key: cord-0832905-wo0t4wdl
authors: Dubois, Melanie M.; Campbell, Jeffrey I.; Savage, Timothy J.; Hood-Pishchany, M. Indriati; Sharma, Tanvi S.; Petty, Carter R.; Lamb, Gabriella S.; Nakamura, Mari M.
title: Comorbidities associated with hospitalization and progression among adolescents with symptomatic COVID-19
date: 2022-02-28
journal: J Pediatr
DOI: 10.1016/j.jpeds.2022.02.048
sha: 35b941d44c11ea806826d3d6dc1f6f8e34174648
doc_id: 832905
cord_uid: wo0t4wdl

OBJECTIVES: To identify subgroups likely to benefit from monoclonal antibody and antiviral therapy, we evaluated the relationship between comorbidities and hospitalization among US adolescents with symptomatic COVID-19. STUDY DESIGN: We analyzed the relationship between presence of comorbidities and need for hospitalization within 28 days of COVID-19 diagnosis for adolescents 12-17 years old in the Pediatric COVID-19 U.S. registry, a multicenter retrospective cohort of US pediatric patients with COVID-19. Comorbidities assessed included obesity, chronic kidney disease (CKD), diabetes, immunosuppressive disease or treatment, sickle cell disease (SCD), heart disease, neurologic disease/neurodevelopmental disorders, and pulmonary disease (excluding patients with mild asthma). We used multivariable logistic regression to determine race/ethnicity-adjusted associations between comorbidities and hospitalization. RESULTS: 1877 patients met inclusion criteria, of whom 284 (15%) were hospitalized within 28 days of COVID-19 diagnosis. In a race/ethnicity-adjusted model, the following comorbidities were independently associated with increased odds of hospitalization: SCD (aOR 6.9 [3.0-15.9]), immunocompromising condition (aOR 6.4 [3.8-10.8]), obesity (aOR 3.2 [2.1-4.9]), diabetes (aOR 3.0 [1.4-6.2]), neurologic disease (aOR 2.8 [1.8-4.3]), and pulmonary disease (excluding mild asthma) (aOR 1.9 [1.2-3.1]). Heart disease and chronic kidney disease were not independently associated with hospitalization. CONCLUSIONS: SCD, immunocompromising conditions, obesity, diabetes, neurologic disease, and pulmonary disease (excluding mild asthma) were associated with hospitalization for symptomatic COVID-19. Adolescents with acute COVID-19 and these comorbidities should be prioritized for consideration of therapy to avert hospitalization.

As of January 2022, more than 8 million children in the US have been infected w ith SARS-CoV-2.(1) New SARS-CoV-2 variant strains have added to the burden of COVID-19 among children and adolescents, despite vaccinations authorized for children aged 5 and older (2) (3) (4) Even though most children and adolescents with COVID-19 are asymptomatic or experience mild illness, some progress to severe disease.(5) Data documents higher COVID-19associated hospitalization rates in adolescents 12-17 years old than in younger children [5] [6] [7] [8] [9] [10] [11] years old.(6) Evaluation of SARS-CoV-2-associated deaths from February to July 2020 in patients <21 years old showed that the most common underlying conditions were obesity, asthma, and developmental disorders. (7) Although chronic comorbidities have been associated with severe disease in children, (5, (8) (9) (10) (11) (12) data are limited regarding which conditions most strongly predict progression, To inform rational use of COVID-19 therapies, there is a need to understand risk for progression to severe disease among adolescents with COVID-19. Patients ≥12 years old weighing at least 40 kilograms with mild or moderate COVID-19 and ≥1 specified comorbidity are eligible to receive monoclonal antibodies (mAbs), including sotrovimab, or antivirals, including nirmaltrelvir plus ritonavir, to prevent progression to severe disease. (13) (14) (15) bamlanivimab plus etesevimab and casirivimab plus imdevimab have demonstrated reduced activity against the SARS-CoV-2 Omicron variant. (16) (17) (18) The Food and Drug Administration's (FDA) Emergency Use Authorizations (EUAs) for mAbs identify the following conditions as risk factors for hospitalization: elevated body mass index (for adolescents, ≥85 th percentile for age and sex); pregnancy; chronic kidney disease (CKD); diabetes; immunocompromising conditions; sickle cell disease (SCD); cardiovascular disease (including congenital heart disease); chronic lung diseases; sickle cell disease; neurodevelopmental disorders; and medical-related J o u r n a l P r e -p r o o f technology dependence. (13) However, multicenter expert guidance on pediatric use of mAb therapy has underscored the paucity of evidence supporting these risks . (19) In addition, given the rise in COVID-19 cases and emergence of the Omicron variant, there is a limited supply of active mAbs and preventative antiviral therapy, driving a need to identify adolescents most likely to benefit from treatment.

we analyzed data from the Pediatric COVID-19 U.S. Registry to identify predictors of hospitalization among adolescents diagnosed with COVID-19.

The Pediatric COVID-19 U.S. Registry is a multicenter retrospective cohort study of US patients <21 years old diagnosed with laboratory-confirmed COVID- 19 

The Boston Children's Hospital IRB also approved submission of deidentified case information and this analysis of registry data as an exempt study.

We included childrens aged 12-17 years diagnosed with symptomatic acute COVID-19 (defined as any symptoms that prompted COVID-19 testing up to 3 days prior to testing), reported to the registry between April 1, 2020 and April 30, 2021, for whom 28-day followup data were submitted by August 5, 2021. We excluded patients who were already hospitalized for any reason before the day of testing, hospitalized for a reason other than COVID-19 by day 28, or had missing 28-day follow-up hospitalization status.

