BACKGROUND

Although postnatal transfer patterns among high-risk (eg, extremely preterm or surgical) infants have been described, transfer patterns among lower-risk populations are unknown. The objective was to examine transfer frequency, indication, timing, and trajectory among very and moderate preterm infants.

METHODS

Observational study of the US Vermont Oxford Network all NICU admissions database from 2016 to 2021 of inborn infants 280/7 to 346/7 weeks. Infants’ first transfer was assessed by gestational age, age at transfer, reason for transfer, and transfer trajectory.

RESULTS

Across 467 hospitals, 294 229 infants were eligible, of whom 12 552 (4.3%) had an initial disposition of transfer. The proportion of infants transferred decreased with increasing gestational age (9.6% [n = 1415] at 28 weeks vs 2.4% [n = 2646] at 34 weeks) as did the median age at time of transfer (47 days [interquartile range 30–73] at 28 weeks vs 8 days [interquartile range 3–16] at 34 weeks). The median post menstrual age at transfer was 34 or 35 weeks across all gestational ages. The most common reason for transfer was growth or discharge planning (45.0%) followed by medical and diagnostic services (30.2%), though this varied by gestation. In this cohort, 42.7% of transfers were to a higher-level unit, 10.2% to a same-level unit, and 46.7% to a lower-level unit, with indication reflecting access to specific services.

CONCLUSIONS

Over 4% of very and moderate preterm infants are transferred. In this population, the median age of transfer is later and does not reflect immediate care needs after birth, but rather the provision of risk-appropriate care.

What’s Known on This Subject:

Postnatal transfer patterns among high-risk (eg, extremely preterm or surgical) infants have been described. However, transfer patterns among lower-risk infant populations such as very and moderate preterm infants, are not well understood.

What This Study Adds:

The median age of transfer among very and moderate preterm infants suggests transfers facilitate access to care needed later in the hospitalization. Over 4% of infants in this population undergo transfer, with indications highlighting the provision of risk-appropriate care.

Regionalization of perinatal care, which directs patients to hospitals with optimal capabilities for their condition, is associated with improved infant outcomes.1 4  Many studies of regionalized care focus on the birth hospital, finding that morbidity and mortality are lower when the highest risk infants, those who are extremely preterm (<28 weeks’ gestation) or very low birth weight (VLBW; ≤1500 g), are born at hospitals with higher-level NICUs.5 10  Although the optimal birth location for extremely preterm infants is clear, similar data regarding very and moderate preterm infants are less clear in part given the variation in NICU services and capabilities within each level of care. Another key component to regionalized care is neonatal transport.11  Consistent with the emphasis on birth hospital resources, studies of neonatal transport often focus on emergent transfers of extremely preterm infants or infants with severe congenital anomalies, who have urgent needs for subspecialty, and/or surgical care shortly after birth.12 17 

Although very (280/7–316/7 weeks’ gestation) and moderate (320/7–336/7 weeks’ gestation) preterm infants represent more than 25% of infants in NICUs, their care is understudied.18,19  Compared with term infants, very and moderate preterm infants are at increased risk for adverse health outcomes during and after the birth hospitalization.18,20  Evidence surrounding birth location, delivery room management, clinical care, transfer procedures, discharge practices, and long term outcomes for different preterm subgroups (eg, very, moderate, and late) is still emerging.21 24  Despite the personal effects of transfers on parents and families, such as separation and psychosocial stress, as well as health system function, such as transfer networks, hospital capacity, and bed availability,25 28  it remains unknown whether transfer patterns of less preterm infants mirror those of extremely preterm infants.

This study’s objective was to examine the transfer frequency, timing, indication, and trajectory (eg, to a higher-level, similar level, or lower-level unit) of initial transfers among infants, 280/7 to 346/7 weeks’ gestation, within the Vermont Oxford Network (VON) database of all NICU admissions. Understanding transfer patterns and NICU utilization may highlight opportunities to optimize risk-appropriate care delivery for this understudied and at-risk population.

VON is a nonprofit, voluntary worldwide community of practice dedicated to improving the quality, safety, and value of care for newborns. This observational cohort study uses prospectively collected data from the VON database of all NICU admissions29  that includes: (1) VLBW infants who are admitted anywhere in a hospital or die anywhere in a hospital within 28 days of birth; and (2) infants >1500 g admitted to a NICU (a location within a hospital where continuous positive airway pressure or intermittent mechanical ventilation can be provided, not including the delivery room or other infant stabilization location) within 28 days of birth, and (3) infants >1500 g who die anywhere in the hospital within 28 days of birth.30  Participating VON members submit data on eligible infants until death, discharge from the hospital, or transfer to another center. All submitted data undergo automated checks for quality and completeness and are deidentified.30  The Institutional Review Board at the University of Vermont determined that use of the VON database for this study was not human subjects research.

Infants eligible for inclusion in the study cohort were those who were inborn at US VON centers contributing to all NICU admissions database at 280/7 to 346/7 weeks’ gestation who survived to NICU admission from January 1, 2016 to December 31, 2021. Infants born 340/7 to 346/7 weeks’ gestation were included to serve as a reference population as these infants are often routinely admitted to the NICU. Infants still hospitalized at 1 year were excluded, as they still have the potential for transfer.

