+ MODEL

Pediatrics and Neonatology xxx (xxxx) xxx

Available online at www.sciencedirect.com

ScienceDirect

journal homepage: http://www.pediatr-neonatol.com

Original Article

Clinical outcomes of different patent ductus

arteriosus treatment in preterm infants born
between 28 and 32 weeks in Taiwan
Hao-Wei Chung a,b,c, Shu-Ting Yang d, Fu-Wen Liang e,f,g,
Hsiu-Lin Chen d,h,* on behalf of Taiwan Premature Infant
Follow-up Network

a
Department of Pediatrics, College of Medicine, Kaohsiung Medical University, Kaohsiung, Taiwan
b
Department of Biological Science and Technology, National Yang Ming Chiao-Tung University, Hsinchu,
Taiwan
c
Department of Pediatrics, Kaohsiung Municipal Siaogang Hospital, Kaohsiung, Taiwan
d
Division of Neonatology, Department of Pediatrics, Kaohsiung Medical University Hospital, Kaohsiung
Medical University, Kaohsiung, Taiwan
e
Department of Public Health, College of Health Sciences, Kaohsiung Medical University, Kaohsiung,
Taiwan
f
Department of Medical Research, Kaohsiung Medical University Hospital, Kaohsiung Medical
University, Kaohsiung, Taiwan
g
Center for Big Data Research, Kaohsiung Medical University, Kaohsiung, Taiwan
h
Department of Respiratory Therapy, College of Medicine, Kaohsiung Medical University, Kaohsiung,
Taiwan

Received Oct 27, 2022; received in revised form Dec 9, 2022; accepted Dec 20, 2022
Available online - - -

Key Words Background: The patent ductus arteriosus (PDA) treatment in very preterm infants is contro-
bronchopulmonary versial. This study focused on preterm infants born at 28e32 weeks of gestation and analyzed
dysplasia; the association between various PDA treatments and clinical outcomes.
conservative Methods: We conducted a retrospective cohort study of infants born at 28e32 weeks of gesta-
treatment; tion between 2016 and 2019 at 22 hospitals in the Taiwan Premature Infant Follow-up Network.
necrotizing We categorized the infants into four groups according to treatment strategies: medication, pri-
enterocolitis; mary surgery, medication plus surgery, or conservative treatment.
patent ductus Results: A total of 1244 infants presented with PDA, and 761 (61.1%) were treated. Medication
arteriosus was the predominant treatment (50.0%), followed by conservative treatment (38.9%), medica-
treatment; tion plus surgery (7.6%), and primary surgery (3.5%). The risk of mortality was not reduced in
the active treatment group compared to the conservative treatment group. There was a higher
Abbreviations: BPD, bronchopulmonary dysplasia; PDA, patent ductus arteriosus; NEC, necrotizing enterocolitis.
* Corresponding author. Division of Neonatology, Department of Pediatrics, Kaohsiung Medical University Hospital,100, TzYou 1st Rd.,
Kaohsiung City 80756, Taiwan.
E-mail address: ch840062@kmu.edu.tw (H.-L. Chen).

https://doi.org/10.1016/j.pedneo.2022.12.004
1875-9572/Copyright ª 2023, Taiwan Pediatric Association. Published by Elsevier Taiwan LLC. This is an open access article under the CC BY-
NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

Please cite this article as: H.-W. Chung, S.-T. Yang, F.-W. Liang et al., Clinical outcomes of different patent ductus arteriosus treatment in
preterm infants born between 28 and 32 weeks in Taiwan, Pediatrics and Neonatology, https://doi.org/10.1016/j.pedneo.2022.12.004
 + MODEL
H.-W. Chung, S.-T. Yang, F.-W. Liang et al.

prevalence of severe intraventricular hemorrhage, necrotizing enterocolitis (NEC), and any de-
preterm infant
gree of bronchopulmonary dysplasia (BPD) in both the primary surgery and medication plus sur-
gery groups than in the conservative treatment group. After adjustment, both the primary
surgery and medication plus surgery groups still had higher odds ratios for the occurrence of
NEC and any degree of BPD.
Conclusions: Compared with active PDA treatment, conservative treatment for PDA did not in-
crease the risk of mortality and morbidity in very preterm infants born at 28e32 weeks of gestation.
The risks and benefits of surgery (PDA ligation) in these infants must be considered cautiously.
Copyright ª 2023, Taiwan Pediatric Association. Published by Elsevier Taiwan LLC. This is an open
access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/
4.0/).

