Outcome of Infants with Unrepaired Heart Disease Admitted tothe Pediatric Intensive Care Unit: Single-center DevelopingCountry Perspective

Shazia Samad Mohsin, MB, BS, DCH,* Anwarul Haque, MD,* Abdul Sattar Shaikh, MB, BS,*Surraiya Bano, MB, BS,† and Babar Sultan Hasan, MD*

Departments of *Pediatric and Child Health and †Emergency Medicine, The Aga Khan University, Karachi, Pakistan

A B S T R A C T

Objective. Congenital heart disease (CHD) has an incidence of ~0.8–1%. Outcome of previously diagnosed CHDpatients awaiting surgery (either correction or palliation) in a developing country setting is unknown. We strive todetermine the outcome of patients with CHD awaiting surgery who present to pediatric intensive care unit (PICU)setting with an acute illness.Design. Retrospective cross-sectional chart review.Setting. Pediatric intensive care unit of The Aga Khan University Hospital, Karachi, Pakistan.Patient. Medical records of infants (1–12 months) with CHD awaiting surgery presenting to the PICU with an acuteillness between January 2009 and June 2012 were included. Newly diagnosed CHD patients, those not requiringPICU admission, and those transferred to another hospital were excluded.Results. A total of 34 infants met the inclusion criteria. Median age at presentation was 5 months. Seventy-fourpercent of the infants had CHD lesion characterized by increased pulmonary blood flow (shunt lesions). Thoughnone of the patients met the strict criteria for sepsis or pneumonia, 74% were admitted with a diagnosis ofpneumonia or sepsis. Only 15% of patient had congestive heart failure as an admitting diagnosis. Oxygen therapywas given to 94% of these patients. Fifty-nine percent of these patients expired during the admission, 95% of thoseexpired had multiorgan dysfunction.Conclusion. Patients with CHD awaiting surgery and who admitted to the PICU with acute illness are at high riskfor mortality. Stringent criteria to diagnose pneumonia or sepsis should be used in these patients.

Key Words: Infants; Unrepaired Congenital Heart Disease; Mortality

Introduction

With an incidence of 1%, congenital heartdisease (CHD) is the most common con-

genital anomaly in pediatrics.1 In developed coun-tries, significant advances in diagnosis and earlysurgical management of infants with CHD haveled to a significant decrease in morbidity and mor-tality.2 There has been a 60% decrease in mortalitybetween 1995 and 2005.3 In fact, populationprevalence studies done in western countries haveshown that the number of adults with CHDexceeds those of infants with CHD.4 On the con-trary, CHD contributes significantly to infantmorbidity and mortality and poses an increasingburden on health resources of developing coun-tries.5 Late presentations, associated comorbidi-

ties, and late repair are some of the contributingfactors leading to this suboptimal outcome inthese countries.6 Infants diagnosed with CHDawaiting surgery (either correction or palliation), dueto either lack of resources or expertise for earlierrepair, may be at a high risk of death.7 Data on theoutcome of such patients, especially when present-ing with an acute illness to an intensive caresetting, are unknown. We strive to determine theoutcome of patients with CHD awaiting surgerywho present to pediatric intensive care unit(PICU) setting with an acute illness.

Methods

This is a retrospective single-center cohort study.Infants (1–12 months) with CHD awaiting surgerywere included if they were admitted to PICU at AgaFunding: None.

116

© 2013 Wiley Periodicals, Inc.Congenit Heart Dis. 2014;9:116–121

Khan University Hospital (AKUH) betweenJanuary 2009 and June 2012. Aga Khan UniversityHospital is a private tertiary care hospital and theonly hospital in the province of Sindh providingneonatal and pediatric congenital heart care. Thepopulation of Sindh is ~35 million (25 million inurban cities and 10 million in rural setting) and thusAKUH patient presentation is from both rural andurban settings. With one cardiothoracic surgeonand three pediatric cardiologists, AKUH performs300–320 surgeries per year. For total correction ofan acyanotic lesion, a minimum of 5-kg weight wasa requirement at our center. Recently (by Septem-ber 2011), the minimum weight requirement wasdropped to 2.5 kg for total correction of acyanoticlesions. Detailed records of patients presenting tothe emergency room (ER) and PICU were notavailable prior to January 2009. Institutional reviewboard approval was obtained. Newly diagnosed orrepaired CHD patients, those who did not requirePICU admission, and those who were transferredto another PICU were excluded. Using the Ameri-can Academy guidelines,8 patients requiring fre-quent observation, cardiorespiratory support in the

