Journal Reading - The Relationship Between Preoperative Feeding Exposure and Post Operative Outcomes in Infants with CHD

gizi

1. STRUCTURE
JURNAL READING
The Relationship Between Preoperative Feeding
Exposures and Postoperative Outcomes in Infants
with Congenital Heart Disease
Jasmeet Kataria-Hale,MD; Acacia Cognata, MD; Joseph Hagan, ScD; Jill Zender, CPNP-AC; Paige
Sheaks, PA-C; Scott Osborne,MD; Jeramy Roddy,MD; Amy Hair, MD.
PROGRAM PENDIDIKAN DOKTER SPESIALIS
DIVISI NUTRISI DAN PENYAKIT METABOLIK ANAK
RSUP DR KARIADI
Pembimbing:
dr. Rina Pratiwi, MSi.Med, Sp.A(K)
dr. Anna Mariska, Sp.A
DisusunOleh:
Dian Wulandari

2. TitleThe Relationship Between Preoperative Feeding Exposures and Postoperative Outcomes in Infants
With Congenital Heart Disease
ResearchersJasmeetKataria-Hale,MD;AcaciaCognata,MD;JosephHagan,ScD;JillZender,CPNP-AC;Paige
Sheaks,PA-C;ScottOsborne,MD;JeramyRoddy,MD;AmyHair,MD
Publication2021
InstitutionsDepartment of Pediatrics, Texas Children’s Hospital, Houston, TX. Section of Neonatology, Texas
Children’s Hospital, Houston, TX. Baylor College of Medicine, Texas Children’s Hospital, Houston,
TX. Lovelace Women’s Hospital, Albuquerque, NM. Section of Pediatric Critical Care Medicine,
Texas Children’s Hospital, Houston, TX.
AttachmentDOI:10.1097/PCC.0000000000002540
JOURNAL IDENTITY

3. ABSTRACT
Objectives : To evaluate the association of pre-operative risk factors and
post-operative outcomes in infants with complex congenital heart
disease
Design:Single-centerretrospectivecohortstudy
Setting:NeonatalICUandcardiovascularICU
Patients : Infants of all gestational ages, born at Texas Children’s Hospital
between 2010 and 2016, with complex congenital heart disease
requiring intervention prior to discharge.
Interventions: None

4. ABSTRACT
MeasurementsandMainResults:Atotalof399patientswereenrolledinthestudy.
Preoperativeriskfactorsincludedfeeding,typeoffeeding,feedingroute,andcardiaclesion.
Postoperativeoutcomesincludednecrotizingenterocolitis,hospitallengthofstay,anddaystofull
feeds.Theoccurrencerateofpostoperativenecrotizingenterocolitis(allstages)was8%.
Preoperativefeeding,typeoffeeding,feedingroute,andcardiaclesionwerenotassociatedwith
higheroddsofpostoperativenecrotizingenterocolitis.Cardiaclesionswithductal-dependent
systemicbloodflowwereassociatedwithahospitallengthofstayof19.6dayslongerthanthose
withductal-dependentpulmonarybloodflow(p<0.001)and2.9dayslongertoreachfullfeeds
thanthosewithductal-dependentpulmonarybloodflow(p<0.001),aftercontrollingfor
prematurity.Nasogastricfeedingroutepreoperativelywasassociatedwithalengthofstayof29.8
dayslongerthanthosefedbymouth(p<0.001)and2.4dayslongertoachievefullfeeds(p<
0.001),aftercontrollingforprematurityandcardiaclesion.Preoperativedietitselfwasnot
associatedwithsignificantchangeinlengthofstayordaystoreachfullfeeds.