1. Introduction Foundation of Taiwan since 1995 to improve the quality of

care, education, and research of preterm infants in Taiwan.
Approximately 50% of very low birth weight (VLBW) infants Data from the TPFN database were collected and analyzed.
have patent ductus arteriosus (PDA).1,2 In 1959, Burnard was Demographic, perinatal, and neonatal data of VLBW infants
the first to report an association between respiratory distress born in 22 neonatal care centers in Taiwan were recor-
syndrome (RDS) and persistent PDA patency in preterm in- ded.16 VLBW is defined as the birth weight (BW) of
fants.3 Since then, an increasing number of studies have 401e1500 g, all data were reported using this uniform
highlighted the association of hemodynamically significant definition from member hospitals. In this retrospective
PDA with death, bronchopulmonary dysplasia (BPD), peri- nationwide, multicenter cohort study, data on 2559 VLBW
ventricular leukomalacia (PVL), intraventricular hemorrhage infants, born between gestational age (GA) 28 and 32 weeks
(IVH), and necrotizing enterocolitis (NEC) in VLBW infants.4,5 from January 1, 2016, to December 31, 2019, were obtained
Therefore, even if the cause was uncertain, the aggressive from the TPFN database. PDA was defined in the TPFN
closing of a persistent PDA was the primary treatment database only when there was an evidence of a shunt on
strategy for VLBW infants 30 years ago.6,7 However, in the Doppler echogenic examination with a systolic murmur.
last decade, treatment, regardless of the method, such as Meanwhile, the infants must have one of the following
medication prophylaxis with indomethacin and ibuprofen or clinical symptoms: hyperdynamic precordium, bounding
early routine surgical ligation, did not lead to improvements pulses, wide pulse pressure, pulmonary vascular conges-
in any short- or long-term benefit aside from PDA closure tion, or cardiomegaly. Infants with major birth defects and
itself for VLBW infants.8e11 In addition, preterm infants may with PDA who died within 48 h after birth were excluded
develop adverse effects to treatment, such as pneumothorax from this study (Fig. 1). This study was approved by the
or vocal cord palsy after surgical ligation or bleeding and ethical standards of the Institutional Review Board of
oliguria after treatment with medication, and the treatment Kaohsiung Medical University Hospital (IRB number:
itself might be associated with an increased risk of morbid- KMUHIRB-SV(I)-20200,079, date: 2020/11/6).
ities of prematurity, such as BPD and NEC.10,12,13 Therefore,
the trend of conservative treatment for persistent patency 2.2. Clinical outcomes
of PDA is increasing.14 Nonetheless, there are conflicting
results for patients who do not receive active interventions Major outcomes were strictly defined and recorded,
for persistent PDA in different study designs.15 Although including death before discharge, severe IVH (defined as
VLBW infants born between 28 and 32 weeks of gestation grade 3 or 4 IVH),17 PVL, NEC (defined as stage 2 according
have an increased possibility of spontaneous closure, less to Bell’s criteria18), severe retinopathy of prematurity
severe disease, and are less clinically complex, there is (ROP) (defined as stage  III), and treated ROP (defined as
limited information on the risks and benefits of different stage  III and received treatment). Any degree of BPD was
treatment strategies for PDA compared to the wait-and-see defined as at least 28 days of oxygen dependence during
strategy (conservative treatment) for these neonates. This hospitalization.19 Positive pressure ventilation (PPV) sup-
study focuses on applying real-world data in a national port at 36 weeks of gestation was defined as using a high-
retrospective cohort to compare the clinical outcomes be- flow nasal cannula, nasal continuous positive airway pres-
tween active and conservative treatments for PDA in VLBW sure, or mechanical intubation at 36 weeks of post-
infants born between 28 and 32 weeks of gestation. menstrual age (PMA). The duration of PPV support,
neonatal intensive care unit (NICU) days and PMA at
discharge were also recorded.
2. Methods
2.3. Different treatment strategies for PDA
2.1. Study population
Infants with PDA who received only medication (indo-
The Taiwan Premature Infant Follow-up Network (TPFN) methacin, ibuprofen, or acetaminophen) for PDA were
database has been funded by the Premature Baby categorized into the medication treatment group. The

2
 + MODEL
Pediatrics and Neonatology xxx (xxxx) xxx

Fig. 1 Flow chart of enrolled neonates from TPFN database.

TPFN database had no information on the route (intrave- 2.4. Covariates

nous or oral) of the prescribed medication used for closing
the PDA. Infants with PDA who underwent surgical ligation Demographic, perinatal, and neonatal data were reported
without medication treatment were categorized into the for the following characteristics: maternal age, BW, GA,
surgery alone group. If the infants with PDA had records of multiple gestations, Apgar score at 1 and 5 min, sex,
receiving medication treatment and surgical ligation, they gestational diabetes, clinical chorioamnionitis, use of
were categorized into the medication plus surgery group. antenatal steroids, mode of delivery, surfactant adminis-
Infants with PDA who did not receive any medication for tration, and presence or absence of RDS. Prolonged pre-
PDA closure (i.e., indomethacin, ibuprofen, or acetamino- term membranes (PPROM) were defined as the rupture of
phen) or surgical ligation were categorized into the con- membranes >18 h before labor and 37 weeks. Early-onset
servative treatment group. The number of patients in these sepsis was defined as positive blood culture results and
four groups is listed in Fig. 1. clinical signs of infection in neonates.