form of positive pressure (invasive and noninva-sive), and inotropic support were admitted to thePICU. Information was gathered from patient’smedical records and PICU database. The admit-ting diagnosis determined by the ER physicians wasrecorded off the medical chart. Stringent diagnos-tic criteria and clinical definitions (Table 1) werethen used for the purpose of this study to assign adiagnosis. Mortality, presence of multiorgan dys-function (Table 1), requirement of more than twoinotropes, systemic hypotension, surgical treat-ment during the same admission, and time intervalbetween diagnosis and last clinic visit to the PICUadmission were recorded. Data were entered andanalyzed using SPSS version 19.0 (IBM Corpora-tion, Armonk, NY, USA). Mann–Whitney U testfor continuous variables and chi-square or Fisher’sexact tests were used for categorical variables. A Pvalue of <.05 was used for significance.

Results

Eighty-eight patients with CHD awaiting surgeryhad a total of 111 ER admissions during the study

Table 1. Clinical Definitions of Diagnostic Criteria for this Study

Variable Study Definition

Fever Temperature >38°C8

Sepsis (with or withoutpositive blood culture)

Presence of two or more of the criteriaFeverTachycardia if heart rate >2 SD for ageRespiratory rate >2 SD of the age limitTotal leukocyte count elevated or depressed for ageRaised CRP8

Pneumonia Diagnosis requires at least two clinical findings plus fever and tachypnea and laboratory and radiographicevidence

Clinical findings:CoughNew onset of lower respiratory tract secretionsChange in character of secretionsIncrease in the quantity of secretionsSuctioning requirementsAuscultatory findings of pneumonia or consolidation (rales, bronchial breath sounds, egophony,

decreased breath sounds)Dyspnea (or appearance of being “air hungry”)Hypoxemia (PO2 60 mm Hg in room air)

Vital signsFeverTachypnea (defined by age group)

Laboratory evidenceWBC > 15 000 and 10% bands or WBC < 4 000

Radiographic evidenceWithin 48 h before institution of therapy, the chest radiograph should show the presence of a new

infiltrate(s) consistent with infection (interstitial, bronchial, alveolar), consolidation, cavitation, abscess,or pneumatocele9

Congestive heart failurewith pulmonary edema

Presence of tachycardia, tachypnea, lung crepitation, gallop, hepatomegaly, and chest x-ray showingcardiomegaly and pulmonary plethora

Multiorgan dysfunctionsyndrome

Presence of two or more organ dysfunction (respiratory, renal, neurologic, hematologic, or hepatic)8

Systemic hypotension Systolic or diastolic blood pressures <2 SD for age8

CRP, C reactive protein; WBC, white blood cell.

Congenit Heart Dis. 2014;9:116–121

Unrepaired Heart Disease in Infants 117

period. Acyanotic heart disease was present in 75(85%) of the patients. Ninety-six (86%) of the ERadmissions were due to cardiorespiratory-relatedsymptoms. Other reasons for ER presentationincluded acute gastroenteritis in 10 (9%), upperrespiratory tract infection in three (3%), and gas-troesophageal reflux and measles in one (1%) each.A total of 34 (39%) of these patients were admittedto the PICU and were included in this study. Thedemographic of patients admitted to the PICU isshown in Table 2. The majority of 25 (74%)patients belonged to the group of CHD withincreased pulmonary blood flow. Out of these, 14lesions were simple defects, i.e., atrial septaldefect, ventricular septal defect (VSD), and a com-bination of VSD with atrial septal defector a patent ductus arteriosus or VSD with coarc-tation, while 11 were complex acyanotic defects,i.e., unobstructed total anomalous venous return,truncus arteriosus, complete atrioventricularcanal defect, and double outlet right ventricle(Figure 1). The diagnosis given at the time ofadmission was sepsis or pneumonia in 25 (74%)and congestive heart failure due to CHD in fivepatients (15%) only. Using strict definitions(Table 1), none of these 25 patients fulfilled thecriteria of sepsis or pneumonia; however, five(20%) had a positive blood culture. Oxygentherapy in the ER was given to 32 (94%) of thepatients including those who had a left to rightshunt and increased pulmonary blood flow. Theprimary reason of this therapy was a clinical diag-