Transfers were identified using the birth hospital initial disposition. In the VON Manual of Operations, this variable reflects the first time the infant was discharged or transferred from the VON member hospital and includes discharged from the hospital, died, remained hospitalized, or transferred on or before their first birthday. Per the definition, infants who were transferred between units within 1 hospital are not considered transfers. To maintain a consistent definition for transfer, analyses examined the first transfer for each infant.

Other variables examined included gestational age at birth (by week), infant sex, small for gestational age (SGA; defined as a birth weight <10th percentile for gestational age), presence of congenital anomalies, maternal race and ethnicity (given prior data showing differences in transfer patterns by race31 ), age at time of transfer, in days and postmenstrual age (PMA) as well as PMA at ultimate discharge home, and the primary reason for transfer, including (1) growth and discharge planning defined as transfer to another hospital for continuing care in preparation for eventual discharge home; (2) medical or diagnostic services defined as transfer to another hospital to receive medical care or diagnostic tests that are not available at the current hospital and notes that if an infant is transferred for a diagnostic work-up that results in surgery, the reason for transfer remains medical/diagnostics services; (3) surgery defined as an infant transferred to another hospital for surgery even if the surgery was not actually performed after transfer; (4) chronic care defined as transfer to another institution for long term chronic care; (5) extracorporeal membrane oxygenation (ECMO) defined as transfer to another hospital for ECMO; and (6) other for which the transfer does not meet any of the criteria for reasons 1 through 5.30  ECMO was combined with other because only 27 infants who met eligibility criteria were transferred for ECMO.

We also examined transfers by trajectory, which were defined as transfers that were to a higher-level, similar level, or lower-level unit. Based on surveys completed by members, VON data categorizes NICUs into 4 types: Type A with restrictions on ventilation, Type A, Type B, and Type C. Type A NICUs with restrictions on ventilation transfer infants to another hospital for assisted ventilation (continuous positive airway pressure or mechanical ventilation) based on infant characteristics (ie, gestational age) or the duration of ventilation and are similar to the American Academy of Pediatrics (AAP) level II definition. Type A NICUs without ventilation restrictions do not perform 1 of 8 surgeries (omphalocele repair, ventriculoperitoneal shunt, tracheoesophageal fistula or esophageal atresia repair, bowel resection or reanastomosis, meningomyelocele repair, cardiac catheterization, cardiac surgery requiring bypass) and Type B NICUs perform at least 1 of the aforementioned 8 surgeries except cardiac surgery requiring bypass; both are similar to the AAP level III definition. Type C NICUs perform cardiac surgery requiring bypass and are similar to the AAP level IV definition. VON asks for the name of the hospital to which infants are transferred. If an infant was transferred to a hospital that was not a VON member, we used the AAP neonatology directory, Neonatology Solutions Web site, and individual hospital websites to identify the probable level of care. For consistency across VON members and nonmembers, transfer trajectory and NICU type were referred to and assigned by AAP level.

Characteristics of eligible infants who were transferred, discharged from the hospital, or died were reviewed. Transfer patterns assessed included median age at time of transfer, both in days and PMA, PMA at ultimate discharge, reason for transfer, and transfer trajectory. Given that clinical needs and thus the nature of transfers likely differ by gestational age, examination of transfer patterns were stratified by week of gestation at birth. We employed a generalized estimating equation logistic regression model with a log link to estimate risk ratios for transfer based on the covariates of gestational age (by week, referencing infants born at 34 weeks), infant sex, SGA, congenital anomalies, maternal race and ethnicity (referencing non-Hispanic white), PMA at initial disposition, and birth hospital NICU type (referencing level IV units). Reason for transfer was not included in the model as infants who were discharged from the hospital or died had no reason for transfer, which perfectly predicts the outcome of no transfer. Analyses were completed using SAS version 9.4.

Across 467 US VON member hospitals that participate in the all-admission database, there were 294 229 eligible infants born between 280/7 and 346/7 weeks’ gestation, of whom 12 552 (4.3%) had an initial disposition of transfer. Of those transferred, 1565 (12.5%) were transferred back to their initial hospital. When considering birth hospital NICU type, 17 052 (5.8%) of infants were born at a hospital with level II unit, 203 675 (69.6%) with a level III unit, and 72 074 (24.6%) with a level IV unit, of which 720 (4.2%), 8327 (4.1%), and 3354 (4.7%) were transferred respectively. Additional infant characteristics by initial disposition are shown in Table 1.