3
 + MODEL
H.-W. Chung, S.-T. Yang, F.-W. Liang et al.

2.5. Statistical analysis steroid use, 5-min Apgar score, surfactant use, and RDS
receiving IPPV to explore the effects of different PDA
Intergroup comparisons of the demographic and clinical treatment strategies on infant mortality and significant
characteristics and outcomes among the different PDA morbidity in infants born at 28e32 weeks of gestation
treatments for these four groups were performed using the (Fig. 2). With conservative treatment as the reference
c2 test for categorical variables and analysis of variance group, there were no differences in mortality risks, severe
(ANOVA) for continuous variables. Adjusted odds ratios IVH, PVL, severe ROP, and treated ROP between active
(aORs) and 95% confidence intervals (CIs) for major mor- treatment and conservative treatment after adjusting for
bidities and mortality were estimated using logistic confounders. The risk of the composite outcome of death or
regression analysis after adjusting for potential con- need for PPV support at PMA 36 weeks was significantly
founders. All statistical tests were two-sided, and a p-value increased in the surgery alone group (aOR: 3.50; 95% CI:
< 0.05 was considered statistically significant. The Bon- 1.57e8.31, p Z 0.003) and the medication plus surgery
ferroni correction was applied to control for multiple group (aOR: 2.91; 95% CI: 1.66e5.17, p < 0.001) but not in
comparisons to correct false positives. Data were analyzed the medication treatment group (aOR: 1.03; 95% CI:
using the R software (version 4.1.0; R Foundation for Sta- 0.76e1.40, p Z 0.85) compared with the conservative
tistical Computing, Vienna, Austria). treatment group. In addition, the medicine plus surgery and
surgery alone groups had higher risks of any degree of BPD.
NEC compared with the conservative treatment group, but
3. Results there was no significant difference between the medication
and conservative treatment groups (BPD, aOR: 5.58; 95% CI:
The proportion of patients diagnosed with PDA in the TPFN 3.07e10.6, p < 0.001; aOR: 5.14; 95% CI: 2.26e13.30,
database was 49.2% (1261/2559). Seventeen patients were p < 0.001; NEC, aOR: 4.84; 95% CI: 1.78e13.2, p Z 0.002,
further excluded because of death before PDA treatment aOR: 3.88; 95% CI: 0.95e13.50, p Z 0.04, respectively).
(<48 h after birth). Finally, 1244 VLBW infants with PDA
born at GA 28e32 weeks were eligible for analysis (Fig. 1).
There were 50% (622/1244) who received medication 4. Discussion
treatment, 38.9% (483/1244) in the conservative treatment
group, 7.6% (95/1244) who received medication plus sur- In this nationwide cohort study, we explored the effects of
gery treatment, and 3.5% (44/1244) who underwent surgery different PDA treatment strategies on mortality and major
alone. Comparisons of the maternal and neonatal charac- morbidity in infants born at 28e32 weeks of gestation. The
teristics among the four groups are shown in Table 1. GA, infants in the conservative treatment group had a lower
BW, use of antenatal steroids, 5-min Apgar score, surfac- percentage of death or need for PPV at PMA 36 weeks,
tant use, and RDS of patients who received intermittent shorter days of PPV and NICU days, and earlier PMA while at
positive pressure ventilation (IPPV) were significantly discharge compared with the other three groups with active
different among the four groups. Infants who underwent treatment for PDA.
surgery alone were the most immature, had a lower BW, Although conservative treatment for VLBW with PDA is
the lowest 5-min Apgar score, and most often received becoming increasingly popular, a variety of treatment
antenatal steroids and surfactants after birth. strategies are still reported because there is no precise
The number of enrolled infants with different PDA indication of who requires active treatment.20,21 In a
treatments according to GA is shown in Table 2. As we can retrospective study of VLBW infants in California, the au-
see, there was an increasing trend of increased GA in the thors concluded that there was an increasing trend toward
conservative treatment group. In contrast, there was a not treating patients (60.5%e78.3%) for VLBW infants with
decreasing trend in increased GA in the medication treat- PDA compared to more intensive treatments.14 A nation-
ment group. wide study in Korea between 2015 and 2018 also showed
Table 3 shows the outcomes of different PDA manage- that the trend for treating PDA in VLBW infants with med-
ment strategies in very preterm infants born at 28e32 ical treatment decreased significantly from 2015 to 2018
weeks. There was no significant difference in the incidence (62%e53%), and the improved clinical outcomes in VLBW
of mortality, PVL, severe ROP, or treated ROP among the infants would benefit from the increased use of conserva-
four groups (Table 3). Compared with the other three tive management for PDA.22 In Europe, a birth cohort study
groups, the infants in the conservative treatment group had of very preterm infants (31 weeks of gestation) from 19
a lower percentage of death or need for PPV at PMA 36 countries revealed regional variations in PDA treatment
weeks, shorter PPV and NICU days, and earlier PMA at from 10 to 39% between regions, and treatment for PDA was
discharge. The infants who underwent surgery alone had a associated with a higher adjusted risk ratio of BPD or death
higher percentage of death or need for PPV at PMA 36 in all regions.23 In our study, we revealed that preterm in-
weeks, higher incidences of any degree of BPD, severe IVH, fants born between 28 and 32 weeks of gestation who
and NEC, and longer PPV days and NICU stay than those in received any PDA treatment were smaller, younger, and
the conservative treatment group. The post hoc analysis had more severe respiratory distress with an indication to
showed no differences in NEC, severe IVH, or any degree of receive more active PDA treatment (Tables 1 and 2). These
BPD between the medication and conservative treatment findings indicated that even though there is a high possi-
groups. bility of spontaneous closure for ductus arteriosus in VLBW
We further performed a multivariable logistic regression infants born at 28 weeks GA1,25,26 the clinicians in
analysis adjusted for GA, birth body weight, antenatal neonatal intensive care units in Taiwan might be still afraid