5. ABSTRACT
Conclusions:Althoughcardiaclesionswithductal-dependentsystemicbloodfloware
consideredhighriskandmayincreaselengthofstayanddaystoachievefullfeeds,they
arenotassociatedwithahigherriskofpostoperativenecrotizingenterocolitis.Nasogastric
routeisnotassociatedwithasignificantlyhigherriskofnecrotizingenterocolitis,butlonger
lengthofstayanddaystoreachfullfeeds.Thesefindingschallengeourperioperative
managementstrategiesincaringfortheseinfants,astheymayincurmorehospitalcosts
andresourceswithoutsignificantmedicalbenefit.
Key words : congenital heart defect; ductus arteriosus; enteral feeding;infant; length of
stay; necrotizing enterocolitis

6. Lackofstandardizationinpreoperativeenteralnutritionforinfantswithcomplex
CHD,partlyduetounclearimpactonperioperativeoutcomes.Limitedevidence
on:whentostartfeeding,whattofeed,howtofeed.Feedingdecisionsmostly
basedonproviderdiscretion.Effectonkeypost-opoutcomes(NEC,hospital
LOS,daystofullfeeds)notwellstudied.Noconsensusonoptimaltimingof
preopfeeding.
INTRODUCTION
FearofNECfromsuspectedmesenterichypoperfusionhasledtosignificant
variationinpractice.Although,insomeretrospectivestudies:noclearincrease
inNECwithpreopfeeding.Humanmilkimprovesoutcomesinpreterm
neonates,butevidenceinCHD/vulnerableinfantsstillneeded.Aninternational
survey:mainreasonstowithholdpreopfeeds:concernforintestinal
hypoperfusion,umbilicalarterialcatheterinplace,prostaglandininfusion+
vasoactivesupport.

7. •Somecentersalloworalfeedingonlypreoptoavoidmasking“intestinalangina”.This
strategyexcludes:preterminfants,ventilatedinfants,infantsonCPAP.
•Lackoforalfeedingàdelayedoralskilldevelopment
•Absenceofanyenteralfeedingàintestinalatrophyàgutpermeabilityincreasesà
infectionriskincreasesàpoorpost-opgutmotility
•Earlypre-opGIstimulationmay:
-Improvepost-opfeed
tolerance/absorption
-Reducecomplications
-ShortenhospitalLOS
-Speedachievementoffullfeeds
INTRODUCTION
•Studyaim:assessassociation
betweenpre–opfeeding
exposures(feeding,typeof
feeding,andfeedingroute),and
post-opoutcomes(NEC,LOS,
anddaystofullfeeding)

8. MATERIAL & METHODS
Study type and setting
Retrospective cohort study
Infants with complex CHD
Texas Children’s Hospital
Study period: Jan 1, 2010 –Jan 1, 2016
IRB approved
Infants after 2016 excluded to avoid confounders (new NICU feeding protocol in
2017, new CICU with different admission protocols)
Reason for 2016 cutoff
2017: NICU feeding protocol introduced
2017: new CICU developed → could change which infants were admitted and how
they were managed

9. MATERIAL & METHODS
Exclusion Criteria
Admitted >72 hoursafter birth
Major non-cardiac anomalies affecting
gut/perfusion:
Heterotaxy
Omphalocele
Gastroschisis
Bowel atresia
Hirschsprung disease
Imperforate anus
Diaphragmatic hernia
Hypoxic-ischemic encephalopathy
Cardiac anomalies not requiring
intervention
Any type of cardiac intervention was
allowed (including transcatheter)
Inclusion Criteria
All gestational ages
Infants with these cardiac physiologies:
Biventricular lesion with ductal-dependent
pulmonary blood flow (dd-PBF)
Biventricular lesion with ductal-dependent
systemic blood flow (dd-SBF)
Single ventricle (SV)with dd-PBF
Single ventricle (SV)with dd-SBF
SV withoutductal-dependent PBF/SBF
d-transposition of the great arteries (d-
TGA)
Truncus arteriosus
Aortopulmonary window
Severe Ebsteinanomaly requiring PGE
TOF with absent pulmonary valve