TABLE 1

Infant Characteristics by Initial Disposition

Discharged From the Hospital, N = 278 420Transferred, N = 12 552Died, N = 3257
Gestational age at birth, weeks (n = 294 229)    
 28 (n = 14 670) 12 678 (86.4) 1415 (9.6) 577 (3.9) 
 29 (n = 16 703) 14 965 (89.6) 1283 (7.7) 455 (2.7) 
 30 (n = 21 923) 20 027 (91.4) 1497 (6.8) 399 (1.8) 
 31 (n = 28 499) 26 472 (92.9) 1608 (5.6) 419 (1.5) 
 32 (n = 42 223) 39 760 (94.2) 2017 (4.8) 446 (1.1) 
 33 (n = 58 868) 56 355 (95.7) 2086 (3.5) 427 (0.7) 
 34 (n = 111 343) 108 163 (97.1) 2646 (2.4) 534 (0.5) 
Maternal race and ethnicity (n = 291 769)    
 Hispanic (n = 47 945) 45 602 (95.1) 1762 (3.7) 581 (1.2) 
 Non-Hispanic Asian (n = 12 611) 11 972 (94.9) 518 (4.1) 121 (1.0) 
 Non-Hispanic Black (n = 67 049) 63 507 (94.7) 2774 (4.1) 768 (1.2) 
 Non-Hispanic Native American (n = 3306) 3148 (95.2) 111 (3.4) 47 (1.4) 
 Non-Hispanic other (n = 6210) 5913 (95.2) 220 (3.5) 77 (2.4) 
 Non-Hispanic white (n = 154 648) 146 031 (94.4) 6987 (4.5) 1630 (1.1) 
Sex (n = 294 198)    
 Male (n = 156 064) 147 317 (52.9) 6912 (55.1) 1835 (56.5) 
 Female (n = 138 134) 131 088 (47.1) 5632 (44.9) 1414 (43.5) 
Birth wt, grams, median (IQR) (n = 294 229) 1927 (1562, 2254) 1660 (1280, 2050) 1500 (1135, 1990) 
Small for gestational age (n = 294 173)    
 Yes (n = 24 855) 22 460 (8.1) 1691 (13.5) 704 (21.8) 
 No (n = 269 318) 255 935 (91.9) 10 851 (86.5) 2532 (78.2) 
Antenatal steroid exposure (n = 293 808)    
 Yes (n = 243 166) 229 805 (82.7) 10 911 (87.1) 2450 (75.6) 
 No (n = 50 642) 48 222 (17.3) 1628 (12.9) 792 (24.4) 
Multiple birth (n = 294 215)    
 Yes (n = 78 876) 74 942 (26.9) 3385 (27.0) 549 (16.9) 
 No (n = 215 339) 203 466 (73.1) 6195 (73.0) 2708 (83.1) 
Congenital anomaly (n = 294 156)    
 Yes (n = 278 375) 6947 (2.5) 2272 (18.1) 1756 (54.0) 
 No (n = 271 428) 271 428 (97.5) 10 259 (81.9) 1494 (46.0) 
Birth hospital NICU type (n = 292 801)    
 Level II (n = 17 052) 16 289 (5.9) 720 (5.8) 43 (1.3) 
 Level III (n = 203 675) 193 223 (69.8) 8327 (67.2) 1682 (52.0) 
 Level IV (n = 72 074) 67 207 (24.3) 3354 (27.1) 1513 (46.7) 
PMA at initial disposition, weeks, median (IQR) (n = 294 223) 37 (36, 38) 35 (34, 37) 33 (31, 35) 
Discharged From the Hospital, N = 278 420Transferred, N = 12 552Died, N = 3257
Gestational age at birth, weeks (n = 294 229)    
 28 (n = 14 670) 12 678 (86.4) 1415 (9.6) 577 (3.9) 
 29 (n = 16 703) 14 965 (89.6) 1283 (7.7) 455 (2.7) 
 30 (n = 21 923) 20 027 (91.4) 1497 (6.8) 399 (1.8) 
 31 (n = 28 499) 26 472 (92.9) 1608 (5.6) 419 (1.5) 
 32 (n = 42 223) 39 760 (94.2) 2017 (4.8) 446 (1.1) 
 33 (n = 58 868) 56 355 (95.7) 2086 (3.5) 427 (0.7) 
 34 (n = 111 343) 108 163 (97.1) 2646 (2.4) 534 (0.5) 
Maternal race and ethnicity (n = 291 769)    
 Hispanic (n = 47 945) 45 602 (95.1) 1762 (3.7) 581 (1.2) 
 Non-Hispanic Asian (n = 12 611) 11 972 (94.9) 518 (4.1) 121 (1.0) 
 Non-Hispanic Black (n = 67 049) 63 507 (94.7) 2774 (4.1) 768 (1.2) 
 Non-Hispanic Native American (n = 3306) 3148 (95.2) 111 (3.4) 47 (1.4) 
 Non-Hispanic other (n = 6210) 5913 (95.2) 220 (3.5) 77 (2.4) 
 Non-Hispanic white (n = 154 648) 146 031 (94.4) 6987 (4.5) 1630 (1.1) 
Sex (n = 294 198)    
 Male (n = 156 064) 147 317 (52.9) 6912 (55.1) 1835 (56.5) 
 Female (n = 138 134) 131 088 (47.1) 5632 (44.9) 1414 (43.5) 
Birth wt, grams, median (IQR) (n = 294 229) 1927 (1562, 2254) 1660 (1280, 2050) 1500 (1135, 1990) 
Small for gestational age (n = 294 173)    
 Yes (n = 24 855) 22 460 (8.1) 1691 (13.5) 704 (21.8) 
 No (n = 269 318) 255 935 (91.9) 10 851 (86.5) 2532 (78.2) 
Antenatal steroid exposure (n = 293 808)    
 Yes (n = 243 166) 229 805 (82.7) 10 911 (87.1) 2450 (75.6) 
 No (n = 50 642) 48 222 (17.3) 1628 (12.9) 792 (24.4) 
Multiple birth (n = 294 215)    
 Yes (n = 78 876) 74 942 (26.9) 3385 (27.0) 549 (16.9) 
 No (n = 215 339) 203 466 (73.1) 6195 (73.0) 2708 (83.1) 
Congenital anomaly (n = 294 156)    
 Yes (n = 278 375) 6947 (2.5) 2272 (18.1) 1756 (54.0) 
 No (n = 271 428) 271 428 (97.5) 10 259 (81.9) 1494 (46.0) 
Birth hospital NICU type (n = 292 801)    
 Level II (n = 17 052) 16 289 (5.9) 720 (5.8) 43 (1.3) 
 Level III (n = 203 675) 193 223 (69.8) 8327 (67.2) 1682 (52.0) 
 Level IV (n = 72 074) 67 207 (24.3) 3354 (27.1) 1513 (46.7) 
PMA at initial disposition, weeks, median (IQR) (n = 294 223) 37 (36, 38) 35 (34, 37) 33 (31, 35) 