4
 + MODEL
Pediatrics and Neonatology xxx (xxxx) xxx

Table 1 Demographic and clinical characteristics among different PDA management in very preterm infants born at 28e32
weeks.
Medication Conservative Medication Surgery P-value
treatment treatment plus surgery alone
(n Z 622) (n Z 483) (n Z 95) (n Z 44)
GA, weeks 29.3  1.2c 29.9  1.3 29.0  1.2c 28.9  1.3c <0.001
Birth weight, g 1163.8  226.3c 1235.6  205.6 1089.7  237.7c 1028.5  281.4c <0.001
Male sex 313 (50.3) 256 (53.0) 41 (43.2) 26 (59.1) 0.224
Cesarean section 491 (78.9) 364 (75.4) 73 (76.8) 38 (86.4) 0.260
Multiple gestations 200 (32.2) 164 (34.0) 32 (36.7) 12 (27.3) 0.078
PPROM 138 (22.2) 132 (27.3) 24 (25.3) 9 (20.5) 0.229
Chorioamnionitis 15 (24.1) 16 (3.31) 3 (3.16) 0 (0.00) 0.623
Antenatal MgSO4 326 (52.1) 241 (49.9) 50 (52.6) 26 (59.1) 0.625
Antenatal steroid 509 (85.2)b 362 (75.0) 82 (86.3)a 39 (88.6)a 0.004
1-min Apgar score 6.0  1.9 6.2  2.0 6.1  1.7 5.5  2.2 0.073
5-min Apgar score 7.9  1.5 8.1  1.6 8.0  1.1 7.3  2.0a 0.010
Intubation at birth 128 (20.6) 72 (14.9) 22 (23.2) 7 (15.9) 0.057
Surfactant use 288 (51.0)c 112 (25.28) 41 (48.8)c 25 (61.0)c <0.001
RDS 565 (90.8) 443 (91.7) 84 (88.4) 41 (93.2) 0.715
RDS received IPPV 175 (32.2)c 104 (21.5) 26 (29.9)c 13 (29.6)c 0.002
Early-onset sepsis 11 (1.8) 12 (2.5) 0 (0.0) 0 (0.0) 0.415
Data are presented as mean  standard deviation or n (%).
PDA: patent ductus arteriosus; GA: gestational age; PPROM: Prolonged preterm membranes; RDS: respiratory distress syndrome; IPPV:
intermittent positive pressure ventilation.
a
Post hoc analysis showed p < 0.05 compared with conservative treatment.
b
Post hoc analysis showed p < 0.01 compared with conservative treatment.
c
Post hoc analysis showed p < 0.001 compared with conservative treatment.