10. MATERIAL & METHODS
Cardiac grouping
Lesions categorized by physiology
Echo + surgical records reviewed
Classified as:
SV vs biventricular
Ductal-dependent pulmonary vs ductal-dependent
systemic
d-TGA, truncus, aortopulmonary window → kept as
separate groups (different physiology)
Data Collection:
From pharmacy database: infants who received
prostaglandin E1
From echo database: infants with the listed CHD
lesions
Two lists were merged, duplicates removed
Each chart reviewed for inclusion
Data collected (from admission → discharge)
Demographics: race, sex, gestational age, birth
weight
Cardiac diagnosis
Postoperative NEC
Days from surgery to discharge (hospital LOS)
Days from surgery to full feeds
Preoperative feeding route
Type of feeds (incl. human milk)
Human milk = mother’s own milk, donor milk, or
combination
Preoperativelocation&generalfeedingpolicy
InfantsadmittedtoCICUorlevel4NICU
Feedsstartedonlyifclinicallystable
Decisionmadebymultidisciplinaryteam:cardiology,
neonatology,cardiacICU,neonataldietitian
Decisionsindividualized:whentostart,whattofeed,
volume,rateofadvancement,route
Usualpractice:feedpreopifhemodynamicallystable
andnotneedingrespiratorysupport,evenifon
prostaglandin

11. MATERIAL & METHODS
Feedingdetails(usualpractice)
Prefermother’sownmilkfirst
Ifunavailable→unfortifiedformulaordonorhumanmilk(basedonfamily/physicianchoice)
Startattrophic20mL/kg/day
Advanceby20mL/kg/day
Totalfluids:120–140mL/kg/day
Noformalinstitutionalfeedingguidelineatthattime
Maxenteralvolume=provider’sdiscretion
Feedsnotfortifieduntil:volume>100mL/kg/dayandweightgaininadequate
Feeding route selection (key point)
Route depended on clinical status and risk of intestinal angina
High-risk infants → oral feeds only
If low-risk but oral not possible (CPAP, mechanical ventilation, preterm) → NG tube feeds

12. MATERIAL & METHODS
Outcome definitions
NEC defined by Bell’s modified criteria
Each NEC case reviewed by 2 neonatologists + 1 radiologist
Disagreements → discussion → 3rd party if needed
If perforation → surgery/pathology used to distinguish NEC vs spontaneous perforation
Hospital LOS = days from surgery → discharge (not birth → discharge) to avoid bias from
variable time to surgery
Days to full feeds = days from surgery → 120 mL/kg/day enteral without TPN
Statistical analysis
Postop NEC: Fisher exact test
LOS and days to full feeds: Wilcoxon rank-sum
To adjust for prematurity:
NEC → logistic regression
LOS, days to full feeds → linear regression
Software: SAS 9.4 (SAS Institute, Cary, NC)

13. Results
StudySample
Total infants enrolled: 399
Demographics:
Male: 60%
Caucasian: 84%
Non-Hispanic: 65%
Mean gestational age: ~38 weeks
Cardiac lesion profile:
Ductal-dependent (overall): 71.7%
About 50%had ductal-dependent systemic blood flow (dd-
SBF)
Next most common: ductal-dependent pulmonary blood flow
(dd-PBF)

14. RESULTS
Hospital length of stay (LOS):
No LOS difference between infants fed vs
not fedpreop (p = 0.216)
Infants with dd-SBFstayed 14.6 days longer
than dd-PBF (p < 0.001)
Infants fed via nasogastric (NG) tubestayed
29.8 days longerthan those fed by mouth
(p < 0.001)
Human milk vs formula preop → no LOS
difference(p = 0.755)
Time to full feeds postop:
Preop fed: 8.2 ±5.6 days
Not preop fed: 8.9 ±6.3 days
Difference not significant(p = 0.079)
dd-SBF infants took 2.9 days longerthan dd-
PBF to reach full feeds (p < 0.001)
NG-fed infants took 2.4 days longerthan
orally fed to reach full feeds (p < 0.001)
Preop diet type(human milk vs formula) →
no effecton time to full feeds (p = 0.490)
Postoperative NEC:
Total NEC cases: 32 infants (8%)
Stage I NEC: 7/32 (22%)
All NEC cases occurred in dd-SBF or dd-
PBFlesions
More NEC in dd-SBF than dd-PBF, but not
statistically significant
Preoperative feeding:
Fed preoperatively: 62%
Of those fed: 44%got human milk
Of those fed: 66%fed by mouth
Effect of preop feeding on NEC:
No significant difference in postop NEC
between fed vs not fed
OR = 0.58(95% CI 0.29–1.19), p = 0.183
(after adjusting for prematurity)
Preop diet typeand feeding routealso not
significantly associated with NEC