The proportion of infants transferred decreased with increasing gestational age (9.6% [n = 1415] at 28 weeks vs 2.4% [n = 2646] at 34 weeks) as did the median age at time of transfer (47 days [interquartile range (IQR) 30–73] at 28 weeks vs 8 days [IQR 3–16] at 34 weeks). The median PMA at transfer was 34 or 35 weeks across all gestational ages (Table 2). The median PMA upon discharge home among infants who were initially transferred ranged from 38 weeks for infants born at ≥31 weeks to 41 weeks for infants born at 28 weeks.

TABLE 2

Timing of Initial Transfer, Indication of Transfer, and Age at Ultimate Discharge Among Those Transferred (n = 12 552) by Gestational Age at Birth

Gestational Age at Birth
28293031323334
Days to initial transfer – median (IQR) 47 (30–73) 39 (21–61) 31 (14–53) 19 (9–37) 13 (7–30) 9 (5–20) 8 (3–16) 
PMA at initial transfer-weeks – median (IQR) 35 (33–39) 35 (32–38) 35 (32–38) 34 (33–37) 34 (33–37) 35 (34–36) 35 (35–37) 
Growth or discharge planning 35 (33–37) 34 (33–36) 34 (33–36) 34 (33–35) 34 (34–35) 35 (34–36) 36 (35–36) 
Medical or diagnostic services 35 (30–40) 35 (31–40) 35 (31–39) 35 (32–38) 35 (33–39) 35 (34–39) 35 (35–39) 
Surgery 36 (32–40) 37 (32–40) 37 (32–40) 36 (32–39) 36 (33–40) 35 (34–39) 35 (35–39) 
Chronic care 41 (34–50) 37 (33–42) 40 (33–50) 45 (40–49) 34 (34–39) 35 (34–47) 40 (35–45) 
Other (unspecified or ECMO) 35 (33–36) 35 (33–42) 35 (32–36) 33 (32–35) 34 (33–36) 34 (34–35) 35 (35–35) 
PMA at ultimate discharge weeks – median (IQR) 41 (38–45) 40 (38–44) 39 (37–43) 38 (37–42) 38 (36–41) 38 (36–41) 38 (37–42) 
Gestational Age at Birth
28293031323334
Days to initial transfer – median (IQR) 47 (30–73) 39 (21–61) 31 (14–53) 19 (9–37) 13 (7–30) 9 (5–20) 8 (3–16) 
PMA at initial transfer-weeks – median (IQR) 35 (33–39) 35 (32–38) 35 (32–38) 34 (33–37) 34 (33–37) 35 (34–36) 35 (35–37) 
Growth or discharge planning 35 (33–37) 34 (33–36) 34 (33–36) 34 (33–35) 34 (34–35) 35 (34–36) 36 (35–36) 
Medical or diagnostic services 35 (30–40) 35 (31–40) 35 (31–39) 35 (32–38) 35 (33–39) 35 (34–39) 35 (35–39) 
Surgery 36 (32–40) 37 (32–40) 37 (32–40) 36 (32–39) 36 (33–40) 35 (34–39) 35 (35–39) 
Chronic care 41 (34–50) 37 (33–42) 40 (33–50) 45 (40–49) 34 (34–39) 35 (34–47) 40 (35–45) 
Other (unspecified or ECMO) 35 (33–36) 35 (33–42) 35 (32–36) 33 (32–35) 34 (33–36) 34 (34–35) 35 (35–35) 
PMA at ultimate discharge weeks – median (IQR) 41 (38–45) 40 (38–44) 39 (37–43) 38 (37–42) 38 (36–41) 38 (36–41) 38 (37–42) 

The most common reason for transfer was growth or discharge planning (45.0%) followed by medical or diagnostic services (30.2%), though this varied by gestational age. The proportion of infants transferred for growth or discharge planning increased with increasing gestational age, accounting for 35.0% of transfers among infants born at 28 weeks and peaking with 50.8% of transfers among those born at 33 weeks (Fig 1). The median PMA at transfer for growth or discharge planning ranged from 34 to 36 weeks PMA and was 35 weeks PMA across all gestational age groups for infants transferred for medial or diagnostic services. Consistent with the indication, chronic care transfers were generally at an older PMA. (Table 2).

FIGURE 1

Title: Reason for transfer among those transferred (n = 12 552) by gestational age at birth.

FIGURE 1

Title: Reason for transfer among those transferred (n = 12 552) by gestational age at birth.