nosis of pneumonia or sepsis. Of the patient whoadmitted to the PICU, 20 (59%) expired. Thepredominant cause of death was cardiorespiratoryfailure due to multiorgan dysfunction. Thirty-twopatients required mechanical ventilation in PICUand median time period from admission to start ofmechanical ventilation was 4 (1, 48) hours. Whencompared with those who survived in the PICU,those who expired received more than two ino-tropes (P < .001), 18 (90%) had multiorgan dys-function (P = .003) and 17 (85%) had systemichypotension (P = .002) (Figure 2). There was nodifference seen for patient’s age, weight, and timeinterval between first diagnosis and last clinic visitto PICU admission for those who survived com-pared with those who expired. Surgical correctionwas done in 18 (50%) patients. A total of 18patients were taken for surgery during the sameadmission, nine (50%) of these patients expired.Complete correction was attempted in seven(39%) and palliation with a pulmonary arterybanding was performed in seven (39%) patients,while four (22%) of the cyanotic patients receiveda Blalock–Taussig shunt (Table 2). Of those whoexpired, two had simple acyanotic lesions and sixhad complex acyanotic lesions, while one hadcyanotic heart disease. Of the remaining 16patients, who were treated conservatively, 11(68%) expired (four with simple, five with complexacyanotic lesions, and two with cyanotic lesions).Surgery in the same admission did not affect theprimary outcome (P = .66).

Discussion

The majority of the CHD patients awaiting surgeryin our cohort had an acyanotic left to right shuntlesion. When presented to the ER with an acuteillness, 39% of them were sick enough to beadmitted to the PICU. Patients with CHD whoadmitted to the PICU with acute illness are at highrisk for mortality in this single-center experiencefrom an underdeveloped country. Though sepsisand pneumonia were the leading diagnosis andoxygen therapy was administered to the majorityof these patients, none met the strict sepsis criteriaas described by International Pediatric SepsisConsensus Conference.9,10

Among the reported 1% incidence of all CHD,500–600/100 000 live births will need some kindof surgical or catheter-based intervention for theirheart disease.1 Early recognition and treatment ofCHD are important because clinical presentationand deterioration may be sudden and some cor-

Table 2. Demographics of Patients with Previously Diag-nosed Congenital Heart Disease Awaiting Surgery Admittedto the Pediatric Intensive Care Unit (n = 34)

*Median Age at Presentation (in months) 5 (1, 12)Median weight at presentation (in kilograms) 3.5 (2, 8)Median z-score for weight at presentation -4.1 (-9.6, -0.93)Median PICU stay (in days) 14 (3, 50)Median time of death after surgery

(postoperative day in days)10 (2,15)

Total surgeries performed 18Palliation (four expired) 11 (61%)

Blalock–Taussig shunt 4Pulmonary artery banding 7

Total correction (five expired) 7 (39%)VSD closure 2PDA and COA repair 1TAPVR repair 4

Median time period from diagnosis to PICUadmission (in months)

2 (0.4, 12)

Median time period from last clinic visit toPICU admission (in months)

1 (0.4, 3)

Mortality 20 (59%)

*Continuous data are presented as median (minimum, maximum values).COA, coarctation; VSD, ventricular septal defect; PDA, patent ductus arterio-sus; TAPVR, total anomalous pulmonary venous return; PICU, pediatric inten-sive care unit.

Mohsin et al.118

Congenit Heart Dis. 2014;9:116–121

rectable defects may lead to mortality.11 Patientswith early appropriate management haveimproved survival due to stable preoperativestate.12,13 Rapid advances have taken place in devel-oped countries in the past six decades and over95% infants born with critical heart disease cansurvive infancy with early diagnosis and timelymanagement.14,15 CHD still pose an increasingburden on health resources of many developingcountries. Nepal has reported 50% mortality andSri Lanka and India 20% each in patients withCHD.16,17 Heart failure due to unrepaired CHD,especially acyanotic shunt lesions, may mimicsymptoms of pneumonia and sepsis, leading to

incorrect diagnosis and sometimes mismanage-ment increasing their morbidity and mortality.18

Many physicians in the developing world wouldfavor medical management over early surgical cor-rection (especially of acyanotic shunt lesions) toavoid subjecting these critically sick infants to asurgical procedure in the presence of presumedinfection and immune suppression.19 Other factorsfor delay in correction may be lack of awarenessregarding the surgical expertise available in thecountry and the misconception that 10-kg weightis necessary to undergo repair.20 Bhatt et al. sug-gested that attempts to resolve preoperative respi-ratory infection completely would not be possible

Figure 1. Spectrum of acyanotic lesions presenting to PICU. VSD, ventricular septal defect; ASD, atrial septal defect; PDA,patent ductus arteiosus; CAVSD, complete atrioventricular septal defect; TAPVR, total anomalous pulmonary venous return;TA, truncus arteriosus; COA, coarctation of aorta; DORV, double outlet right ventricle; PICU, pediatric intensive care unit.