15. RESULTS

16. RESULTS

17. DISCUSSION
Mainfindings:
Lesionswithductal-dependentsystemicbloodflow(dd-SBF)didnothaveasignificantlyhigherriskof
postoperativeNECthandd-PBFafteradjustingforprematurity.
But:dd-SBFlesionswerelinkedtolongerhospitalstaysandlongertimetofullfeeds.
Preoperativenasogastric(NG)feedingwasindependentlyassociatedwithlongerLOSandlongertimetofull
feeds(evenafteradjustingforprematurity).
Typeofpreopdiet(humanmilkvsformula)wasnotlinkedtobetter/worseoutcomes.
Contextfromliterature:
Ductal-dependentCHDbabiesaregenerallyconsideredhigherriskforNEC,especiallydd-SBF,duetolowersystemic
perfusion.
ButmanyotherfactorsmayinfluenceNEC:prematurity,illnessseverity,feedingregimen,bypasstime,postopcomplications.
Interpretation:
These“high-risk”lesions(esp.dd-SBF)→tendtohavelongerLOSandlongertimetofullfeeds,even
aftercontrollingforprematurity→suggestspostoperativefactorsareimportant.
Postopfactorsthatmaycontribute:prolongedventilation,infections,ECMO,neurologiccomplications→
needtobestudiedmore.

18. DISCUSSION
Onpreopfeeding&NEC:
SimplyfeedingvsnotfeedingpreopwasnotclearlyassociatedwithpostopNEC.
Thissupportsearlierstudiesthatthetheoreticalriskoffeed-relatedNECdoesn’talwaysmatchobservedrisk.
Priorworkshowedpreophumanmilk↓preopNEC,butinthisstudypreophumanmilkdidnotreducepostopNEC
→postopfeedingmaymattermore.
Possibledoseeffectofhumanmilkwasnotevaluated.
On oral vs NG feeding:
Oral feeding in fragile infants is known to help oral skills, shorten LOS, and improve later development (supported by
Cochrane + CHD studies).
In this study, preop NG feeding(vs oral) → longer LOSand longer time to full feeds.
Likely mechanism: no oral stimulation→ baby can’t develop suck/oral skills during a critical window.
They didn’t check how many still needed a tube at discharge, but delayed oral readiness is plausible.
Important nuance:
NG tube use may be a marker of sicker babies(preterm, on ventilator, CICU) → so NG may reflect illness severity
and provider caution, not just the route itself.
This shows the need for standardized preop feeding protocolsto reduce practice variation.

19. DISCUSSION
Clinical implication:
Fear of “intestinal angina” with tube feeding may be overestimated.
Feeding route did notincrease NEC risk, but NG route was linked to longer stay→ so if gut
risk is low, oral feeding may be preferable, but NG is still acceptable and likely safe re: NEC.
Limitations:
Single-center, retrospective→ limited generalizability.
Majority were dd-SBF→ may bias results.
NEC numbers small → possible type II error(i.e. missed a real difference).
Postop factors (ventilation, infections, ECMO, neuro problems) not fully controlled.
Randomized trial is unlikely; a prospective, standardizedstudy from birth → discharge is needed.

20. CONCLUSION
PreoperativemanagementininfantswithcomplexCHDmattersforpostoperativeoutcomes.
Eventhoughinfantswithdd-SBFareconsideredhighriskandoftenmanagedvery
conservatively,someoftheseconservativestrategiesaren’twellsupportedbydata.This
studydidnotfindanincreasedriskofpostoperativeNECininfantswho:werefed
preoperatively,orwerefedvianasogastric(NG)tube.However,infantsfedviaNGtubehad:
longerhospitallengthofstay(LOS),andmoredaystoreachfullfeeds.Therefore,when
decidinghowtofeedpreoperatively,cliniciansshouldalsothinkabout:neurodevelopmental
impact(oralskills,feedingreadiness),hospitalcosts(longerLOS).Thestudysupportsthe
ideathatstandardizedperioperativefeeding/managementprotocolscouldreducepractice
variationandimproveoutcomesinthishigh-riskgroup.

