ABSTRACT
Background and Aim
In 2018, a pasteurised donor human milk (PDHM) service was established within Australia's national blood service, Lifeblood, and subsequently expanded to a national service. Before 2018, five milk banks operated in Australia, but most infants admitted to Australian neonatal intensive care units (NICUs) still did not have access to PDHM. This study aimed to compare infant outcomes before and after Australian NICUs had access to an assured supply of PDHM from Lifeblood.
Methods
Cohort study utilising Australian and New Zealand Neonatal Network data for infants born < 32 weeks' gestation admitted to 15 NICUs across Australia over a 3‐year period (2018–2020; n = 4395 infants). The primary study factor was NICU access to an assured supply of PDHM from Lifeblood. Neonatal outcomes were compared in periods with “PDHM available” versus “PDHM not available”. The primary outcome was the proportion of infants born < 28 weeks' gestation who developed NEC. Secondary outcomes included duration of parenteral nutrition, retinopathy of prematurity, infection, death, length of stay, weight at discharge, and breast milk feeding at discharge.
Results
After adjusting for confounding risks, NEC relative risk was 38% lower for infants born < 28 weeks' (aRR 0.62, 95% CI 0.46–0.83) when Lifeblood PDHM was available. Weight at discharge (2828.8 vs. 2939.56 g, p = 0.001) was lower when PDHM was available. Breast milk feeding at discharge, infections and death were not associated with PDHM availability.
Conclusions
Reliable access to PDHM was associated with reduced risk of NEC for infants born < 28 weeks' gestation in Australian NICUs.
Brief Points
- What is already known on this topic
-
○When maternal breast milk is not available, international evidence shows feeding with pasteurised donor human milk (PDHM) instead of infant formula reduces the risk of necrotising enterocolitis for infants born very preterm (< 32 weeks gestation) or low birthweight (< 1500 g).
-
○The benefits and risks of PDHM may vary in different settings. As PDHM is a limited and relatively expensive resource, it is important that its value in the local Australian context is studied and clearly defined.
-
○
- What this paper adds
-
○In the Australian context, being admitted to a neonatal intensive care unit with access to a reliable source of PDHM is associated with reduced risk of necrotising enterocolitis and retinopathy of prematurity for infants born very preterm.
-
○Further work is needed to ensure that PDHM is always used as part of a strategy which emphasises maternal breastfeeding and lactation support.
-
○
1. Introduction
Mother's own milk (MOM) is the best source of nutrition for preterm infants. Many preterm infants in a neonatal intensive care units (NICU), however, are temporarily unable to be fed MOM and are given supplemental nutrition. Pasteurised donor human milk (PDHM) is recommended by the World Health Organisation [1] when sufficient MOM is not available.
Internationally, using PDHM instead of infant formula for infants born < 32 weeks' and/or < 1800 g when sufficient MOM is not available is associated with approximately 2‐fold reduction in necrotising enterocolitis (NEC) [2] and improved feed tolerance [3]. Access to PDHM has also been associated with improved health service outcomes, such as reduced length of stay [4, 5], primarily resulting from a reduction in NEC and its complications. However, reduction in NEC is not observed with PDHM use in all settings [6, 7] and its impact on late‐onset sepsis [2, 6], retinopathy of prematurity (ROP) [4], and adverse long term neurodevelopmental outcomes [8] remains uncertain.
The major risk associated with PDHM is slower early growth [2], although multi‐nutrient fortification may reduce this risk [9]. There is also a risk that access to PDHM may unintentionally adversely impact breastfeeding rates [10] if not provided in conjunction with strong support for maternal lactation.
Most infants admitted to Australian NICUs did not have access to PDHM before 2018. While five milk banks were operating [11], only 25% of high‐risk preterm infants had access to PDHM [4]. Australia's national blood service, Australian Red Cross Lifeblood (Lifeblood), began operating a centralised milk service in September 2018 to provide PDHM to preterm infants in NICUs who did not have access to another source. The service was expanded to a national service over several years. At the end of 2020, 11 NICUs had received a reliable supply of PDHM from Lifeblood, without any episode of interruption or supply shortage.
As relatively few infants had been able to receive PDHM in Australia until 2018, it is important to define the potential benefits and risks of PDHM use as it becomes widely available [12]. This study aimed to evaluate the impact of reliable PDHM supply in neonatal units caring for infants born very and extremely preterm, in terms of proportion of infants with NEC as well as other clinical outcomes and breastfeeding outcomes at hospital discharge. The primary outcome was the proportion of infants born < 28 weeks' gestation who developed NEC.
2. Methods
2.1. Study Design
A population‐based cohort study using the Australian and New Zealand Neonatal Network (ANZNN) database for 2018–2020. ANZNN prospectively collects data on all infants admitted to Australian NICUs born < 32 weeks and/or < 1500 g.
2.2. Setting
22 of 23 Australian NICUs gave permission for their ANZNN data to be used in for the purpose of this study. To accurately assess the impact of access to a sufficient, reliable supply of PDHM, we excluded seven NICUs that either already received a regular supply of PDHM from an existing milk bank prior to 2018 (n = 5) or did not report data for all study years (n = 2).
We included only NICUs (n = 15) that began to receive regular PDHM supply from Lifeblood during the study period (n = 11) or NICUs that did not have any access to a regular supply of PDHM up until the end of 2020 (n = 4).
The five NICUs with a regular supply of PDHM prior to 2018 were excluded as we had no capacity to objectively determine the reliability of PDHM supply from other milk banks during the study period. We were able to assess ‘reliability’ of supply from Lifeblood through routinely collected Lifeblood records: all PDHM orders placed by hospitals were delivered in full, on time over the study period. There were no unfilled or partially filled orders, which would indicate interrupted supply or supply shortages.
Across participating units, infants were eligible to receive PDHM if they were born < 32 weeks' gestation or weighing < 1500 g and did not have sufficient MOM, with access usually continued up to 34 weeks' corrected age.
2.3. Participants
Infants born < 32 weeks' gestation between 1 January 2018 and 31 December 2020 and admitted to an eligible Australian NICU (n = 15).