Close modal

Of the 12 552 infants transferred, 12 151 were transferred to places where level of care could be identified. In this cohort, 42.7% of transfers were to a higher-level unit, 10.2% to a same level unit, and 46.7% to a lower-level unit (Table 3). The majority (87.0%) of transfers to a higher-level unit were from level III to level IV units, whereas the majority (63.2%) of similar level transfers were between level III units. When examining the distribution of transfers to a lower-level unit, 32.8% were from level III to II units and 27.0% were from level IV to level II units (Table 3). Figure 2 depicts the reason for transfer by transfer trajectory. Transfers to higher-level units were most often for medical or diagnostic services (60.5%), whereas for transfers to lower-level units were for growth or discharge planning (85.6%), which is consistent with regionalized care and access to specific services (Fig 2).

TABLE 3

Distribution of Transfer Trajectory Among 12 151 Very and Moderate Preterm Infants

Transfer TrajectoryN%
To a higher-level 5188 42.7 
To a similar-level 1242 10.2 
To a lower-level 5721 46.7 
To a higher-level   
 II to III 109 2.1 
 II to IV 568 11.0 
 III to IV 4511 86.9 
To a similar-level   
 II to II 11 0.9 
 III to III 785 63.2 
 IV to IV 446 35.9 
To a lower-level   
 II to I 18 0.3 
 III to II 1876 32.8 
 III to I 982 17.2 
 IV to III 588 10.3 
 IV to II 1546 27.0 
 IV to I 711 12.4 
Transfer TrajectoryN%
To a higher-level 5188 42.7 
To a similar-level 1242 10.2 
To a lower-level 5721 46.7 
To a higher-level   
 II to III 109 2.1 
 II to IV 568 11.0 
 III to IV 4511 86.9 
To a similar-level   
 II to II 11 0.9 
 III to III 785 63.2 
 IV to IV 446 35.9 
To a lower-level   
 II to I 18 0.3 
 III to II 1876 32.8 
 III to I 982 17.2 
 IV to III 588 10.3 
 IV to II 1546 27.0 
 IV to I 711 12.4 
FIGURE 2

Reason for transfer by trajectory of transfer (n = 12 151).

FIGURE 2

Reason for transfer by trajectory of transfer (n = 12 151).

Close modal

In the model assessing risk of transfer (Supplemental Table 4), the strongest risk factor was congenital anomalies (adjusted risk ratio [aRR] 4.90, 95% confidence interval [CI] 3.64–6.60). Gestational age remained an important predictor after adjusting for other infant characteristics (28 weeks: aRR 3.86, 95% CI 3.15–4.73 vs 33 weeks: aRR 1.44, 95% CI 1.31–1.59; 34 weeks reference). Male infant sex and SGA, characteristics more common among infants who are transferred, remained associated in the model. Conversely, each additional day of PMA at initial disposition, which was highest among infants discharged from the hospital, was associated with a lower risk of transfer. In this population, the maternal race and ethnicities of Hispanic and non-Hispanic Black were associated with a lower risk of transfer (Hispanic aRR 0.77, 95% CI 0.64–0.94; non-Hispanic Black aRR 0.81, 95% CI 0.68–0.97; non-Hispanic white reference). There was no significant difference in transfer risk for the non-Hispanic groups of Asian, Native American, or other. The birth hospital NICU level of care was not associated with risk of transfer (level II: aRR 1.18, 95% CI 0.64–2.16; level III: 1.01, 95% CI 0.57–1.78; level IV reference).

In this US cohort, over 4% of infants born between 28 and 34 weeks’ gestation were transferred. Transfers in this population largely occurred later in the hospitalization, differing from higher-risk infants. Transfer reason and trajectory differed by gestational age with patterns reflecting access to specific services, which are likely influenced by unit transfer policies. After adjustment for infant characteristics, the birth hospital’s NICU level of care was not associated with an increased risk of transfer, which may reflect both appropriate prenatal assessments with respect to birth location (eg, prenatal referral based on gestational age) and the heterogenous care needs (eg, later access to both higher- and lower-level care) in this population.

In this cohort, transfer occurred at median age of nearly 7 weeks among infants born at 28 weeks and more than 1 week among infants born at 34 weeks, later than reported in more preterm populations.14  With nearly 70% of these infants born in a hospital with a level III unit, these data collectively suggest that many of these infants were born in hospitals with appropriately matched resources for stabilization and initial care. Although regionalization is a longstanding priority in neonatal care delivery,11  deregionalization remains an undercurrent with ongoing establishment of small, level III units with variable services.32 34  Although such deregionalization may significantly affect very high risk infant populations who benefit from birth and care in high volume, high level NICUs,8  there may be a differential effect on less preterm infants who may not need or benefit from being born in the highest level centers.23 

Transfers in this population present some unique challenges when assessing opportunities to improve care delivery for these infants and their families. Reasons for transfer are heterogenous. This is evident in the most common reason for transfer, growth or discharge planning. If or when an infant will meet growth or discharge planning transfer criteria is difficult to predict, and such criteria likely vary across hospitals. Furthermore, within this group of infants, some may be transferred to access more involved feeding support at a higher-level unit and others to establish feeding skills and weight gain at a lower-level unit. Similarly, transfer for medical or diagnostic services may result in a range of interventions and the ability to predict both the occurrence and timing of these services is difficult. Although predicting the timing of clinical changes will remain challenging, examination of transfer policies may shed light on when and why these infants are transferred as well as how these transfers and their timing affect infant outcomes and family experiences.

