There is limited information about neurodevelop-mental outcomes after surgery for CHD from low- and middle-income countries (LMICs). Of late, many centres from LMICs have started performing corrective surgical procedures in neonates and infants with CHD, with outcomes comparable to those from developed nations [7,8]. However, the demographic profile and pre-operative characteristics are very different in LMICs and these could significantly impact long-term outcomes [7-12]. An analysis of Indian children with uncorrected CHD has shown that they are at increased risk of developmental delay [13]. These can have a significant impact on neurodevelopmental outcomes despite high- quality surgical expertise and post-operative intensive care.

This prospective study from a tertiary-care pediatric cardiac center examines the short term neuro-developmental outcomes in infants undergoing corrective surgery for CHD.

Cardiac diagnosis with the pre-operative risk score based on the Risk Adjustment for Congenital Heart Surgery 1 (RACHS-1) classification was recorded. Details of prenatal diagnosis, delivery and mode of transport to the cardiac center (monitored/unmonitored) were recorded. Duration of ventilation and prostaglandin (PGE1) infusion (hours), ICU and hospital stay (days), hematocrit, sepsis (culture positive/clinical), oxygen saturation and nature of the surgical procedure (planned/urgent) were recorded.

Duration of cardiopulmonary bypass (CPB), aortic cross clamp (ACC) and total circulatory arrest (minutes), mean perfusion pressure during CPB (mm Hg), minimum hematocrit, lowest temperature, modified/continuous ultrafiltration, major hypoxic events (PaO2 <55 mm Hg), cardiac arrest, arrhythmia with hemodynamic compromise, air embolism and need for re-institution of CPB were noted. The CPB flow rates and mode of ultra-filtration were based on surgical preferences. Direct cerebral vascular monitoring was not done.

Duration of mechanical ventilation and inotropic support (hours), ICU and hospital stay (days) were noted. Delayed sternal closure, re-intubations, pulmonary hypertension (PH) crisis, cardiac arrests requiring resuscitation, focal neurological deficits, post-operative sepsis and seizures were noted. Prolonged ventilation, and prolonged ICU stay were defined as >48 hours and >7 days, respectively.

Statistical analysis: Student’s t test was used to compare continuous risk factors by normal and delayed group. Chi-square test was used to find the association between demographic and perioperative categorical factors by PDI and MDI (delay) category. One way ANOVA was used to compare the PDI and MDI by RACHS category. Multivariate binary logistic regression analysis (stepwise) was used to estimate the Odds Ratio with 95% Confidence Intervel and adjusting for potential confounding variables. The cut-off point for statistical significance was set an alpha level of 5%. Statistical analysis was done using IBM SPSS 20.0 (SPSS Inc, Chicago, USA).

Of the 162 infants (92 males) included, 52 (32.1%) were neonates. There was no in-hospital mortality. Four patients (2.4%) died on follow-up; 3 had residual cardiac issues, and in one cause of death was unknown (Fig. 1). RACHS category 1 and 2 constituted 59.3% (96) of the surgical procedures (Web Table I). Prenatal diagnosis was made in 4.3% of patients.

Fig. 1 Flow of participants in the study.

The distribution of acyanotic (80, 49.4%) and cyanotic CHD were similar. Preoperative weight, height, head circumference were abnormal in 110 (67.9%), 48 (29.6%) and 60 (37%) children, respectively. Preoperative ICU care was needed in 56 (36.4%) with a median stay of 4 (1-30) days; 15 (9.3%) had pre-operative sepsis. The median age at surgery was 60 days (2-365).

Major intraoperative complications occurred in 10 infants (6.2%). One had a major hypoxic event while 9 patients (5.5%) had arrhythmias with hemodynamic compromise. Re-institution of CPB was required in 7 infants (4.3%). The median CPB and ACC time were 109 min (41-457) and 58 min (10-274), respectively (Web Table I). The mean (SD) hematocrit and perfusion pressure on CPB was 28.4 (3.11) %and 38.4 (4.11) mm Hg. The median duration of postoperative ventilation and inotrope was 42 (3 - 552) and 72 (2 - 432) hours, respectively. Delayed sternal closure was done in 8 infants (4.9%). Mean (SD) postoperative ICU stay was 7.3 (5.41) days. Postoperative sepsis occurred in 32 infants (19.8%). Re-intubation was required in 15 (9.3%) infants.

Anthropometric indices improved on one-year follow-up with abnormal weight for age and weight/height Z scores in 23% (35) and 16.4% (25), respectively. Height-for-age and head circumference Z scores were abnormal in 27.6% (42) and 23% (35), respectively.