2.4. Data Sources
Data was acquired from the ANZNN database. The ANZNN collects standardised, audited data on outcomes for all high‐risk newborns admitted to Australian NICUs. Dates of NICU access to Lifeblood PDHM supply was determined by month and year of first PDHM delivery from Lifeblood internal records.
2.5. Variables
The primary outcome was the proportion of infants born < 28 weeks' gestation who developed NEC in a time‐period when NICUs were receiving Lifeblood PDHM, compared to when the NICUs were not receiving Lifeblood PDHM. The ANZNN defines NEC as (1) NEC diagnosis at surgery or post‐mortem, or (2) radiological diagnosis of NEC—a clinical history plus pneumatosis intestinalis, or portal vein gas, or a persistent dilated loop on serial X‐rays, or (3) clinical diagnosis of NEC—a clinical history plus abdominal wall cellulitis and palpable abdominal mass [13].
Secondary outcomes were the proportion of infants born < 32 weeks' gestation who developed NEC, infections, weight at discharge to home or at term, duration of parenteral nutrition, breast milk feeding at discharge, retinopathy of prematurity, death prior to discharge and length of NICU stay.
Other study variables included infant sex, year of admission, delivery mode, birth weight, antibiotic provided to mother, probiotics given to the infant, respiratory support at 36 weeks' postmenstrual age, receipt of donor milk, and hospital. Variables are defined in Table S1.
2.6. Statistical Analysis
Descriptive statistics were used to characterise variables for extremely preterm (< 28 weeks' gestation) and all preterm infants born < 32 weeks' gestation.
For the analysis, infants were categorised into two groups based on the availability of Lifeblood PDHM in the unit at the time that they were admitted, irrespective of whether the individual infant received PDHM. Time periods were categorised as “PDHM not available in NICU” or “PDHM available in NICU”. Across the 15 units in the analysis, 270 months were included in each of the “PDHM not available in NICU” and “PDHM available in NICU” periods.
We fitted a log‐binomial model under the generalised linear model (GENMOD) framework to examine the association between PDHM access and outcomes with binary responses, including NEC, infection, weight at discharge, breast milk feeding at discharge, retinopathy of prematurity, and death prior to discharge. Relative risk (RR) and 95% confidence intervals (CI) were calculated after adjusting for sex, year of admission, delivery type, birth weight, antibiotic provided to mother, and probiotics given to the infant. As continuous outcome variables (duration of parenteral nutrition and length of NICU stay) were skewed, nonparametric Mann–Whitney U tests were used to estimate the difference in two groups and calculate p‐values. Data analysis was performed using SAS version 9.4.
2.7. Sample Size
To estimate the required duration of the study we did a study size calculation prior to commencement. ANZNN 2015 registry data indicated that 1000 infants were born < 28 weeks in 2015 and that 8.2% developed NEC [14].
Sample size estimation was based on a staged roll out of PDHM across NICUs included in the cohort study. Assuming a medium effect size (50% reduction in NEC from 8.2% to 4.1% in extremely preterm infants), with assumption of Type‐I error of 0.05, and Type‐II error of 0.2 (power = 0.8), and an allocation ratio of 3‐1 (intervention to routine‐care) for a two‐tail z‐test, the study reaches desired power at 36 months.
2.8. Ethics Statement
Ethical approvals were granted by the South Eastern Sydney Local Health District (2019/ETH10608) and Lifeblood (Klein 12112020) Human Research Ethics Committees.
3. Results
From 2018 to 2020, there were 4395 infants born < 32 weeks in an eligible NICU. Demographic and clinical characteristics of infants admitted to these NICUs (n = 15) are shown in Table 1.
TABLE 1.
Demographic and clinical characteristics of infants admitted to neonatal intensive care units (NICUs) in Australia included in the analysis from 2018 to 2020 (n = 15).
| Variables | Extremely preterm infants < 28 weeks Lifeblood PDHM NOT AVAILABLE # (%) | Extremely preterm infants < 28 weeks Lifeblood PDHM AVAILABLE # (%) | All preterm infants < 32 weeks Lifeblood PDHM NOT AVAILABLE # (%) | All preterm infants < 32 weeks Lifeblood PDHM AVAILABLE # (%) | |
|---|---|---|---|---|---|
| n | 716 | 722 | 2213 | 2182 | |