The plurality of infants in this cohort were transferred to a lower level of care, likely reflecting transfers for convalescent care. Among VLBW infants, convalescent care transfer rates were 4.5%.35  In this study, of the 4.3% of infants with an initial disposition of transfer, almost 36% of those (approximately 1.5% of the total cohort) were transferred to a lower level of care. This analysis focused on the initial transfer, thus back-transfers (defined as a second transfer; eg, an infant returning to their initial birth hospital after subspeciality evaluation or “failing” an initial transfer to a lower-level unit and returning to the higher-level unit) were not captured in this study.36,37  Given the different patient phenotypes and care patterns associated with back-transfers and multiple transfers, future consideration and dedicated studies in this population and on this topic may highlight opportunities to optimize care for these infants.

Although transfers are meant to advance care, they are not necessarily benign. Transfers can adversely affect infant physiology and are associated with morbidities in preterm infants up to 34 weeks’ gestation.38,39  Transfers are also associated with a decreased likelihood of breastfeeding at discharge, though these data do not account for transfer trajectory.40  Breastfeeding is a relevant issue for this population and a common measure of preterm infant NICU care quality.22,41  Additionally, transfers may be further from where a family lives and may limit family involvement in care and discharge planning. Considering adverse clinical events, the provision of family-centered care and discharge preparedness may be potential balancing measures to consider in future work evaluating transfers.

In these data, transfer patterns and risk differed by a variety of patient characteristics, including gestational age, indication for transfer, and maternal race and ethnicity, which is similar to previous studies of neonatal transport networks for VLBW infants.31  The variation in transport patterns for different populations suggests that the characteristics of the transport networks themselves may account for some of the observed differences between populations.28  Transfer policies may influence transport networks, and although state-level data surrounding transfer policies and reimbursement are available,42  knowledge of unit transfer policies are limited. Interestingly, in this cohort the median PMA of transfer was 34 to 35 weeks for very different indications (eg, growth and discharge planning versus medical and diagnostic services versus surgery), which may suggest that transfer timing is influenced by gestational-age based unit or hospital policies. Defining and understanding the influence of transport networks and policies on transfer patterns may highlight new opportunities to streamline care.

This study has limitations and strengths. This is an observational study of prospectively collected data and there are likely unmeasured variables and confounders that would enrich study findings. Although unavailable, additional details regarding the type of surgery or distance a family lives from a NICU would further contextualize transfers. Future studies integrating the clinical status before and after transfer may also add depth to the evaluation and understanding of transfers. This study focused on the initial transfer, which may limit study generalizability, however this also minimizes bias as infants with multiple transfers are likely a distinct subgroup. This study used the AAP levels of care (I–IV) for analysis, which are similar though not identical to the VON NICU types (A–C). Although slightly different, utilizing American Academy of Pediatrics (AAP) levels increases generalizability and improves interpretability of findings. This study leveraged a convenience sample in VON’s all NICU admission database. Although not a true population-based sample, the analytic cohort included 467 NICUs and nearly 295 000 infants. Given the number of infants and variety of units captured, these findings are likely generalizable beyond those included in the analytic cohort and likely reflect transfer patterns across the United States. These findings are relevant to many neonatal care settings since infants born between 28 and 34 weeks’ gestation may be admitted (or transferred to) a level II special care nursery or higher level III or IV NICU. Further, it is plausible that the phenotype of transfers seen in this study’s population may extend to some extremely (26–27) or late (35–36) preterm infants. Thus, implications of developing transfer policies to optimize the care of very and moderate preterm infants may have spillover effects on other infant populations. As pediatricians and neonatologists work to improve care delivery and outcomes for infants, recognizing the frequency, timing, and reason for transfers as well as potential downstream effects on infants and their families (eg, more convenient convalescent care versus subspecialty care a greater distance from home) are important considerations when designing and implementing transfer policies and delivering effective neonatal care.

In this cohort, over 4% of preterm infants were transferred, often multiple days to weeks after birth. These transfer patterns reflect risk-appropriate assessment of initial stabilization and care needs as well as regionalized care delivery through later transfers to access specific services. These data may inform future efforts to optimize the timing of transfers for patients and their families and future studies reflecting on the clinical status, transport networks, and transfer policies may highlight opportunities to streamline care for these infants.

We are indebted to our colleagues who submit data to the Vermont Oxford Network on behalf of infants and their families. The list of centers contributing data to this study are listed in Supplemental Table 5.

Dr Handley conceptualized and designed the study and drafted the initial manuscript; Dr Salazar contributed to the design of the study and interpretation of the data; Dr Kunz contributed to the design of the study and interpretation of the data; Dr Lorch conceptualized and designed the study; Dr Edwards conceptualized and designed the study and conducted the analyses; and all authors critically reviewed and revised the manuscript, approved the final manuscript as submitted, and agree to be accountable for all aspects of the work.

FUNDING: Drs Handley and Salazar are supported by the National Institutes of Health (K23HD109426 to SCH and T32HL098054 to EGS); Dr Kunz is supported by the Agency for Healthcare Research and Quality (K08HS025749); Dr Edwards receives salary support from Vermont Oxford Network; and Dr Lorch received no external funding. The NIH and ARHQ had no role in the design or conduct of the study.

CONFLICT OF INTEREST DISCLOSURE: The authors have indicated they have no conflicts of interest relevant to this article to disclose.

