Evaluation of a Hospitalized Pediatric COVID-19 Cohort from Indian National Clinical Registry of COVID-19

The impact of the COVID-19 pandemic has been widespread, and no one has been spared. COVID-19 severely affected the geriatric population and adults with comorbidity; children have also been affected and have suffered from the pandemic [1, 2]. Children are a particularly vulnerable group, especially the younger ones, considering their limited ability to express and take care of themselves [3].

Since January 2020, many case series as well as retrospective analyses describing the clinical features and course of COVID-19 in Indian children have been published. However, the majority of these studies recruited a limited number of children, and described the course during the first or second wave of the pandemic [4,5,6,7,8,9]. Data regarding the especially vulnerable population of neonates and infants are scarce [10, 11]. Factors, modifiable or not, associated with COVID-related mortality have not been well studied in the pediatric population. In the current analysis, the authors present the details of hospitalized COVID-19 patients belonging to the age group 0–18 y from 42 government and private hospitals across India, spanning all three waves of the pandemic experienced in the country. They have also identified the factors associated with in-hospital mortality in pediatric COVID-19 patients.

The National Clinical Registry for COVID-19 (NCRC) serves as a platform for on-going prospective data collection, which has been developed and maintained by the Indian Council of Medical Research (ICMR) in collaboration with the Ministry of Health & Family Welfare (MoHFW), Government of India, the All India Institute of Medical Sciences, New Delhi (AIIMS), and the ICMR-National Institute of Medical Statistics (ICMR-NIMS). The structure of the registry is available in the public domain (https://www.icmr.gov.in/tab1ar1.html). The details of the functioning of NCRC have been published elsewhere [12]. Across the network of the NCRC, participating hospitals recruited consecutive in-patients who had SARS-CoV-2 infection confirmed by real-time polymerase chain reaction (RT-PCR), nucleic acid amplification test (NAAT), or rapid antigen test (RAT). Data pertaining to participants 0–18 y of age were retrieved for the purpose of this record-based analysis.

Sociodemographic, clinical, laboratory, and hospital outcome data were analyzed. Categorical data have been presented as frequency and proportions, and continuous data with mean ± standard deviation or median (interquartile range), as appropriate. The logistic-regression model was used to identify factors associated with death among children. The death of a COVID-19 pediatric patient due to any cause during hospital stay was considered the outcome of interest for these analyses. Patients who were transferred to another hospital or left against medical advice (LAMA) were excluded from the outcome analyses, though their baseline characteristics were analyzed. Age, gender, anthropometric indices, pre-existing comorbidities, breathing difficulty, and central nervous system symptoms at admission as well as severity assessment by the WHO ordinal scale were used as explanatory variables in univariate analyses [13]. The variables with significant associations (p < 0.05) and those with known clinical or contextual importance were included in the multivariate logistic-regression model. Data analysis was carried out using STATA v14 (College Station, TX, US). The anthropometric z scores were calculated using the Anthro and Anthroplus software (WHO, Geneva).

Necessary approvals for the current investigation were obtained from the Central Ethics Committee for Human Research at ICMR (CECHR) as well as from the respective institutional ethics committees of each of the participating centers. Considering the observational nature of the registry, and the usage of anonymized data from the routine case records of the patients, a waiver of consent was granted by the aforementioned ethics committees.

From 1^st September 2020 to 20^th February 2022, 1244 hospitalized COVID-19 patients aged 0–18 y were enrolled in the NCRC from 42 centers across India (Supplementary Fig. S1). While almost half of the patients belonged to the age group of 10–18 y, around one-fifth of the patients were infants and neonates; the youngest being 1-d old. Approximately 2/3^rd of the patients in the 0–5 y age bracket had weight-for-age and weight-for-height z scores between −2 and +2. Weight for height also followed a similar pattern. Body mass index (BMI) was calculated among children of 6–18 y age. BMI of 77% children was between −2 and +2 standard deviations. A history of contact with a COVID-positive patient could be elicited in 22.1% (275/1244) of participants. The baseline demographic characteristics at admission are detailed in Table 1. Seventy-six percent of the children were admitted at a WHO ordinal scale of 3 (Table 1).