| Sex | Male | 395 (55.2%) | 422 (58.4%) | 1220 (55.1%) | 1224 (56.1%) |
| Female | 321 (44.8%) | 300 (41.6%) | 992 (44.8%) | 958 (43.9%) | |
| Ambiguous | 0 (0%) | 0 (0%) | 1 (0%) | 0 (0%) | |
| Year of birth | 2018 | 452 (63.1%) | 46 (6.4%) | 1387 (62.7%) | 134 (6.1%) |
| 2019 | 167 (23.3%) | 338 (46.8%) | 494 (22.3%) | 981 (45.0%) | |
| 2020 | 97 (13.5%) | 338 (46.8%) | 332 (15.0%) | 1067 (48.9%) | |
| Delivery mode | Vaginal | 257 (35.9%) | 244 (33.8%) | 650 (29.4%) | 678 (31.1%) |
| Instrumental | 17 (2.4%) | 11 (1.5%) | 68 (3.1%) | 50 (2.3%) | |
| CS with labour | 182 (25.4%) | 165 (22.9%) | 525 (23.7%) | 430 (19.7%) | |
| CS without labour | 260 (36.3%) | 302 (41.8%) | 969 (43.8%) | 1024 (46.9%) | |
| Unknown | 0 (0%) | 0 (0%) | 1 (0%) | 0 (0%) | |
| Birth weight (mean, SD) | Extreme LBW < 1000 g | 553 (77.2%) | 543 (75.2%) | 697 (31.5%) | 679 (31.1%) |
| Very LBW 1000–1499 g | 161 (22.5%) | 177 (24.5%) | 921 (41.6%) | 887 (40.7%) | |
| LBW 1500–2499 g | 2 (0.3%) | 2 (0.3%) | 564 (26.8%) | 614 (28.1%) | |
| Normal ≥ 2500 g | 0 (0%) | 0 (0%) | 1 (0%) | 2 (0.1%) | |
| Maternal antibiotic | Yes | 402 (56.1%) | 507 (70.2%) | 1083 (48.9%) | 1414 (64.8%) |
| No | 314 (43.9%) | 215 (29.8%) | 1128 (51.0%) | 768 (35.2%) | |
| Unknown | 0 (0%) | 0 (0%) | 2 (0.1%) | 0 (0%) | |
| Probiotic | Yes | 661 (92.3%) | 675 (93.5%) | 1883 (85.1%) | 2047 (93.8%) |
| No | 55 (7.7%) | 47 (6.5%) | 330 (14.9%) | 135 (6.2%) | |
| Respiratory support | Yes | 381 (53.2%) | 397 (55.0%) | 633 (28.6%) | 675 (30.9%) |
| No | 333 (46.5%) | 321 (44.5%) | 1577 (71.3%) | 1502 (68.8%) | |
| Unknown | 2 (0.3%) | 4 (0.6%) | 3 (0.1%) | 5 (0.2%) | |
| Major surgery | Yes | 80 (11.2%) | 81 (11.2%) | 125 (5.6%) | 136 (6.2%) |
| No | 636 (88.8%) | 641 (88.8%) | 2088 (94.4%) | 2046 (93.8%) | |
| Early breast milk feeding | Yes | 662 (92.5%) | 679 (94.0%) | 2001 (90.4%) | 2042 (93.6%) |
| No | 54 (7.5%) | 43 (6.0%) | 212 (9.6%) | 140 (6.4%) | |
| Infant received donor milk | Yes | 73 (10.2%) | 439 (60.8%) | 181 (8.2%) | 1376 (63.1%) |
| No | 643 (89.8%) | 283 (39.2%) | 2032 (91.8%) | 806 (36.9%) | |
| Breast milk feeding at discharge | Yes | 430 (60.1%) | 390 (54.0%) | 1555 (70.3%) | 1437 (65.9%) |
| No | 286 (39.9%) | 332 (46.0%) | 658 (29.7%) | 745 (34.1%) | |
| Parenteral nutrition (days) | Median, (range) | 13 (0–365) | 17 (0–240) | 9 (0–365) | 9 (0–240) |
| Infection | Yes | 190 (26.5%) | 217 (30.1%) | 282 (12.7%) | 316 (14.5%) |
| No | 526 (73.5%) | 505 (69.9%) | 1931 (87.3%) | 1866 (85.5%) | |
| Late onset infection | Yes | 130 (18.2%) | 153 (21.2%) | 177 (8.0%) | 200 (9.2%) |
| No | 586 (81.8%) | 569 (78.8%) | 2036 (92.0%) | 1982 (90.8%) | |
| Age at discharge home (days) | Median, (range) | 96 (50–365) | 99 (45–415) | 61 (15–413) | 62 (19–416) |
| Death before discharge | Yes | 107 (14.9%) | 96 (13.3%) | 149 (6.7%) | 132 (6.0%) |
| No | 609 (85.1%) | 626 (86.7%) | 2064 (93.3%) | 2050 (94.0%) | |
| Necrotising enterocolitis (NEC) | Yes | 59 (8.2%) | 48 (6.6%) | 79 (3.6%) | 66 (3.0%) |
| No | 657 (91.8%) | 674 (93.4%) | 2134 (96.4%) | 2116 (97.0%) | |
| Retinopathy of prematurity (ROP) | Yes | 425 (59.4%) | 405 (56.1%) | 584 (26.4%) | 550 (25.2%) |
| No | 194 (27.1%) | 228 (31.6%) | 961 (43.4%) | 1082 (49.6%) | |
| Unknown | 97 (13.5%) | 89 (12.3%) | 668 (30.2%) | 550 (25.2%) |
Of these 4395 infants, 2182 (49.6%) were born during a period when the NICU had access to PDHM from Lifeblood (Figure S1).
In the time periods where NICUs had access to Lifeblood PDHM, 63.1% of admitted infants were recorded as receiving PDHM. Figure 1 shows the proportion of infants receiving PDHM in individual units ranged from 38% to 89%.
FIGURE 1.
Proportion of admitted infants born < 32 weeks' gestation receiving donor milk before and after receiving supply of PDHM from Lifeblood during the study period (2018–2020).
3.1. Infants Born < 28 Weeks' Gestation
Between 2018 and 2020, 1438 infants were born < 28 weeks' gestation, with 722 (50.2%) born when the NICU had access to Lifeblood PDHM. The 107 cases of NEC occurred in infants born < 28 weeks' gestation: NEC rate 7.4% overall.
NEC rates were significantly lower in infants born extremely preterm when PDHM was reliably available in the unit where they were admitted. The rate of NEC was 6.65% in the period when Lifeblood PDHM was available compared to 8.24% when the units did not receive Lifeblood PDHM. The rate of NEC was 38% lower when Lifeblood PDHM was available (aRR 0.62, 95% CI 0.46–0.83) (Table 2).
TABLE 2.