Figure 2. Differences in frequency of multiorgan dysfunction, usage of �2 inotropes and systemic hypotension in patientswith CHD awaiting surgery who expired in the PICU vs. those who survived in the PICU. MOD, multiorgan dysfunction; PICU,pediatric intensive care unit; CHD, congenital heart disease.

Congenit Heart Dis. 2014;9:116–121

Unrepaired Heart Disease in Infants 119

in the presence of significant increased pulmonaryblood flow.21 Patients with CHD awaiting surgeryadmitted to the PICU in our cohort had highmorbidity and mortality depicting a very high riskin these patients. We suggest that early repair maybe more beneficial than conservative medicaltherapy when surgical expertise is available. Afaqet al.22 also recommend early repair over conser-vative management in their study.

Though the majority of our patients wereadmitted with a diagnosis of sepsis or pneumonia,none fulfilled the strict criteria as described byInternational Pediatric Sepsis Consensus Confer-ence.9 Such diagnosis may delay surgical correc-tion and may even lead to mismanagementespecially in pulmonary vascular resistance-dependent left to right shunt lesions. Ninety-fivepercent of infants were provided oxygen therapy inthe ER. Oxygen in the presence of left to rightshunt lesions can worsen the clinical scenario ofheart failure by decreasing the pulmonary vascularresistance and subsequently further increasingthe pulmonary blood flow.23 We propose the usageof strict criteria prior to diagnosing pneumonia orsepsis in these patients. Once diagnosed withpneumonia, oxygen therapy should be very cau-tiously used with clear goals for weaning.

Our results also showed that surgical repairafter PICU admission did not affect the primaryoutcome (P = .6). Forty-five percent of the mor-tality group patients underwent surgical correc-tion at the same admission, but the unstablepreoperative state did not improve the outcome.

Limitations

As we looked at patients who presented to a single-center ER, there may be a selection bias. Patientswho were in more significant heart failure andrelatively more unstable presented to the ER thusfalsely magnifying the morbidity and mortality.On the other hand, we may be underestimatingthe true problem as many of the CHD patientsawaiting surgery may have presented to other hos-pitals and thus not included in this study. Being aretrospective cross-sectional study, one cannotaccount for hidden confounders like socioeco-nomic condition, comorbid conditions likeimmune deficiency, or other congenital anomaliesleading to an increased mortality. As there was nodifference in the weight of patients who expired vs.survived, we assume that at presentation they hada similar state of well-being.

Conclusion

Patients with CHD awaiting surgery and whoadmitted to the PICU with acute illness are at highrisk for mortality. Stringent criteria to diagnosepneumonia or sepsis should be used in thesepatients.

Acknowledgement

We like to thank Farheen Ayub for her technical assistancein making the figures for the article.

Authors Contribution

Shazia Samad Mohsin contributed in study design, dataanalysis, drafting the article, revising the article, andfinal approval. Anwarul Haque was involved in studydesign, data interpretation, statistics, revising thearticle, and final approval. Abdul Sattar Shaikh contrib-uted in data collection, data analysis, drafting, revising,and final approval of the article. Surraiya Bano wasinvolved in data collection, data interpretation, articlerevision, and its final approval. Babar S. Hasan contrib-uted in study concept, data analysis, revising the article,critical analysis, and final approval.

Corresponding Author: Babar Hasan, MD, FAAP,Assistant Professor of Pediatrics, Consultant PediatricCardiology, Department of Pediatric and Child Health,The Aga Khan University, Stadium Road, P.O. Box3500, Karachi 74800, Pakistan. Tel: (+92) 3362559473;Fax (+92) (21) 4934294; E-mail: babar.hasan@aku.edu

Conflict of interest: None.

Accepted in final form: March 5, 2013.

References

1 Hoffman J, Kaplan S. The incidence of congenitalheart disease. J Am Coll Cardiol. 2002;39:1890–1900.

2 Khairy P, Ionescu-Ittu R, Mackie AS, Abrahamow-icz M, Pilote L, Marelli AJ. Changing mortality incongenital heart disease. J Am Coll Cardiol.2010;56:1149–1157.