Only 68.6% of the children were symptomatic at admission. Among the symptomatic patients, fever was the most commonly reported symptom for all age groups. In children ≥ 5 y of age, fever was followed by a dry cough (33.3%) and breathing difficulty (22.3%). A substantial proportion of younger children between 1 and 5 y of age manifested gastrointestinal symptoms like vomiting (22.5%), diarrhea (17.7%), and abdominal pain (5.4%). Runny nose and wet cough were also reported in over 10% of children in this age group. Infants presented with fever, cough, and breathing difficulty, as well as gastrointestinal symptoms like vomiting (19.4%) and diarrhea (16.4%). Neurological symptoms such as altered consciousness (1.9%) and seizures (7.6%) were also observed in some infants. Neonates were largely asymptomatic (60.4%). The symptomatic neonates had fever, cough, breathing difficulty, and vomiting as their most common symptoms. Table 2 shows the symptom profile among symptomatic patients at baseline, segregated by age groups.

Out of the 1244 admitted children, 260 (21%) had at least one comorbidity. The most common comorbidity among children over 1 y of age was hematologic disorders and malignancy, while cardiac disease was more common among infants. Neonates did not present with comorbidities. Supplementary Table S1 shows the frequency and proportion of comorbidities across the various age groups.

Table 3 shows the laboratory profile of patients at admission. The laboratory markers were largely normal, except for the raised inflammatory markers. Figure 1 shows the treatment profile of the enrolled patients. The most commonly prescribed drug was azithromycin [n = 956/1244 (23.2%)], followed by steroids [n = 190/1244 (15.3%)], doxycycline [n = 175/1244 (14.1%)], ivermectin [n = 171/1244 (13.8%)], and anticoagulants [n = 111/1244 (8.9%)]. The rest of the drugs were administered to less than 5% of the patients. Oxygen supplementation was required by 296/1244 (23.8%) of patients, while noninvasive and mechanical ventilation were required by 51/1244 (4.1%) and 78/1244 (6.3%), respectively. The use of steroids was lowest among neonates (9.2%) with highest usage among the 10–18 y age group (17.8%). Oxygen requirement as well as invasive and noninvasive ventilation were highest among children less than 1 y of age (data not shown).

Treatment profile among patients aged 0–18 y enrolled in the NCRC, n = 1244. Values are expressed in percentages

Full size image

The data for outcome analysis were available for 1085 children. Of them, 93.8% (n = 1018) were discharged, and 6.2% (n = 67) died in the hospital. Children who were transferred to other hospitals (n = 127) or those who left against medical advice (n = 32) were excluded from the outcome analysis.

The proportion of deaths was highest in the age group 1 mo to 1 y [n = 15/120 (12.5%)], followed by neonates [n = 6/83 (7.2%)], 5–9 y [n = 11/173 (6.4%)], 10–18 y [n = 31/540 (5.7%)], and 1–4 y of children [n = 4/169 (2.4%)].

The median duration of hospital stay was 4 d (IQR: 2,9) and 6 d (IQR: 4,9) among survivors and nonsurvivors, respectively. Of the children who died, 50.8% had at least one comorbidity. Table 4 shows the proportion of outcomes across age categories, gender, pre-existing comorbidities, selected symptom categories, baseline clinical status, and nutritional status of the patients, along with unadjusted and adjusted odds ratios. The outcome was not significantly different across the age and gender categories. Bivariate logistic regression showed higher odds of occurrence of death among patients with chronic kidney disease (CKD), malignancy, and chronic neurological disease, shortness of breath/fast breathing at admission, altered sensorium or seizures at admission, or admission with the WHO ordinal scale 4 and above. On the multivariate model, the odds of occurrence of death were high among patients with malignancy (aOR: 8.9, 95% CI: 2.4, 32.3), altered sensorium or seizures at admission (aOR: 6.8, CI: 1.9, 24.6), and WHO ordinal scale 4 or above at admission (aOR: 19.6, CI: 8.0, 47.8), after adjusting for age, gender, other comorbidities, and symptoms. Anthropometric measures did not affect the occurrence of poor outcomes.

The patients were grouped based on the date of admission to the hospital for the three waves of the pandemic, with the first wave considered as the phase that lasted until 1^st February 2021, the second wave being the one from 2^nd February 2021 to 15^th December 2022, and the third wave from 16^th December to 20^th February 2022. The number of children admitted during the first, second, and third waves were 422 (33.9%), 607 (48.8%), and 215 (17.3%), respectively. The proportion of patients admitted with comorbidities and mortality across the three waves was not significantly different, though a significant shift in age towards the under-5 group was observed at the time of the third wave (Supplementary Table S2).