Association between availability of pasteurised donor human milk from Lifeblood and NEC in extremely preterm infants admitted to neonatal intensive care units (NICUs) in Australia from 2018 to 2020 (n = 15 NICUs).
| Variables | NEC number | NEC % | RR | Adjusted RR a | |
|---|---|---|---|---|---|
| Lifeblood PDHM available | Yes (after) | 48/722 | 6.65 | 0.80 (0.56–1.16) | 0.62 (0.46–0.83) |
| No (before) | 59/716 | 8.24 | Ref | Ref | |
| Sex | Male | 60/817 | 7.34 | Ref | Ref |
| Female | 47/621 | 7.57 | 1.03 (0.71–1.49) | 1.00 (0.71–1.39) | |
| Ambiguous | — | — | |||
| Year | 2018 | 32/498 | 6.43 | Ref | Ref |
| 2019 | 37/505 | 7.33 | 1.14 (0.72–1.80) | 1.51 (1.10–2.09) | |
| 2020 | 38/435 | 8.74 | 1.35 (0.86–2.13) | 1.93 (1.31–2.83) | |
| Delivery mode | Vaginal | 25/501 | 4.99 | Ref | Ref |
| Instrumental | 3/28 | 10.71 | 1.15 (0.69–6.68) | 2.35 (0.69–7.96) | |
| CS with labour | 34/347 | 9.80 | 1.96 (1.19–3.23) | 1.95 (1.22–3.13) | |
| CS without labour | 45/562 | 8.01 | 1.60 (0.99–2.57) | 1.52 (0.88–2.61) | |
| Unknown | — | ||||
| Birth weight | Extreme LBW | 94/1096 | 8.58 | Ref | Ref |
| Very | 13/338 | 3.85 | 0.45 (0.25–0.79) | 0.44 (0.25–0.79) | |
| LBW | 0/4 | 0 | — | ||
| Normal | — | ||||
| Maternal antibiotic | Yes | 65/909 | 7.15 | 0.90 (0.62–1.31) | 0.98 (0.71–1.36) |
| No | 42/529 | 7.94 | Ref | Ref | |
| Unknown | |||||
| Probiotic | Yes | 101/1336 | 7.56 | 1.28 (0.57–2.85) | 1.36 (0.64–2.91) |
| No | 6/102 | 5.88 | Ref | Ref | |
| Unknown | — | — |
Adjusted for sex, year, delivery mode, birth weight, maternal antibiotic, probiotics and hospital.
Weight at discharge to home or at term was lower when Lifeblood PDHM was available, compared to when it was not (2828.8 ± 492.6 g vs. 2939.56 ± 482.2 g, p = 0.001).
The median duration of parenteral nutrition was longer when Lifeblood PDHM was available (17 vs. 13 days, p = 0.014).
Median length of stay was not different when Lifeblood PDHM was available (96 vs. 99 days, p = 0.750).
The rate of retinopathy of prematurity (ROP) was 9% lower when Lifeblood PDHM was available (aRR 0.91, 95% CI 0.83–0.99) (Table 3).
TABLE 3.
Other outcomes (aRR) when PDHM from Lifeblood is available, compared to the period it is not in cohort admitted to neonatal intensive care units (NICUs) in Australia from 2018 to 2020 (n = 15 NICUs).
| Outcome | Extremely preterm (aRR a ) | All preterm (aRR a ) |
|---|---|---|
| All infection | 1.09 (0.92–1.28) | 1.09 (0.93–1.27) |
| Late onset infection | 1.18 (0.94–1.50) | 1.12 (0.46–1.86) |
| Breast milk feeding at discharge | 1.06 (0.91–1.23) | 0.95 (0.89–1.04) |
| Died before discharge | 0.97 (0.78–1.20) | 1.05 (0.75–1.46) |
| Retinopathy of prematurity (ROP) | 0.91 (0.83–0.99) | 0.90 (0.83–0.99) |
Adjusted for sex, year, delivery mode, birth weight, maternal antibiotic, probiotics and hospital.
Breast milk feeding at discharge was lower when Lifeblood PDHM was available (54.0% vs. 60.1%); however, this difference was not statistically significant. Other secondary outcomes, including infections and death before discharge, were also not different.
3.2. All Infants Born < 32 Weeks' Gestation
Among all 4395 infants born < 32 weeks' gestation, the rate of NEC was also significantly lower when Lifeblood PDHM was available, compared to the period it was not available (3.02% v 3.57%) (aRR 0.63, 95% CI 0.46–0.89) (Table 4). Findings for secondary outcomes across the whole cohort were very similar to the < 28‐week group, with a significant reduction in ROP (aRR 0.90, 95% CI 0.83–0.99) (Table 3).
TABLE 4.
Association between availability of pasteurised donor human milk from Lifeblood and NEC in all preterm infants (< 32 weeks gestation) admitted to neonatal intensive care units (NICUs) in Australia from 2018 to 2020 (n = 15 NICUs).
| Variables | NEC number | NEC % | RR | Adjusted RR a | |
|---|---|---|---|---|---|
| Lifeblood PDHM available | Yes (after) | 66/2182 | 3.02 | 0.85 (0.61–1.17) | 0.63 (0.46–0.89) |
| No (before) | 79/2213 | 3.57 | Ref | Ref | |
| Sex | Male | 81/2444 | 3.31 | Ref | Ref |
| Female | 64/1950 | 3.28 | 0.99 (0.71–1.36) | 0.94 (0.72–1.22) | |
| Ambiguous | 0/1 | 0 | — | — | |
| Year | 2018 | 43/1521 | 2.83 | Ref | Ref |
| 2019 | 53/1475 | 3.59 | 1.27 (0.85–1.89) | 1.60 (1.06–2.40) | |
| 2020 | 49/1399 | 3.50 | 1.24 (0.83–1.85) | 1.73 (1.13–2.65) | |
| Delivery mode | Vaginal | 38/1328 | 2.86 | Ref | Ref |
| Instrumental | 3/118 | 2.54 | 0.88 (0.28–2.83) | 1.33 (0.50–3.54) | |
| CS with labour | 40/955 | 4.19 | 1.46 (0.94–2.26) | 1.5 (1.04–2.17) | |
| CS without labour | 64/1993 | 3.21 | 1.12 (0.75–1.66) | 1.07 (0.75–1.55) | |
| Unknown | 0/1 | 0 | — | ||
| Birth weight | Extreme LBW | 99/1376 | 7.19 | Ref | Ref |
| Very | 33/1808 | 1.83 | 0.25 (0.17–0.37) | 0.25 (0.17–0.38) | |
| LBW | 13/1208 | 1.08 | 0.15 (0.08–0.26) | 0.15 (0.08–0.29) | |
| Normal | 0/3 | 0 | — | — | |
| Maternal antibiotic | Yes | 87/2497 | 3.48 | 1.13 (0.82–1.58) | 1.2 (0.94–1.55) |
| No | 58/1896 | 3.06 | Ref | Ref | |
| Unknown | 0/2 | 0 | |||
| Probiotic | Yes | 133/3930 | 3.38 | 1.31 (0.93–2.34) | 0.98 (0.57–1.66) |
| No | 12/465 | 2.58 | Ref | Ref | |
| Unknown | — | — |
Adjusted for sex, year, delivery mode, birth weight, maternal antibiotic, probiotics and hospital.