of the association of hemodynamically significant PDA with and how to treat PDA in different hospitals is the conflicting
complications of prematurity, such as death, BPD, or evidence regarding morbidity and mortality among
neurological complication in these preterm infants.4,5 That different PDA treatment strategies for VLBW infants. The
lead to the decision to use active treatment for PDA OR for mortality, IVH, or NEC in the conservative treatment
because of no consensus on the precise indication of who group compared with any medication was not significantly
requires active treatment. Furthermore, as there is evi- different in a network meta-analysis that included 68 RCTs
dence to show that there are no differences in mortality or of medication treatment for preterm infants less than 37
morbidity associated with conservative management weeks.10 Nonetheless, another meta-analysis of 12 cohort
compared with that of active treatment,15 including in our studies that included preterm infants <32 weeks of gesta-
study, the risks and benefits of active treatment for PDA in tion or VLBW infants with PDA showed that no active
infants born between 28 and 32 weeks of gestation must be intervention was associated with lower odds of BPD, IVH,
considered cautiously. The consensus for indication of who and NEC, but a higher OR of mortality than any active PDA
requires active treatment should be discussed among the treatment.15 The study of national epidemiologic data on
experts in Taiwan. infants treated for PDA in Korea also showed that surgical
Although the capability of functional PDA closure is and/or medical treatments are associated with a higher risk
associated with hemodynamic forces, pulmonary vascular of morbidity.22 In our study, among VLBW infants born be-
resistance, vessel structure maturity, and biochemical tween 28 and 32 weeks of gestation, although there were
oxygen-sensing mechanisms, GA at birth is inversely pro- no differences in the OR of mortality, severe IVH, PVL,
portional to the time it takes to achieve closure.24 One of treated ROP, ROP, NEC, any degree of BPD, or PPV at PMA
the reasons for the lack of consensus regarding when, who, 36 weeks between infants with conservative treatment and

Table 2 The numbers of enrolled infants among different PDA treatments by gestational age.
28 weeks 29 weeks 30 weeks 31 weeks 32 weeks
(N Z 333) (N Z 359) (N Z 269) (N Z 164) (N Z 119)
Medication treatment 55.3% (184/333) 55.2% (198/359) 48.7% (131/269) 39.0% (64/164) 37.8% (45/119)
Conservative treatment 25.8% (86/333) 34.5% (124/359) 43.9% (118/269) 53.1% (87/164) 57.1% (68/119)
Medicine plus Surgery 12.6% (42/333) 6.7% (24/359) 5.9% (16/269) 5.5% (9/164) 3.4% (4/119)
Surgery alone 6.3% (21/333) 3.6% (13/359) 1.5% (4/269) 2.4% (4/164) 1.7% (2/119)
PDA: patent ductus arteriosus; GA: gestational age.

5
 + MODEL
H.-W. Chung, S.-T. Yang, F.-W. Liang et al.

Table 3 Outcome among different PDA management in very preterm infants born at 28e32 weeks.
Medication Conservative Medicine plus Surgery P-value
treatment treatment Surgery alone
(n Z 622) (n Z 483) (n Z 95) (n Z 44)
Mortality 19 (3.0) 15 (3.1) 3 (3.16) 2 (4.6) 0.890
PPV at PMA 36 Weeks 221 (35.5)a 136 (28.2) 61 (64.2)c 31 (70.5)c <0.001
Death/PPV at PMA 36 Weeks 234 (37.6)b 148 (30.6) 62 (65.3)c 32 (72.7)c <0.001
Any degree of BPD 223 (35.9) 161 (33.3) 73 (76.8)c 36 (81.8)c <0.001
Duration of PPV, days 47.9 (33.1)c 38.21 (24.6) 77.8 (52.8)c 82.9 (43.6)c <0.001
Severe IVH (Grade 3) 23 (3.7) 18 (3.7) 7 (7.4) 8 (18.2)c 0.001
PVL 38 (6.1) 26 (5.4) 8 (8.4) 5 (11.6) 0.272
Severe ROP (Stage 3) 24 (3.9) 20 (4.1) 6 (6.3) 4 (9.1) 0.230
Treated ROP 27 (10.2) 17 (9.2) 5 (13.9) 3 (18.8) 0.467
NEC grade  IIB 22 (3.5) 13 (2.7) 10 (10.5)b 6 (13.6)b 0.003
NICU days 46.1 (32.8)b 39.4 (22.6) 72.2 (48.5)c 83.3 (49.6)c <0.001
PMA at discharge, weeks 38.9 (4.9)b 37.9 (3.2) 41.8 (7.0)c 43.7 (7.0)c <0.001
Data are presented as mean  standard deviation or n (%).
PDA: patent ductus arteriosus; VLBW: very low birth weight; PPV: Positive pressure ventilation including high-flow nasal cannula, nasal
continuous positive airway pressure, intermittent positive pressure ventilation. IVH: intraventricular hemorrhage; PVL: periventricular
leukomalacia; ROP retinopathy of prematurity; NEC: necrotizing enterocolitis, NICU: neonatal intensive care unit; PMA: post maternal
age, BPD: bronchopulmonary dysplasia.
a
Post hoc analysis showed p < 0.05 compared with conservative treatment.
b
Post hoc analysis showed p < 0.01 compared with conservative treatment.
c
Post hoc analysis showed p < 0.001 compared with conservative treatment.