In this report, the authors have described the clinical presentation and course of COVID-19 illness in a large, representative cohort of 1244 pediatric patients admitted to 42 hospitals across India, which included 98 neonates and 142 infants. The majority of the participants were asymptomatic, and the mortality was lower compared to that observed in cohorts of adults.

About one-third (33.2%) participants in this cohort were asymptomatic, especially the neonates. Other Indian studies have also reported similar observations [6]. It seems that many of the pediatric admissions were either for monitoring or as an add-on diagnosis upon some other existing condition. Among symptomatic children, fever was the most common symptom across the age spectrum. This pattern mirrors that of the adult COVID-19 patients [12]. However, some interesting additional symptoms were observed in children, especially the younger ones. Gastrointestinal symptoms, including vomiting, diarrhea, and abdominal pain were seen in 18.76% of the under-5 s. Previously published systematic reviews have shown a varied frequency of gastrointestinal symptoms in the United Kingdom, the United States, and Wuhan [14]. The high expression of angiotensin-converting enzyme receptor 2 (ACE-2), and transmembrane protease serine 2 (TMPRSS2) in the gastric, duodenal, and rectal epithelium, and in the enterocytes of the ileum and colon are hypothesized to facilitate the entry of the virus in the gastrointestinal system [14]. Destruction of absorptive enterocytes by COVID-19 leads to an altered intestinal permeability, consequently leading to diarrhea [15]. Rash was another notable symptom in this age group. Neonates also presented with neurological abnormalities such as altered consciousness and convulsions. Other researchers have also reported neurological manifestations and complications [11].

In the present study, mortality in COVID-19 pediatric patient due to any cause during hospital stay was 6.2%. Such a figure is considerably lower than what is reported among hospitalized adult COVID-19 patients in India and abroad [12, 16, 17]. Other Indian reports on the pediatric population, mainly from tertiary care centers, have reported mortality ranging from 3.2% to 11.2% [6, 18, 19]. The present study, being a representative cohort, gives a holistic estimate of poor outcomes across treatment centers and the multiple waves of the pandemic. Lower mortality in children as compared to adults admitted with COVID-19 has been noted worldwide [17]. It may be attributed to the fact that children have a lower ACE2 expression in lungs, lower proinflammatory cytokine response, and are capable of mounting stronger innate immune response [20].

Overall, comorbidities were reported in only 21% of the present pediatric cohort. However, more than half of the kids who died had some coexisting medical condition. This differs slightly from other reports, mainly from tertiary care centers, where a large proportion of the admitted patients had some comorbidities [6]. The clientele of patients and the threshold of admission might be the contributing factors. Coexisting malignancy were associated with higher odds of in-hospital mortality in the present cohort. As in adults, studies evaluating factors associated with mortality in COVID-positive children have found comorbidity to be an important factor [21, 22]. The present results differ from a few studies conducted among pediatric cancer patients, which showed that malignancy was not associated with poor outcomes or mortality among COVID-19 patients [23, 24]. Altered sensorium and WHO ordinal scale ≥ 4 at admission were associated with significantly higher adjusted odds of mortality among the present study subjects. A recently published study from the US also predicted severe disease among pediatric COVID-19 patients who have abnormal vital signs at the time of admission [25].

Furthermore, malnutrition has also been linked with COVID-19 severity [26]. However, lower anthropometric indices were not associated with any adverse outcome in the present cohort. The majority of the admitted children were within normal limits for weight for height or BMI z scores. The present cohort also provides a kaleidoscopic view of the three waves of the COVID-19 pandemic in children in India. Severity or mortality did not differ, but more children in the under-5 group were admitted to the hospital during the third wave. This increase in the proportion of under-5 s coincided with factors such as the unvaccinated status of the young, the reopening of schools, etc.

The current analysis from the National Clinical Registry for COVID-19 is a large, multicentric, widely representative analysis which describes the demographic, anthropometric, clinical and outcome characteristics of hospitalized pediatric COVID-19 patients in India.

Considering this is a record-based study in hospitals, the information collected was reliant on the accuracy of the records maintained. Also, the patients who were transferred to other institutes or who had left against medical advice were not followed up and could not be included in mortality analysis as their outcomes were unknown.

The multicentric cohort of admitted Indian children showed that COVID-19 illness was milder in children than adults, with one-third being asymptomatic. This pattern was consistent over the three waves of the pandemic. Factors associated with higher odds of death were malignancy, altered sensorium, and a WHO ordinal scale ≥ 4 at admission.