4. Discussion
In preterm infants, NEC is a debilitating illness with a high risk of mortality and morbidity. Interventions which reduce NEC are very important, particularly as Australian units are admitting more infants born ≤ 24 weeks' gestation [15], who are at highest risk.
During the period when NICUs had access to a sufficient supply of PDHM from Lifeblood, NEC rates were lower in infants born extremely preterm (< 28 weeks' gestation) and all infants born < 32 weeks' gestation. This is consistent with the international evidence [7, 16, 17, 18, 19], though effect size in this cohort study (aRR 0.67) was not as large as reported in a recent Cochrane review (2‐fold reduction in NEC in infants < 32 weeks receiving PDHM instead of infant formula) [2].
The smaller effect size observed in this study, compared to published randomised controlled trials, may be explained by the fact that several NICUs later supplied by Lifeblood had already been accessing small quantities of donor milk from other sources. Notably, the ANZNN definition of ‘donor milk receipt’ does not exclude peer‐shared human milk, and units may have used different definitions of donor milk when reporting to the database. In the ‘pre‐intervention’ period, between 0% and 23% of infants in these NICUs received donor milk. PDHM use increased after Lifeblood supply became available but this study may encompass a transition or adoption period where PDHM was not yet consistently offered to all eligible infants or consented to by families when sufficient MOM was unavailable. An average of 60% of infants across the 11 NICUs received donor milk, which is less than expected. A US survey reported that in more than half of participating NICUs, ≥ 80% of very low birthweight infants received PDHM in hospital [20]. Previous research has shown breast milk feeding rates differ for extremely preterm infants among Australian and New Zealand NICUs, with some units achieving high rates of MOM feeding but Indigenous and other minority groups at risk for lower rates of receiving MOM [21]. Moreover, even when donor milk is available families may encounter barriers to accepting it for their child due to lack of their own or their healthcare provider's knowledge of PDHM use, concerns about the safety or quality of PDHM, or cultural, religious or social concerns about donated milk or donor identity [22]. It is unknown if lower use in this study is due to higher availability of maternal milk and/or differences in patient population, parental consent rates, or other factors.
Preterm nutrition practises and other NEC prevention practises also vary considerably among and within Australian NICUs, which may also contribute to the results observed in this study. Surveys have shown Australian NICUs significantly differ in their protocols for start and advancement of feeds, use of human milk fortifiers and other supplements, and use and strains of probiotics [23, 24, 25]. Standardised, evidence‐based nutrition protocols have been shown to sustainably improve infant outcomes, including for NEC [26]. Therefore, while access to a sufficient source of PDHM is a critical foundational step, further work to develop and implement standardised nutrition care is also needed to further reduce NEC rates [23].
Weight at discharge was lower for infants born < 28 weeks' gestation when Lifeblood PDHM was available. Reduced or delayed growth is considered the key risk associated with the use of PDHM [5], although human milk may promote better quality growth (in terms of body composition) than infant formula [27]. PDHM typically has lower protein and energy content than preterm MOM or infant formula. A recent study of Lifeblood PDHM found mean protein content was lower than would be expected in preterm MOM or formula [28]. Slower growth may be offset by routine fortification [29], which is now standard in many Australian NICUs. Fortifier use varies considerably among units in Australia, and lower weight at discharge in extremely, but not all, preterm infants may reflect this.
Other secondary outcomes were also impacted by PDHM availability. Risk of ROP was slightly lower when Lifeblood PDHM was available. This is consistent with evidence that a human milk diet reduces ROP [30] and has been reported in some studies of PDHM, although it has not been a consistent finding across studies [4, 31]. In this study, the association between human milk and ROP may be confounded by the lack of detailed data available about the quantity of MOM received by infants who were also fed PDHM. Further research is needed to confirm this finding.
Extremely preterm infants had a 4 day longer median duration of parenteral nutrition after Lifeblood PDHM became available. This is unexpected, as PDHM is generally associated with improved feed tolerance and randomised controlled trials have found markedly shorter [3] or no difference [6] in duration of parenteral nutrition in PDHM compared to formula supplemented groups.
It is important to ensure that PDHM availability does not adversely impact maternal breastfeeding. Internationally, evidence around the impact of PDHM on MOM feeding is mixed [10, 19, 32]. In this study, rates of breast milk feeding at discharge were lower after PDHM was reliably available, but this difference was not statistically significant when the unit was treated as a random effect to account for variation in feeding practises and maternal lactation support between units. Infants only receive PDHM when insufficient MOM is available, and it is important that PDHM availability does not reduce staff support for maternal lactation. PDHM should always be provided alongside lactation support and clear messaging that PDHM is not a replacement for MOM [33]. Importantly, lactation support in NICUs was also impacted by the COVID‐19 pandemic in 2020. Further research should explore variation in Australian unit practises around PDHM use that might support or unintentionally discourage maternal breastfeeding. For example, one NICU in South Australia reported that average breast milk feeding at discharge increased from 63% to 77% after PDHM became available [34].
This is the first study that investigates the benefits and risks of access to a reliable source of PDHM for infants born < 32 weeks' gestation in Australia. This study is strengthened by using data routinely collected by the ANZNN, which ensures its completeness and validity.