medication treatment, the infants in the conservative in accordance with previous studies, including preterm in-
treatment group had a lower percentage of death or need fants born at GA < 28 weeks. A meta-analysis of 39 cohort
PPV at PMA 36 weeks, shorter days of PPV and NICU days, studies showed that PDA ligation was associated with an
and earlier PMA while at discharge compared with the other increased OR for chronic lung disease.12 A national cohort
three groups with active treatment for PDA. Our results study of preterm infants born at less than 32 weeks of
correspond with the current findings, in which most VLBW gestation in Finland showed that therapies for PDA were
infants born at 28 weeks without any intervention showed associated with severe BPD, especially in preterm infants
no increase in morbidity and mortality.1,27 who received both medication and surgical closure.33
In our study, the mortality risk did not differ between Another national retrospective study of VLBW infants in
the conservative and active treatment strategy groups. In Korea showed that surgery alone and medication plus sur-
previous studies, the lower mortality rate in the active gery had a higher OR of BPD than conservative treatment.22
treatment group may be partly attributable to variations in Our finding was consistent with the study that the risk of
the mortality rate among preterm infants with different NEC was greater in PDA-ligated infants than in non-ligated
GAs.12,28 No active intervention is sometimes the only op- infants.34 Intestinal ischemia is a pathogenic mechanism of
tion because infants are contraindicated from active NEC.35 Preterm infants who undergo surgery have a risk of
treatment due to their poor clinical status.29 We speculated intestinal ischemia because of hemodynamic instability
that active treatment would not reduce the risk of death in post-ligation.34 However, our results showing higher risks of
our study because of a more homogeneous trend of spon- NEC and BPD in the surgery alone and medication plus
taneous PDA closure among VLBW infants at GA 28e32 surgery groups of infants need to be cautiously interpreted
weeks than among extremely preterm infants born at less because there is still evidence showing the benefits of
than 28 weeks. ductal closure within a specific period in preterm in-
For the development of BPD, animal models have fants.36,37 Determining the appropriate timing of ductal
confirmed the causal effect between ligation and chronic closure by surgical ligation might be an important issue for
lung disease,30 and BPD may be related to the exposure these preterm infants. A meta-analysis study by Yan et al.
time to PDA, regardless of whether they exist.31,32 This revealed that early ligation (the cut-off time of early and
could explain the physicians were more willing to consider late ligation was two or three weeks of life) might have
immediate ductal closure when neonates received me- fewer intubation days, earlier full oral feeding, and better
chanical ventilation after birth.25 However, our study nutritional status at PMA 36 weeks compared with late
revealed that both surgery alone and medication plus sur- ligation.38 Because of data limitations, it cannot reach the
gery for PDA were associated with an increased risk of any specific ligation time now. Regardless of the higher in-
degree of BPD and PPV at PMA 36 weeks for VLBW infants cidences of NEC and BPD or more extended NICU stays the
born between 28 and 32 weeks of gestation. Our results are duration of respiratory support in infants who ever

6
 + MODEL
Pediatrics and Neonatology xxx (xxxx) xxx

Fig. 2 The effects of different PDA treatment strategies on infant mortality and major morbidity born at 28e32 weeks of
gestation. Adjusted for gestational age, birth body weight, antenatal steroid use, 5-min Apgar score, surfactant use, and RDS
receiving IPPV.

underwent surgery might reflect selection bias in a retro- medication treatment. Without the intervention timeline,
spective study. Infants with a lower GA and more medically it was difficult to determine whether the morbidities were
complex clinical conditions have increased rates of active confounders or outcomes. Second, echocardiographic de-
PDA treatment rather than regional variation in expert tails were not available in the TPFN database. In our study,
opinion or training culture.21 For infants with a more the clinical symptoms we used to describe PDA could not be
mature compensation capability and higher spontaneous used to identify the medical complexity of their clinical
closure rate, further investigation should explore the best conditions at preterm birth or during PDA shunting.39 Third,
PDA treatment strategies for VLBW infants born between 28 we excluded infants with PDA before treatment because
and 32 weeks with more medically complex clinical condi- they died within 48 h after birth. Some have questioned this
tions to balance the benefit and risks of PDA exposure. exclusion. Even though Ficial et al. revealed that 60% of
Our study has some limitations. First, the TPFN database medication treatment for PDA was started between the first
only provided information on PDA management, but not the 24 and 72 h of life.19 We excluded these infants because
clinical severity of PDA (hemodynamic significance), the they may have a contraindication for treating PDA or the
timing, dosage of medication, or natural response to time was too short of deciding how to treat, which we could

7
 + MODEL
H.-W. Chung, S.-T. Yang, F.-W. Liang et al.