At three months after corrective surgery, delayed PDI was seen in 53 (33.5%) and mean (SD) PDI score was 81.2 (33.02); this improved to 92.4 (26.02) at one-year follow-up (22, 14.5% delayed). On univariate analysis, factors associated with delayed PDI at one-year follow-up included gestational age <39 weeks, weight and head circumference Z-score <–2 on follow-up, postoperative ventilation >48 hours, postoperative ICU stay >7 days and longer duration of postoperative inotrope use. Lower mean PDI score was associated with weight (P=0.02) and head circumference Z-score < -2 (P=0.009) at one year follow up.

At three months after corrective surgery, delayed MDI was seen in 31 (19.6%) and mean (SD) MDI score was 90.0 (27.17); this improved to 96.1 (26.05) at one-year follow-up (22, 14.5% delayed). On univariate analysis, factors associated with lower mean MDI scores at one-year follow-up were gestational age <39 weeks (P<0.001) and preoperative weight Z-score <-2 (P=0.03).

On multivariate regression analysis, the variables associated with delayed PDI outcome at one year follow-up were gestational age <39 weeks and one year postoperative head circumference Z score <-2 (Table I). Delayed MDI at one year was associated with higher mean perfusion pressure on CPB (P = 0.05).

TABLE I	Multivariate Regression Analysis of Predictors of Delayed PDI and MDI

We analyzed the short-term neurodevelopmental outcomes in Indian infants undergoing corrective surgery for CHD. Though the PDI and MDI scores improved on one-year follow-up, 14.5% of patients continued to have delay. The motor skills was more affected than the mental skills and cognitive function at both points of time. Patient-specific constitutional factors like early term birth and head circumference on follow-up were associated with neurodevelopmental outcomes rather than peri-operative factors or cardiac diagnosis.

Formal developmental assessment by DASII was not done pre-operatively and it is possible that subtle neurological abnormalities which could have impacted outcomes, and pre-existing developmental delay were missed. We did not have a cohort of non-CHD or unoperated infants for comparison of developmental indices.

The patterns of developmental delay are in accordance with findings of previous studies affecting psychomotor domains more than mental milestones [3,6,13,17]. Several studies have stated that patient-specific factors are important determinants of neurodevelopmental outcome compared to peri-operative factors [2,3,18]. Recent studies have reported the impact of gestational age at delivery on outcomes of neonatal heart surgery [19]. Our data suggests better neurodevelopmental outcomes for infants born at full term compared to early term deliveries [5]. There is a common practice to plan a early (<39 weeks) delivery once a prenatal diagnosis of critical CHD is made to facilitate expedited postnatal cardiac care [20]. Delivery before full term may have a greater impact on brain development and neurodevelopmental outcomes [21,22]. Postnatal studies in term neonates with complex CHD have shown smaller head circumferences compared with normal term neonates [21-23]. Possible reasons attributed to this include cerebral hypo-oxygentation, shared genetic or environmental or placental factors [23]. It is possible that some of these factors may persist despite surgical correction of the CHD, thus influencing neurodevelopmental outcomes. Studies have shown that head circumference at birth is a predictor of head circumference at 1 year of age, thereby explaining our observation of lower head circumference at 1-year follow-up predicting lower PDI [3,24]. Cerebral perfusion pressure, mean cerebral blood flow velocity and regional cerebral oxygen saturation index (rSO2i) would have been better indicators of the adequacy of cerebral blood flow on CPB and in predicting hypoxic ischemic and reperfusion injury [25].

Our results suggest that a pre-operative neurodevelopmental evaluation needs to be done for all patients undergoing surgery for CHD. Neuro-developmental evaluation should be included in the follow-up of all patients, and early intervention program should be initiated whenever deficits are detected. Neurodevelopmental clinics need to function in collaboration with a multidisciplinary team comprising of developmental pediatrician, neurologist, pediatric cardiologist and occupational therapist [26]. Longer follow-up is needed to assess the overall development of these children in higher intellectual domains and executive functions.

In conclusion, psychomotor developmental and mental developmental scores are delayed in 14.5% of infants one year after corrective infant heart surgery and are dynamic in nature. Patient-specific factors like early term birth and lower head circumference at one-year after surgery predicted neurodevelopmental outcomes than cardiac factors.

Contributors: RSS: designed the study, collected and analyzed data and drafted the initial manuscript. TS: did the neurodevelopmental follow up; AS: did the data analysis and revised the manuscript; RKK: contributed to the study design and revised the manuscript; BV: conceptualized and designed the study, analyzed the data, reviewed and edited the manuscript and shall act as the guarantor for the manuscript.

Funding: None; Competing interests: None stated.