4.1. Limitations
This study has limitations due to the nature of the dataset. ANZNN do not collect data on infant formula or fortifier receipt, so we cannot determine overall use of supplemental nutrition or exposure to cow's milk products. Moreover, donor milk use is recorded as a binary variable (yes/no). Information about whether donor milk was offered, but not consented to by the family, is not recorded. Therefore, we cannot determine the proportion of the infant's diet that was PDHM or formula compared to MOM. The amount of supplementation an individual infant requires varies considerably, and the proportion of MOM may be a predictor of infant outcomes [35]. Finally, outcomes in this observational study may have been affected by practise changes during the COVID‐19 pandemic in 2020.
5. Conclusions
Reliable access to PDHM was associated with a decreased rate of NEC and ROP for infants born < 28 weeks' gestation in Australian NICUs. Further work is needed to ensure that PDHM is always used as part of a strategy which emphasises maternal breastfeeding and lactation support.
Author Contributions
Laura D. Klein, Kei Lui, Abrar A. Chughtai, Ian Callander and Vanessa Clifford designed the research, with advice from all authors. Data collected and provided by the Australian and New Zealand Neonatal Network. Abrar A. Chughtai analysed the data. Laura D. Klein, Kei Lui, Abrar A. Chughtai and Vanessa Clifford drafted the manuscript. All authors have read and approved the final manuscript.
Funding
Funding for this study was provided by the Baxter International Foundation and the Australian Department of Health (Grant #4‐EML85BJ). Funders were not involved in the design or publication of this study. Australian governments fund Australian Red Cross Lifeblood to provide blood, blood products and services to the Australian community.
Disclosure
Laura D. Klein, Christine Sulfaro, Richard Brown, David O. Irving, Veronica Hoad, Joanne Pink, and Vanessa Clifford are employees of Australian Red Cross Lifeblood, which supplies pasteurised donor human milk to hospitals in Australia. Lifeblood is a not‐for‐profit organisation that only accepts donations from voluntary, non‐remunerated donors. Robert Guaran is Chair of the Pasteurised Donor Human Milk Clinical Advisory Committee, as a component of his role as a Senior Clinical Advisor‐Neonatology in the NSW Ministry of Health. Robert Guaran is also an unpaid member of the Miracle Babies Foundation Board and Lifeblood Milk Clinical Advisory Board. In the past 3 years, Andrew J. McPhee has received research grant support from Fonterra and travel and conference support from Clover Corp to present a paper at a sponsored session. Ulrike Brandenberg is an unpaid member of the Lifeblood Milk Clinical Advisory Board.
Conflicts of Interest
The authors declare no conflicts of interest.
Supporting information
Table S1: Study variable definitions.
Figure S1: Lifeblood PDHM availability in 15 NICUs during the study period.
Acknowledgements
The authors thank Stephen Wright for his assistance in preparing the power calculations for the study protocol, and the families who have generously donated their excess breast milk to Lifeblood. Thanks to all Advisory Council Members of the ANZNN (*denotes ANZNN Executive). Australia: Scott Morris (Flinders Medical Centre, SA), Peter Schmidt, Manbir Chauhan* (Gold Coast University Hospital, QLD), Larissa Korostenski (John Hunter Children's Hospital, NSW), Mary Sharp, Steven Resnick, Rebecca Thomas, Tobias Strunk* (King Edward Memorial and Perth Children's Hospitals, WA), Jacqueline Stack (Liverpool Hospital, NSW), Pita Birch, Tori Oliver* (Mater Mother's Hospital, QLD), Dan Casalaz, Jim Holberton* (Mercy Hospital for Women, VIC), Alice Stewart, Rod Hunt*, Kenneth Tan* (Monash Medical Centre, VIC), Lucy Cooke (Neonatal Retrieval Emergency Service Southern Queensland, QLD), Lyn Downe (Nepean Hospital, NSW), Jonathan Davis (Newborn Emergency Transport Service, WA), Michael Stewart (Paediatric Infant Perinatal Emergency Retrieval, VIC), Andrew Berry (Newborn & paediatric Emergency Transport Service, NSW), Leah Hickey (Royal Children's Hospital, VIC), Mantho Kgosiemang (Royal Darwin Hospital, NT), Tony De Paoli, Naomi Spotswood* (Royal Hobart Hospital, TAS), Srinivas Bolisetty, Kei Lui* (Royal Hospital for Women, NSW), Eveline Staub (Royal North Shore Hospital, NSW), Mark Greenhalgh, (Royal Prince Alfred Hospital, NSW), Pieter Koorts (Royal Brisbane and Women's Hospital, QLD), Sue Jacobs (Royal Women's Hospital, VIC), Amy Keir* (SAAS MedSTAR Kids, SA), Clare Collins (Sunshine Hospital, VIC), Andrew Numa (Sydney Children's Hospital, NSW), Hazel Carlisle (The Canberra Hospital, ACT), Nadia Badawi, Himanshu Popat (The Children's Hospital at Westmead, NSW), Gary Alcock (Townsville University Hospital, QLD), Melissa Luig* (Westmead Hospital, NSW), Alison Kent (Women's & Children's Hospital, SA). New Zealand: Bronwyn Dixon (Christchurch Women's Hospital), Brian Darlow (Christchurch School of Medicine), Peter Fowlie, Andrew Kelly (Dunedin Hospital), Guy Bloomfield (Middlemore Hospital), Mariam Buksh, Malcolm Battin* (Auckland City Hospital), Jutta van den Boom (Waikato Hospital), Helen Miller, Angelica Allermo‐Fletcher, Claire Jacobs* (Wellington Women's Hospital). Other: Victor Samuel Rajadurai* (KK Women's and Children's Hospital, Singapore), Simon Lam (Prince of Wales Hospital, Hong Kong). We also wish to acknowledge ANZNN Executive that are not members of hospitals contributing data: Georgina Chambers* (National Perinatal Epidemiology and Statistics Unit, University of New South Wales); David Barker* (Whangarei Hospital, NZ), Anjali Dhawan* (Blacktown Hospital, NSW), Natalie Merida* (consumer), Denise Harrison* (ACNN). We also thank the members of the Australian and New Zealand Neonatal Network Scientific Research Review Committee: Kenneth Tan, Nicola Austin, Barbara Bajuk, Luke Jardine, Michael Meyer, Tim Schindler, Michael Stark, Tobias Strunk, Javeed Travadi. Open access publishing facilitated by The University of Melbourne, as part of the Wiley ‐ The University of Melbourne agreement via the Council of Australasian University Librarians.