IQR

interquartile range

PMA

post menstrual age

VON

Vermont Oxford Network

1
Institute of Medicine; Committee on the Future of Emergency Care in the United States Health; Board on Health Care Services
.
Emergency Medical Services: At the Crossroads
.
National Academies Press
;
2006
2
Bode
MM
,
O’shea
TM
,
Metzguer
KR
,
Stiles
AD
.
Perinatal regionalization and neonatal mortality in North Carolina, 1968-1994
.
Am J Obstet Gynecol
.
2001
;
184
(
6
):
1302
1307
3
Neto
MT
.
Perinatal care in Portugal: effects of 15 years of a regionalized system
.
Acta Paediatr
.
2006
;
95
(
11
):
1349
1352
4
Rashidian
A
,
Omidvari
AH
,
Vali
Y
, et al
.
The effectiveness of regionalization of perinatal care services--a systematic review
.
Public Health
.
2014
;
128
(
10
):
872
885
5
Paneth
N
,
Kiely
JL
,
Wallenstein
S
,
Marcus
M
,
Pakter
J
,
Susser
M
.
Newborn intensive care and neonatal mortality in low-birth-weight infants: a population study
.
N Engl J Med
.
1982
;
307
(
3
):
149
155
6
Sanderson
M
,
Sappenfield
WM
,
Jespersen
KM
,
Liu
Q
,
Baker
SL
.
Association between level of delivery hospital and neonatal outcomes among South Carolina Medicaid recipients
.
Am J Obstet Gynecol
.
2000
;
183
(
6
):
1504
1511
7
Cifuentes
J
,
Bronstein
J
,
Phibbs
CS
,
Phibbs
RH
,
Schmitt
SK
,
Carlo
WA
.
Mortality in low birth weight infants according to level of neonatal care at hospital of birth
.
Pediatrics
.
2002
;
109
(
5
):
745
751
8
Phibbs
CS
,
Baker
LC
,
Caughey
AB
,
Danielsen
B
,
Schmitt
SK
,
Phibbs
RH
.
Level and volume of neonatal intensive care and mortality in very-low-birth-weight infants
.
N Engl J Med
.
2007
;
356
(
21
):
2165
2175
9
Rautava
L
,
Lehtonen
L
,
Peltola
M
, et al
;
PERFECT Preterm Infant Study Group
.
The effect of birth in secondary- or tertiary-level hospitals in Finland on mortality in very preterm infants: a birth-register study
.
Pediatrics
.
2007
;
119
(
1
):
e257
e263
10
Lasswell
SM
,
Barfield
WD
,
Rochat
RW
,
Blackmon
L
.
Perinatal regionalization for very low-birth-weight and very preterm infants: a meta-analysis
.
JAMA
.
2010
;
304
(
9
):
992
1000
11
Kilpatrick
SJ
,
Papile
L-A
,
Macones
GA
, eds;
American Academy of Pediatrics, American College of Obstetricians and Gynecologists
.
Guidelines for Perinatal Care
. 8th ed.
American Academy of Pediatrics
;
2017
12
Lupton
BA
,
Pendray
MR
.
Regionalized neonatal emergency transport
.
Semin Neonatol
.
2004
;
9
(
2
):
125
133
13
Lee
SK
,
McMillan
DD
,
Ohlsson
A
, et al
.
The benefit of preterm birth at tertiary care centers is related to gestational age
.
Am J Obstet Gynecol
.
2003
;
188
(
3
):
617
622
14
Akula
VP
,
Gould
JB
,
Kan
P
,
Bollman
L
,
Profit
J
,
Lee
HC
.
Characteristics of neonatal transports in California
.
J Perinatol
.
2016
;
36
(
12
):
1122
1127
15
Swartz
MF
,
Cholette
JM
,
Orie
JM
,
Jacobs
ML
,
Jacobs
JP
,
Alfieris
GM
.
Transfer of neonates with critical congenital heart disease within a regionalized network
.
Pediatr Cardiol
.
2017
;
38
(
7
):
1350
1358
16
Purkey
NJ
,
Ma
C
,
Lee
HC
, et al
.
Timing of transfer and mortality in neonates with hypoplastic left heart syndrome in California
.
Pediatr Cardiol
.
2021
;
42
(
4
):
906
917
17
Purkey
NJ
,
Ma
C
,
Lee
HC
, et al
.
Distance from home to birth hospital, transfer, and mortality in neonates with hypoplastic left heart syndrome in California
.
Birth Defects Res
.
2022
;
114
(
12
):
662
673
18
Escobar
GJ
,
McCormick
MC
,
Zupancic
JA
, et al
.
Unstudied infants: outcomes of moderately premature infants in the neonatal intensive care unit
.
Arch Dis Child Fetal Neonatal Ed
.
2006
;
91
(
4
):
F238
F244
19
Edwards
EM
,
Horbar
JD
.
Variation in use by NICU types in the United States
.
Pediatrics
.
2018
;
142
(
5
):
e20180457
20
Stephens
AS
,
Lain
SJ
,
Roberts
CL
,
Bowen
JR
,
Nassar
N
.
Survival, hospitalization, and acute-care costs of very and moderate preterm infants in the first 6 years of life: a population-based study
.
J Pediatr
.
2016
;
169
:
61
8.e3
21
Smyrni
N
,
Koutsaki
M
,
Petra
M
, et al
.
Moderately and late preterm infants: short- and long-term outcomes from a registry-based cohort
.
Front Neurol
.
2021
;
12
:
628066
22
Salazar
