC. Our medical NEC incidence decreased from nine to three cases resulting in a cost-saving of $444,024 over 2 years. Our incidence of surgical NEC increased from 0 to 1 between 2015–2016 and 2020–2021. However, upon further review of this case, NEC was noted to occur before the initiation of enteral feeds; hence, it is unlikely that modifications in feeding guidelines contributed to the infant's clinical course. In contrast to previously published studies, the difference in incidence of medical and surgical NEC in our study did not receive statistical significance. 21 We believe that lack of statistical significance may be owing to low incidence of NEC preintervention as well as the smaller sample size, making our study not adequately powered to detect statistically significant changes.
We found no significant difference in rates of LOS or severe ROP but did note a 5% increase in incidence of BPD although this was not statistically significant. Previous studies assessing the impact of EHM on rates of BPD have shown conflicting results ranging from a net reduction of 9% to an increase in 2% after implementation of an EHM diet.7,11,22 During our analysis window, our unit underwent several updates in respiratory management techniques including but not limited to first-intention high-frequency jet ventilation and the introduction of LISA (Less-Invasive Surfactant Administration) to replace INSURE (INtubation–SURfactant–Extubation) for surfactant administration. Furthermore, the definition of BPD has undergone several updates—older studies have defined BPD as any preterm infant of <32 weeks' gestation requiring treatment with >21% oxygen for at least 28 days, whereas more recent studies have defined BPD as the need for supplemental oxygen at 36 weeks PMA. 23
A newer definition proposed in 2019 categorizes BPD based on respiratory support administered at 36 weeks PMA regardless of prior duration or current level of oxygen therapy. 24 With several updates in respiratory and feeding guidelines, it is difficult to delineate the true source of increase in BPD incidence and is something that will need to further be investigated in future studies.
Although most literature supports the utilization of HMDF over CMDF, it is worth noting that robust randomized controlled trials (RCTs) assessing the difference between HMDF and CMDF are lacking. Attempts at a systematic review have been challenging as some RCTs have supplemented formula for MOM in the control group limiting their inclusion. 25 A single 2018 RCT including 127 infants compared HMDF and CMDF in the absence of formula and found no significant difference in NEC, BPD, LOS, and ROP requiring treatment or death. Furthermore, the study found no difference in incidence of feeding interruption and days on TPN. 26 This highlights the urgent need for additional studies that compare the implementation of HMDF and CMDF with the exclusive use of MOM or DHM as the base feed to assess the true efficacy of the fortifier.
There are several limitations in our study. In our preintervention analytic cohort, 21 infants were excluded owing to incomplete data either because of being born outside of our institution and transferred in or being transferred out for service not provided at our institution. For our economic analysis, we estimated cost incurred with the use of national averages in published literature. Our study was performed in a level III academic NICU in California where cost of goods is traditionally higher than the rest of the country. A unit-specific economic analysis as performed by Johnson et al. would have resulted in a more accurate representation of cost expenditure and potential savings. 15 In addition, we did not account for opportunity costs by the patients' families and society at large.
Furthermore, institutional differences in clinical practice and feeding guidelines will have an impact on the incidence of short-term comorbidities as well as the length of stay and days on PN. Finally, our study involves 116 infants over a 4-year span, which is a relatively small patient cohort to perform an economic analysis.
Conclusion
Our study provides a real-world experience of implementing an EHM diet in a level III NICU and serves to add to existing literature regarding the economic benefits of EHM. Despite the higher price of HMDF and DHM coupled with the overhead costs of establishing a milk laboratory, the implementation of an EHM diet proved to be a cost-effective option for our unit. The cost savings accrued by the clinical benefits of decreased days on PN, length of stay, and lower numbers of medical NEC directly resulted in net economic savings in the long term for the institution. Although certain costs cannot be mitigated, investing in infrastructure to increase MOM utilization can decrease DHM needs allowing further cost avoidance and potentially improving short-term outcomes.
Footnotes
Acknowledgments
The authors acknowledge the neonatologists, neonatal fellows, NICU staff for their hard work and dedication in taking care of sick infants and the nutrition department for their assistance and contribution in writing the article at Los Angeles General Medical Center. The authors also thank Joy Koglin from Prolacta Bioscience who performed the initial economic analysis, but the final economic analysis and interpretation of the results was performed by M.T. and R.C. Prolacta Bioscience staff was not involved in writing the article nor contributed to the content of the article. None of the authors received any financial support from Prolacta Bioscience.
Authors' Contributions
M.T. was responsible for creating the database, data collection, and analysis, results interpretation and writing the article. M.R.C. contributed to database creation, data analysis, result interpretation, and writing the article. L.B. contributed to gathering the demographic data from the database and writing the article. R.R. contributed to study design, results interpretation, and revising the article. R.C. contributed to study design, data analysis, result interpretation, and writing the article.
Disclosure Statement
All the authors have no conflict of interest to disclose in relationship to the work described. Prolacta Bioscience participated in the preliminary cost-saving analysis. This was followed by a complete independent retrospective chart review and independent final statistical analysis by the authors in which Prolacta Bioscience was not involved. All the resulting data and conclusions in this article have been derived by the authors with no input from outside commercial interests. The study was approved by Institutional Review Board at LAC+USC Medical Center (IRB ID: HS-19-00450). The study was performed in accordance with the Declaration of Helsinki.
Funding Information
Joy Koglin from Prolacta Bioscience provided her time in performing the preliminary cost-saving analysis.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