21. Population
Infants (all gestational ages) with complex congenital heart disease (CHD)admitted to Texas Children’s Hospital
Time: Jan 1, 2010 –Jan 1, 2016
All required cardiac intervention before discharge(surgery or transcatheter)
Major physiologies included:
Ductal-dependent systemicblood flow (dd-SBF)
Ductal-dependent pulmonaryblood flow (dd-PBF)
Single ventricle variants
d-TGA, truncus, aortopulmonary window, severe Ebstein on PGE, TOF absent PV
Excluded: >72 h at admission, major GI anomalies (omphalocele, gastroschisis, atresia, Hirschsprung, imperforate anus), heterotaxy,
HIE, CHD not needing intervention.
Intervention / Exposure
Preoperative enteral feeding exposures:
Fed vs not fed preoperatively
Type of feed: human milk vs other (formula/unfortified)
Feeding route: oral vs nasogastric (NG) tube
Cardiac lesion type (esp. dd-SBF) as a clinical exposure
Feeding decisions were not protocolized —based on provider + multidisciplinary team.
CRITICAL APPRAISAL

22. Comparison
Fed preop vsnot fed preop
Human milk vsnon–human milk
Oral feeds vsNG tube
dd-SBF vsdd-PBF lesions
All analyses adjusted for prematurity.
Outcomes
Primary/major outcomes:
Postoperative NEC (Bell’s modified, all stages)
Hospital length of stay (LOS) → surgery → discharge
Days to achieve full feeds → surgery → 120 mL/kg/day enteral without TPN
Key results:
Preop feeding (yes/no): not associated with ↑ postop NEC (OR 0.58; p=0.18)
Diet type (human milk vs formula): not associated with postop NEC or shorter LOS
NG route preop → longer LOS (+~30 days) and longer days to full feeds (+~2.4 d) vs oral, even after adjustment
dd-SBF lesions → longer LOS (+~14–20 d) and longer time to full feeds (+2.9 d) vs dd-PBF, but not more NEC than
dd-PBF.
CRITICAL APPRAISAL

23. Validity (Internal)
Strengths:
Clear inclusion/exclusion criteria
All NEC cases adjudicated by 2 neonatologists + radiologist
Adjusted analyses for an important confounder (prematurity)
Outcomes objectively defined (LOS, d-to-full feeds)
Limitations:
Retrospective, single center→ risk of selection bias
Feeding not randomized → route (NG) may be a proxy for sicker babies (ventilated, CICU) → residual confounding
Small number of NEC events (8%) → low power → possible type II error for NEC, esp. dd-SBF vs dd-PBF
Preop dose of human milk not quantified → can’t test dose–response
Postop factors (infection, ventilation days, ECMO) not fully controlled → may drive LOS.
Importance/Impact
Clinicallyimportantbecausemanycenterswithholdfeedsinductal-dependentCHDoutoffearofNEC.
Thisstudyshows:feedingitself(incl.NG)wasnotlinkedtomorepostopNEC,butwaslinkedto
longerLOS→sothepenaltyismoreresource/developmental,notsafety/NEC.
Highlightsthatlesionphysiology(dd-SBF)isamajordriverofslowerpostoprecovery.
CRITICAL APPRAISAL

24. Applicability
Applicable to: large cardiac centerscaring for neonates with ductal-dependent CHD pre-surgery, especially places
still using “NPO because of NEC fear” strategy.
Caution in applying to:
Centerswith already standardized enteral CHD protocols (results may differ)
Infants with major GI/surgical anomalies (excluded here)
Settings with different postop practices (ECMO, ventilation, timing of surgery)
Practical takeaway:
It is reasonable to start cautious preop feeds in hemodynamically stable CHD infants, even on PGE, because NEC
signal is low.
But: prefer oral when possible (neurodevelopment, shorter LOS); use NG when oral not feasible, but recognize it
may lengthen stay —partly illness-severity related.
Best next step for a unit: make a standardized preop feeding protocol to reduce provider-to-provider variation.
CRITICAL APPRAISAL

25. THANK YOU
Please advise