not differentiate in the TPFN database. Finally, without 9. Benitz WE, Committee on Fetus and Newborn. American
imaging data, we could not determine the dynamic changes Academy of Pediatrics, Patent ductus arteriosus in preterm
in PDA under different treatment strategies. infants. Pediatrics 2016;137. https://doi.org/10.1542/
peds.2015-3730.
10. Mitra S, Florez ID, Tamayo ME, Mbuagbaw L, Vanniyasingam T,
5. Conclusion Veroniki AA, et al. Association of Placebo, indomethacin,
ibuprofen, and acetaminophen with closure of hemodynami-
In this nationwide study, we demonstrate the higher rates cally significant patent ductus arteriosus in preterm infants: a
of active treatments for PDA, and surgery alone and med- systematic review and meta-analysis. JAMA 2018;319:
1221e38.
icine plus surgery for PDA were strongly associated with
11. Clyman RI, Liebowitz M, Kaempf J, Erdeve O, Bulbul A,
more significant risks of NEC, any degree of BPD, and PPV at
Håkansson S, et al. PDA-TOLERATE trial: an exploratory ran-
PMA 36 weeks in VLBW infants born at GA 28e32 weeks. domized controlled trial of treatment of moderate-to-large
Meanwhile, we highlight that conservative treatment (no patent ductus arteriosus at 1 Week of age. J Pediatr 2019;205:
active intervention) was not inferior to active treatment for 41e48.e6.
PDA based on the clinical outcomes during their hospitali- 12. Weisz DE, More K, McNamara PJ, Shah PS. PDA ligation and health
zation among VLBW infants born at GA 28e32 weeks. The outcomes: a meta-analysis. Pediatrics 2014;133:e1024e46.
benefits and risks of surgery alone and medication plus 13. Ohlsson A, Walia R, Shah SS. Ibuprofen for the treatment of
surgery should be carefully considered in these infants. patent ductus arteriosus in preterm or low birth weight (or
both) infants. Cochrane Database Syst Rev 2020;2:CD003481.
14. Ngo S, Profit J, Gould JB, Lee HC. Trends in patent ductus
Data statement arteriosus diagnosis and management for very low birth weight
infants. Pediatrics 2017;139:e20162390.
The data presented in this study is available on request 15. Hundscheid T, Jansen EJS, Onland W, Kooi EMW, Andriessen P,
from the corresponding author. de Boode WP. Conservative management of patent ductus
arteriosus in preterm infants-A systematic review and meta-
analyses assessing differences in outcome measures between
Declaration of competing interest randomized controlled trials and cohort studies. Front Pediatr
2021;9:626261.
16. Su BH, Hsieh WS, Hsu CH, Chang JH, Lien R, Lin CH, et al.
None.
Neonatal outcomes of extremely preterm infants from Taiwan:
comparison with Canada, Japan, and the USA. Pediatr Neo-
Acknowledgments natol 2015;56:46e52.
17. Papile LA, Burstein J, Burstein R, Koffler H. Incidence and
evolution of subependymal and intraventricular hemorrhage: a
We thank Mr. Liu Yu-Chuan and the center for medical study of infants with birth weights less than 1,500 gm. J
informatics and statistics of Kaohsiung Medical University Pediatr 1978;92:529e34.
for the data cleaning and analysis. We also thank the Pre- 18. Bell MJ, Ternberg JL, Feigin RD, Keating JP, Marshall R,
mature Baby Foundation of Taiwan and the Taiwan Society Barton L, et al. Neonatal necrotizing enterocolitis. Therapeutic
of Neonatology for supporting the Taiwan premature infant decisions based upon clinical staging. Ann Surg 1978;187:1e7.
follow-up network. This work was supported by Kaohsiung 19. Jobe AH, Bancalari E. Bronchopulmonary dysplasia. Am J
Medical University Hospital (KMUH107-7R44) in Taiwan. Respir Crit Care Med 2001;163:1723e9.
20. Isayama T, Kusuda S, Adams M, Berti E, Battin M, Helenius K,
et al. International variation in the management of patent
References ductus arteriosus and its association with infant outcomes: a
survey and linked cohort study. J Pediatr 2022;244:24e29.e7.
1. Semberova J, Sirc J, Miletin J, Kucera J, Berka I, Sebkova S, 21. Runte KE, Flyer JN, Edwards EM, Soll RF, Horbar JD, Yeager SB.
et al. Spontaneous closure of patent ductus arteriosus in in- Variation of patent ductus arteriosus treatment in very low
fants 1500 g. Pediatrics 2017;140:e20164258. birth weight infants. Pediatrics 2021;148:e2021052874.
2. Lee JA, Kim MJ, Oh S, Choi BM. Current status of therapeutic 22. Park J, Yoon SJ, Han J, Song IG, Lim J, Shin JE, et al. Patent
strategies for patent ductus arteriosus in very-low-birth-weight ductus arteriosus treatment trends and associated morbidities
infants in Korea. J Korean Med Sci 2015;30(Suppl 1):S59e66. in neonates. Sci Rep 2021;11:10689.
3. Burnard ED. The cardiac murmur in relation to symptoms in the 23. Edstedt Bonamy AK, Gudmundsdottir A, Maier RF, Toome L,
newborn. Br Med J 1959;1:134e8. Zeitlin J, Bonet M, et al. Patent ductus arteriosus treatment in
4. Hamrick SE, Hansmann G. Patent ductus arteriosus of the very preterm infants: a European Population-based cohort
preterm infant. Pediatrics 2010;125:1020e30. study (EPICE) on variation and outcomes. Neonatology 2017;
5. Terrin G, Di Chiara M, Boscarino G, Metrangolo V, Faccioli F, 111:367e75.
Onestà E, et al. Morbidity associated with patent ductus 24. Hamrick SEG, Sallmon H, Rose AT, Porras D, Shelton EL,
arteriosus in preterm newborns: a retrospective case-control Reese J, et al. Patent ductus arteriosus of the preterm infant.
study. Ital J Pediatr 2021;47:9. Pediatrics 2020;146:e20201209.
6. Benitz WE. Treatment of persistent patent ductus arteriosus in 25. Clyman RI, Couto J, Murphy GM. Patent ductus arteriosus: are
preterm infants: time to accept the null hypothesis? J Peri- current neonatal treatment options better or worse than no
natol 2010;30:241e52. treatment at all? Semin Perinatol 2012;36:123e9.
7. Kitterman JA. Patent ductus arteriosus: current clinical status. 26. Su BH, Lin HY, Chiu HY, Tsai ML, Chen YT, Lu IC. Therapeutic
Arch Dis Child 1980;55:106e9. strategy of patent ductus arteriosus in extremely preterm in-
8. Bose CL, Laughon MM. Patent ductus arteriosus: lack of evi- fants. Pediatr Neonatol 2020;61:133e41.
dence for common treatments. Arch Dis Child Fetal Neonatal 27. Chen HL, Yang RC, Lee WT, Lee PL, Hsu JH, Wu JR, et al. Lung
Ed 2007;92:F498e502. function in very preterm infants with patent ductus arteriosus