Contributor Information
Laura D. Klein, Email: lklein@redcrossblood.org.au.
the Australian and New Zealand Neonatal Network:
Scott Morris, Peter Schmidt, Manbir Chauhan, Larissa Korostenski, Mary Sharp, Steven Resnick, Rebecca Thomas, Tobias Strunk, Jacqueline Stack, Pita Birch, Tori Oliver, Dan Casalaz, Jim Holberton, Alice Stewart, Rod Hunt, Kenneth Tan, Lucy Cooke, Lyn Downe, Jonathan Davis, Michael Stewart, Andrew Berry, Leah Hickey, Mantho Kgosiemang, Tony De Paoli, Naomi Spotswood, Srinivas Bolisetty, Kei Lui, Eveline Staub, Mark Greenhalgh, Pieter Koorts, Sue Jacobs, Amy Keir, Clare Collins, Andrew Numa, Hazel Carlisle, Nadia Badawi, Himanshu Popat, Gary Alcock, Melissa Luig, Alison Kent, Bronwyn Dixon, Brian Darlow, Peter Fowlie, Andrew Kelly, Guy Bloomfield, Mariam Buksh, Malcolm Battin, Jutta van den Boom, Helen Miller, Angelica Allermo‐Fletcher, Claire Jacobs, Victor Samuel Rajadurai, Simon Lam, Georgina Chambers, David Barker, Anjali Dhawan, Natalie Merida, Denise Harrison, Kenneth Tan, Nicola Austin, Barbara Bajuk, Luke Jardine, Michael Meyer, Tim Schindler, Michael Stark, Tobias Strunk, and Javeed Travadi
Data Availability Statement
Data used in this study are not publicly available but may be obtained from the Australian and New Zealand Neonatal upon request (https://www.anznn.net/research/researchrequest).
References
- 1. World Health Organization , WHO Recommendations for the Care of the Preterm or Low Birthweight Infant (World Health Organization, 2022). [Google Scholar]
- 2. Quigley M., Embleton N. D., Meader N., and McGuire W., “Donor Human Milk for Preventing Necrotising Enterocolitis in Very Preterm or Very Low‐Birthweight Infants,” Cochrane Database of Systematic Reviews 9 (2024): CD002971. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Cristofalo E. A., Schanler R. J., Blanco C. L., et al., “Randomized Trial of Exclusive Human Milk Versus Preterm Formula Diets in Extremely Premature Infants,” Journal of Pediatrics 163 (2013): 1592–1595. [DOI] [PubMed] [Google Scholar]
- 4. Miller J., Tonkin E., Damarell R. A., et al., “A Systematic Review and Meta‐Analysis of Human Milk Feeding and Morbidity in Very Low Birth Weight Infants,” Nutrients 10 (2018): 707. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Yang R., Chen D., Deng Q., and Xu X., “The Effect of Donor Human Milk on the Length of Hospital Stay in Very Low Birthweight Infants: A Systematic Review and Meta‐Analysis,” International Breastfeeding Journal 15 (2020): 89. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Corpeleijn W. E., de Waard M., Christmann V., et al., “Effect of Donor Milk on Severe Infections and Mortality in Very Low‐Birth‐Weight Infants: The Early Nutrition Study Randomized Clinical Trial,” JAMA Pediatrics 170 (2016): 654–661. [DOI] [PubMed] [Google Scholar]
- 7. Sisk P. M., Lambeth T. M., Rojas M. A., et al., “Necrotizing Enterocolitis and Growth in Preterm Infants Fed Predominantly Maternal Milk, Pasteurized Donor Milk, or Preterm Formula: A Retrospective Study,” American Journal of Perinatology 34 (2017): 676–683. [DOI] [PubMed] [Google Scholar]
- 8. O'Connor D. L., Gibbins S., Kiss A., et al., “Effect of Supplemental Donor Human Milk Compared With Preterm Formula on Neurodevelopment of Very Low‐Birth‐Weight Infants at 18 Months: A Randomized Clinical Trial,” Journal of the American Medical Association 316 (2016): 1897–1905. [DOI] [PubMed] [Google Scholar]
- 9. Brown J. V., Embleton N. D., Harding J. E., et al., “Multi‐Nutrient Fortification of Human Milk for Preterm Infants,” Cochrane Database of Systematic Reviews 6 (2016): Cd000343. [DOI] [PubMed] [Google Scholar]
- 10. Williams T., Nair H., Simpson J., and Embleton N., “Use of Donor Human Milk and Maternal Breastfeeding Rates: A Systematic Review,” Journal of Human Lactation 32 (2016): 212–220. [DOI] [PubMed] [Google Scholar]
- 11. Commonwealth Department of Health , Donor Human Milk Banking in Australia ‐ Issues and Background Paper (Commonwealth of Australia, 2014). [Google Scholar]
- 12. Matthews E., Collins C. T., Ellison V., et al., “Top 10 Research Priorities for Human Milk Banking and Use of Donor Human Milk: A Partnership Between Parents and Clinicians,” Journal of Paediatrics and Child Health 56 (2020): 770–776. [DOI] [PubMed] [Google Scholar]
- 13. ANZNN Network , ANZNN 2021 Data Dictionary (ANZNN, 2020). [Google Scholar]
- 14. Chow S. S. W., Le Marsney R., Creighton P., et al., Report of the Australian and New Zealand Neonatal Network 2015 (ANZNN, 2017). [Google Scholar]
- 15. Chow S. S. W., Creighton P., Chambers G. M., et al., Report of the Australian and New Zealand Neonatal Network 2020 (ANZNN, 2022). [Google Scholar]
- 16. Cohen M., Steffen E., Axelrod R., et al., “Availability of Donor Human Milk Fecreases the Incidence of Necrotizing Enterocolitis in VLBW Infants,” Advances in Neonatal Care 21 (2021): 341–348. [DOI] [PubMed] [Google Scholar]