EG
,
Handley
SC
,
Greenberg
LT
,
Edwards
EM
,
Lorch
SA
.
Measuring quality of care in moderate and late preterm infants
.
J Perinatol
.
2022
;
42
(
10
):
1294
1300
23
Salazar
EG
,
Handley
SC
,
Greenberg
LT
,
Edwards
EM
,
Lorch
SA
.
Association between neonatal intensive care unit type and quality of care in moderate and late preterm infants
.
JAMA Pediatr
.
2023
;
177
(
3
):
278
285
24
Amsalu
R
,
Oltman
S
,
Medvedev
MM
, et al
.
Predicting the risk of 7-day readmission in late preterm infants in California: a population-based cohort study
.
Heal Sci Rep
.
2023
;
6
(
1
):
e994
25
Mullaney
DM
,
Edwards
WH
,
DeGrazia
M
.
Family-centered care during acute neonatal transport
.
Adv Neonatal Care
.
2014
;
14
(
5
Suppl 5
):
S16
S23
26
Aagaard
H
,
Hall
EOC
,
Ludvigsen
MS
,
Uhrenfeldt
L
,
Fegran
L
.
Parents’ experiences of neonatal transfer. a meta-study of qualitative research 2000-2017
.
Nurs Inq
.
2018
;
25
(
3
):
e12231
27
Freedman
S
.
Capacity and utilization in health care: the effect of empty beds on neonatal intensive care admission
.
Am Econ J Econ Policy
.
2016
;
8
(
2
):
154
185
28
Kunz
SN
,
Helkey
D
,
Zitnik
M
, et al
.
Quantifying the variation in neonatal transport referral patterns using network analysis
.
J Perinatol
.
2021
;
41
(
12
):
2795
2803
29
Vermont Oxford Network
.
Expanded database. Available at: https://public.vtoxford.org/data-and-reports/expanded-database/. Accessed July 30, 2021
30
Vermont Oxford Network
.
Manual of Operations Part 2
.
Vermont Oxford Network
;
2021
31
Kunz
SN
,
Zupancic
JAF
,
Rigdon
J
, et al
.
Network analysis: a novel method for mapping neonatal acute transport patterns in California
.
J Perinatol
.
2017
;
37
(
6
):
702
708
32
Howell
EM
,
Richardson
D
,
Ginsburg
P
,
Foot
B
.
Deregionalization of neonatal intensive care in urban areas
.
Am J Public Health
.
2002
;
92
(
1
):
119
124
33
Kastenberg
ZJ
,
Lee
HC
,
Profit
J
,
Gould
JB
,
Sylvester
KG
.
Effect of deregionalized care on mortality in very low-birth-weight infants with necrotizing enterocolitis
.
JAMA Pediatr
.
2015
;
169
(
1
):
26
32
34
Boghossian
NS
,
Geraci
M
,
Phibbs
CS
,
Lorch
SA
,
Edwards
EM
,
Horbar
JD
.
Trends in resources for neonatal intensive care at delivery hospitals for infants born younger than 30 weeks’ gestation, 2009-2020
.
JAMA Netw Open
.
2023
;
6
(
5
):
e2312107
35
Boghossian
NS
,
Greenberg
LT
,
Edwards
EM
,
Horbar
JD
.
Transfer patterns of very low birth weight infants for convalescent care
.
Pediatrics
.
2022
;
149
(
6
):
2021054866
36
Bourque
SL
,
Levek
C
,
Melara
DL
,
Grover
TR
,
Hwang
SS
.
Prevalence and predictors of back-transport closer to maternal residence after acute neonatal care in a regional NICU
.
Matern Child Health J
.
2019
;
23
(
2
):
212
219
37
Kunz
SN
,
Dukhovny
D
,
Profit
J
,
Mao
W
,
Miedema
D
,
Zupancic
JAF
.
Predicting successful neonatal retro-transfer to a lower level of care
.
J Pediatr
.
2019
;
205
:
272
276.e1
38
Shlossman
PA
,
Manley
JS
,
Sciscione
AC
,
Colmorgen
GH
.
An analysis of neonatal morbidity and mortality in maternal (in utero) and neonatal transports at 24-34 weeks’ gestation
.
Am J Perinatol
.
1997
;
14
(
8
):
449
456
39
Helenius
K
,
Longford
N
,
Lehtonen
L
,
Modi
N
,
Gale
C
;
Neonatal Data Analysis Unit and the United Kingdom Neonatal Collaborative
.
Association of early postnatal transfer and birth outside a tertiary hospital with mortality and severe brain injury in extremely preterm infants: observational cohort study with propensity score matching
.
BMJ
.
2019
;
367
:
l5678
40
Sinclair
L
,
Spence
K
,
Galea
C
;
NSW Neonatal Clinical Nurse Consultant Network and NICUs
.
Influence of patterns of neonatal transfers on breastfeeding outcomes
.
J Paediatr Child Health
.
2021
;
57
(
9
):
1473
1478
41
Profit
J
,
Kowalkowski
MA
,
Zupancic
JA
, et al
.
Baby-MONITOR: a composite indicator of NICU quality
.
Pediatrics
.
2014
;
134
(
1
):
74
82
42
Okoroh
EM
,
Kroelinger
CD
,
Lasswell
SM
,
Goodman
DA
,
Williams
AM
,
Barfield
WD
.
United States and territory policies supporting maternal and neonatal transfer: review of transport and reimbursement
.
J Perinatol
.
2016
;
36
(
1
):
30
34

Supplementary data