8
 + MODEL
Pediatrics and Neonatology xxx (xxxx) xxx

under conservative management: an observational study. BMC Nationwide cohort study. J Matern Fetal Neonatal Med 2018;
Pediatr 2015;15:167. 31:2576e83.
28. Mirea L, Sankaran K, Seshia M, Ohlsson A, Allen AC, Aziz K, 34. Othman HF, Linfield DT, Mohamed MA, Aly H. Ligation of patent
et al. Treatment of patent ductus arteriosus and neonatal ductus arteriosus in very low birth weight premature infants.
mortality/morbidities: adjustment for treatment selection Pediatr Neonatol 2020;61:399e405.
bias. J Pediatr 2012;161:689e694.e1. 35. Nankervis CA, Giannone PJ, Reber KM. The neonatal intestinal
29. Weisz DE, Mirea L, Resende MHF, Ly L, Church PT, Kelly E, et al. vasculature: contributing factors to necrotizing enterocolitis.
Outcomes of surgical ligation after unsuccessful Pharmaco- Semin Perinatol 2008;32:83e91.
therapy for patent ductus arteriosus in neonates born 36. Lee JH, Lee HJ, Park HK, Ahn JH, Kim HS, Jang HJ, et al.
extremely preterm. J Pediatr 2018;195:292e296.e3. Surgical ligation of patent ductus arteriosus in preterm neo-
30. Chang LY, McCurnin D, Yoder B, Shaul PW, Clyman RI. Ductus nates weighing less than 1500g: a 9-year single center experi-
arteriosus ligation and alveolar growth in preterm baboons with ence. J Cardiothorac Surg 2020;15:144.
a patent ductus arteriosus. Pediatr Res 2008;63:299e302. 37. Weisz DE, Mirea L, Rosenberg E, Jang M, Ly L, Church PT, et al.
31. Liebowitz M, Clyman RI. Prophylactic indomethacin compared Association of patent ductus arteriosus ligation with death or
with delayed conservative management of the patent ductus neurodevelopmental impairment among extremely preterm
arteriosus in extremely preterm infants: effects on neonatal infants. JAMA Pediatr 2017;171:443e9.
outcomes. J Pediatr 2017;187:119e126.e1. 38. Yan H, Ma F, Li Y, Zhou K, Hua Y, Wan C. The optimal timing of
32. Clyman RI, Hills NK, Liebowitz M, Johng S. Relationship be- surgical ligation of patent ductus arteriosus in preterm or very-
tween duration of infant exposure to a moderate-to-large low-birth-weight infants: a systematic review and meta-anal-
patent ductus arteriosus shunt and the risk of developing ysis. Medicine (Baltimore) 2020;99:e19356.
bronchopulmonary dysplasia or death before 36 weeks. Am J 39. Shepherd JL, Noori S. What is a hemodynamically significant
Perinatol 2020;37:216e23. PDA in preterm infants? Congenit Heart Dis 2019;14:21e6.
33. Härkin P, Marttila R, Pokka T, Saarela T, Hallman M. Morbidities
associated with patent ductus arteriosus in preterm infants.