- 17. Colaizy T. T., Poindexter B. B., McDonald S. A., et al., “Neurodevelopmental Outcomes of Extremely Preterm Infants Fed Donor Milk or Preterm Infant Formula: A Randomized Clinical Trial,” Journal of the American Medical Association 331 (2024): 582–591. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Feinberg M., Miller L., Engers B., et al., “Reduced Necrotizing Enterocolitis After an Initiative to Promote Breastfeeding and Early Human Milk Administration,” Pediatric Quality & Safety 2 (2017): e014. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Kantorowska A., Wei J. C., Cohen R. S., Lawrence R. A., Gould J. B., and Lee H. C., “Impact of Donor Milk Availability on Breast Milk Use and Necrotizing Enterocolitis Rates,” Pediatrics 137, no. 3 (2016): e20153123. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Boundy E. O., Anstey E. H., and Nelson J. M., “Donor Human Milk Use in Advanced Neonatal Care Units ‐ United States, 2020,” Morbidity and Mortality Weekly Report 71 (2022): 1037–1041. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Hilditch C., Collins C. T., Rumbold A., Gomersall J., Middleton P., and Keir A., “Breastmilk Use in Preterm Infants <29 weeks' Gestational Age in Australia, New Zealand and Singapore,” Journal of Paediatrics and Child Health 60, no. 11 (2024): 684–690. [DOI] [PubMed] [Google Scholar]
- 22. Mathias E. G., Patil D. S., Kolakemar A., et al., “Barriers and Facilitators for the Donation and Acceptance of Human Breast Milk: A Scoping Review,” Current Nutrition Reports 12, no. 4 (2023): 617–634. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. McLeod G., Farrent S., Gilroy M., et al., “Variation in Neonatal Nutrition Practice and Implications: A Survey of Australia and New Zealand Neonatal Units,” Frontiers in Nutrition 8 (2021): 642474. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Adams M., Bassler D., Darlow B. A., et al., “International Network for Evaluating Outcomes (iNeo) of Neonates. Preventive Strategies and Factors Associated With Surgically Treated Necrotising Enterocolitis in Extremely Preterm Infants: An International Unit Survey Linked With Retrospective Cohort Data Analysis,” BMJ Open 9, no. 10 (2019): e031086. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Klingenberg C., Aunsholt L., Van Den Akker C. H. P., et al., “International Survey on Enteral Nutrition, Supplementation and Probiotic Practices for Preterm Infants,” Archives of Disease in Childhood. Fetal and Neonatal Edition, ahead of print, February 12, 2026, 10.1136/archdischild-2025-329670. [DOI] [PubMed] [Google Scholar]
- 26. Patel A. L., Panagos P. G., and Silvestri J. M., “Reducing Incidence of Necrotizing Enterocolitis,” Clinics in Perinatology 44, no. 3 (2027): 683–700. [DOI] [PubMed] [Google Scholar]
- 27. Huang P., Zhou J., Yin Y., Jing W., Luo B., and Wang J., “Effects of Breast‐Feeding Compared With Formula‐Feeding on Preterm Infant Body Composition: A Systematic Review and Meta‐Analysis,” British Journal of Nutrition 116 (2016): 132–141. [DOI] [PubMed] [Google Scholar]
- 28. Walter L., Clifford V., Sulfaro C., et al., “Macronutrient Content of Pasteurised Donor Human Milk: Variability Between Batches From Single‐Donor Pools at an Australian Milk Bank,” Journal of Paediatrics and Child Health 59 (2023): 1281–1288. [DOI] [PubMed] [Google Scholar]
- 29. Ginovart G., Gich I., Gutierrez A., et al., “A Fortified Donor Milk Policy Is Associated With Improved In‐Hospital Head Growth and Weight Gain in Very Low‐Birth‐Weight Infants,” Advances in Neonatal Care 17, no. 4 (2017): 250–257. [DOI] [PubMed] [Google Scholar]
- 30. Bharwani S. K., Green B. F., Pezzullo J. C., Bharwani S. S., Bharwani S. S., and Dhanireddy R., “Systematic Review and Meta‐Analysis of Human Milk Intake and Retinopathy of Prematurity: A Significant Update,” Journal of Perinatology 36 (2016): 913–920. [DOI] [PubMed] [Google Scholar]
- 31. Lambley R. G., “Human Donor Milk Also Protects Against Severe Retinopathy of Prematurity,” BMJ 372 (2021): n25. [DOI] [PubMed] [Google Scholar]
- 32. Esquerra‐Zwiers A., Schoeny M. E., Engstrom J., et al., “The Interaction of Donor Human Milk Availability and Race/Ethnicity on Provision of Mother's Own Milk for Very Low Birth Weight Infants,” Breastfeeding Medicine 16 (2021): 46–53. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33. Brandstetter S., Mansen K., DeMarchis A., Nguyen Quyhn N., Engmann C., and Israel‐Ballard K., “A Decision Tree for Donor Human Milk: An Example Tool to Protect, Promote, and Support Breastfeeding,” Frontiers in Pediatrics 6 (2018): 324. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Keir A. K., Summers L., Gillis J., McPhee A. J., and Rumbold A., “Impact of Human Donor Milk on Maternal Milk Use at Discharge: Assessment Using Control Charts,” Archives of Disease in Childhood. Fetal and Neonatal Edition 107 (2022): 452–453. [DOI] [PubMed] [Google Scholar]
- 35. Lund A. M., Löfqvist C., Pivodic A., et al., “Unpasteurised Maternal Breast Milk Is Positively Associated With Growth Outcomes in Extremely Preterm Infants,” Acta Paediatrica 109 (2020): 1138–1147. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Table S1: Study variable definitions.
Figure S1: Lifeblood PDHM availability in 15 NICUs during the study period.
Data Availability Statement
Data used in this study are not publicly available but may be obtained from the Australian and New Zealand Neonatal upon request (https://www.anznn.net/research/researchrequest).