3 Federal Interagency Forum on child and family sta-tistics America’s children: key national indicators ofwell being, 2009. Washington DC: U.S. Govern-ment printing office 2009.

4 Marelli AJ, Mackie AS, Ionescu-Ittu R, Rahme E,Pilote L. Congenital heart disease in general popu-lation: changing prevalence and age distribution.Circulation. 2007;115:163–172.

5 Saxena A. Congenital cardiac surgery in the lessprivileged regions of the world. Expert Rev Cardio-vasc Ther. 2009;7:1621–1629.

Mohsin et al.120

Congenit Heart Dis. 2014;9:116–121

6 Dhandayuthapani G, Chakrabati S, Ranasinghe A,et al. Short term outcome of infants presenting topediatric intensive care unit with new cardiac diag-noses. Congenit Heart Dis. 2010;5:444–449.

7 Pfammatter JP, Stocker FP. Delayed recognition ofhemodynamically relevant congenital heart disease.Eur J Pediatr. 2001;160:231–234.

8 Society of Critical Care Medicine, PediatricSection, Admission Task Force. Guidelines fordeveloping admission and discharge policies forthe pediatric intensive care unit. Pediatrics. 1999;103:840–842.

9 Goldstein B, Giroiv B, Randulph A. InternationalConsensus Conference on Pediatric Sepsis. Interna-tional Pediatric Sepsis Consensus Conference:definitions for sepsis and organ dysfunction in pedi-atrics. Pediatr Crit Care Med. 2005;6:2–8.

10 Langley GM, Bradley JS. Defining pneumonia incritically ill patients and children. Pediatr Crit CareMed. 2005;6:S9–S13.

11 Brown KL, Ridout BA, Hoskote A, Verholst L,Ricci M, Bull C. Delayed diagnosis of congenitalheart disease worsens preoperative condition andoutcome of surgery in neonates. Heart. 2006;92:1298–1302.

12 Masmoudi H, Massin MM. Importance of timelydiagnosis of congenital heart disease. Pediatr Health.2008;2:47–52.

13 William WG. Surgical outcomes in congenital heartdisease: expectations and realities. Eur J CardiothoracSurg. 2005;27:937–944.

14 Moons P, Sluysmans T, De Wolf D, et al. Congeni-tal heart disease in 111225 births in Belgium: birth,prevalence, treatment and survival in the 21st

century. Acta Paediatr. 2009;98:472–477.

15 Bernier PL, Stefanescu A, Samoukovic G, Tcher-venkov IC. The challenge of congenital heartdisease worldwide: epidemiologic and demographicfacts. Semin Thorac Cardiovasc Surg Pediatr Card SurgAnnu. 2010;13:26–34.

16 Shah G, Singh M, Pandey T, Kalakheti BK, Bhan-dari GP. Incidence of congenital heart disease intertiary care hospital. Kathmandu Univ Med J(KUMJ). 2008;6:33–36.

17 Wickramasingrihe P, Lamabadusuriya S, Naren-therin S. Prospective study of congenital heartdisease in children. Ceylon Med J. 2001;46:96–98.

18 Kogon B, Butler H, Kirshbom P, Kanter K, McCo-nnell M. Closure of symptomatic ventricular septaldefects: how early is too early? Pediatr Cardiol.2008;29:36–39.

19 Hewitson J, Brink J, Zilla P. The challenge of pedi-atric cardiac surgeries in the developing world.Semin Thorac Cardiovasc Surg. 2002;14:340–345.

20 Saxena A. Congenital heart disease in India: a statusreport. Indian J Pediatr. 2005;72:595–598.

21 Bhatt M, Rotu SJ, Kumar RK, et al. Management ofinfants with large ventricular septal defects and res-piratory infections requiring mechanical ventilation.J Thorac Cardiovasc Surg. 2004;127:1466–1473.

22 Afaq A, Zia HA, Amanullah MM. Is early correctionof congenital ventricular septal defect a betteroption in developing country. J Pak Med Assoc.2012;60:324–327.

23 Woods WA, McCulloch MA. Cardiovascular emer-gencies in the pediatric patient. Emerg Med ClinNorth Am. 2005;23:1233–1249.

Congenit Heart Dis. 2014;9:116–121

Unrepaired Heart Disease in Infants 121