Race and ethnicity were categorized as White non-Hispanic, Hispanic, Black non-Hispanic, Asian non-Hispanic, other non-Hispanic, and "missing". Age (in years) and sex were also assessed as covariates.

The primary exposure was pre-existing comorbidities specified in the mAb EUAs. (13, 17, 18) Obesity, CKD, diabetes (any type), SCD, heart disease, neurologic disease, and immunocompromising conditions were captured as binary survey variables in the registry; additional details about diagnoses were available for some but not all comorbidities. Using available survey data, we specified a variable for pulmonary disease excluding mild asthma (defined as asthma not requiring daily inhaled corticosteroids). We defined medical technology dependence as use of tracheostomy with or without mechanical ventilation, need for total parenteral nutrition, or use of a gastrostomy/gastrostomyjejunostomy tube. Although technology dependence was evaluated in descriptive analysis, it was removed from analysis of associations due to small sample size. The secondary J o u r n a l P r e -p r o o f exposure was trichotomized total comorbidity burden (0 comorbidities, 1 comorbidity, or 2 or more comorbidities); comorbidity burden did not include technology dependence.

The primary outcome was hospitalization for COVID-19 (which was a clinical determination adjudicated by investigators at time of data entry) at any time between Days 0 (COVID-19 diagnosis) and 28, which served as a proxy for severe disease. The secondary outcomes were intensive care unit admission between Days 0 and 28, hospitalization between Days 1-28 among patients not hospitalized on Day 0, and need for any form of oxygen support between Days 1-28. Secondary outcomes were selected to represent incidence of severe disease and progression of COVID-19 disease among patients presenting with mild-to-moderate illness. When dates of hospitalization or initiation of oxygen support were not specified, they were assumed to have occurred after Day 0; we made this assumption because 97% and 100% of cases with these data missing, respectively, were reported in the 28-day follow-up assessment of hospitalization, which did not include a field for date of hospitalization or date of oxygen initiation, rather than in the initial Day 0-7 survey.

Because of conflicting evidence about the association between different forms of pre-existing pulmonary disease and risk for COVID-19-related hospitalization, (21, 22) we performed a sensitivity analysis using the following variables as the exposure of interest: 1) asthma requiring daily inhaled corticosteroids, 2) pre-existing pulmonary disease except for asthma, and 3) pre-existing pulmonary disease, including mild asthma.

We used frequencies and percentages to describe demographic characteristics, comorbidities, and outcomes. We used the Fisher exact test and chi-squared test (for J o u r n a l P r e -p r o o f categorical variables) and t test (for continuous variables) to assess univariabl e associations between exposures/covariates and outcomes. We evaluated for possible collinearity between comorbidities using point-biserial pairwise correlation coefficients, with a correlation of >0.8 indicating potential for collinearity.

We performed multivariable logistic regression to model associations between exposures and outcomes. All comorbidity variables except for technology dependence (due to small sample size) were included in models because they were hypothesized a priori to independently predict outcomes, and to enable adjustment for presence of multiple comorbidities. Models were adjusted for demographic characteristics (age, sex, and race/ethnicity) when univariable tests of association between these factors and outcomes were significant at P<0.1. Significance in multivariable analysis was set at P<0.05. In cases of zero cells, relevant variables were removed from the models to preserve sample size.

We constructed models to evaluate independent associations between comorbidities and the following outcomes: 1) hospitalization between Day 0-28; 2) ICU admission between Day 0-28; and 3) hospitalization between Day 1-28, excluding patients hospitalized on Day 0. We constructed two additional models to assess relationships between comorbidities specified using the trichotomized comorbidity variable and the outcomes of 4) oxygen requirement between Day 1-28, excluding patients requiring oxygen on Day 0 and 5) hospitalization on Day 0 versus hospitalization on Day 1-28. Model 5 assessed whether the odds of hospitalization associated with comorbidities are greater in the days following diagnosis of COVID-19 than at diagnosis. Models 3, 4, and 5 evaluate risk of disease progression. We used the trichotomized comorbidity burden score rather than individual comorbidities in Models 4 and 5 due to concerns for model overfitting, given small numbers J o u r n a l P r e -p r o o f of outcomes. We performed sensitivity analyses with different definitions of pulmonary disease using Model 1.

Finally, we conducted an exploratory analysis for interaction between comorbidities on risk of hospitalization, hypothesizing that interaction of comorbidities would confer additive risk of hospitalization. We used logistic or exact logistic regression models, including each pair of comorbidities and their interaction term to screen for potentially significant interactions. We then performed a multivariable interaction model that included interaction terms that were significant at the P<0.1 level in the screening analysis, as well as all comorbidities and race/ethnicity.

The Pediatric COVID-19 U.S registry contained 2683 patients aged 12-17 years old with COVID-19, of whom 1877 were included in our primary analysis (Figure 1 ). Mean age was 15.3 years (SD 1.4), and 992 (53%) were female ( Table I) . By Day 28 after COVID-19 diagnosis, 284 patients (15%) were hospitalized, of whom 93 (5%) were hospitalized ≥1 days after diagnosis, suggesting progression of disease. There were no documented deaths.

Among included patients, 448 (24%) had ≥ 1 comorbidity ( Table I; Table II; available at www.jpeds.com). The most common conditions were obesity (7.0%), neurologic disease (6.6%), pulmonary disease excluding mild asthma (5.6%), and immunocompromising conditions (3.7%). In univariable analysis, all comorbidities were significantly associated with hospitalization between Day 0-28 (Table III; Immunocompromising conditions, obesity, pulmonary disease (excluding mild asthma), and heart disease were associated with ICU admission between Day 0 -28 ( Figure   2 , Table IV ). Immunocompromising conditions, SCD, CKD, diabetes, neurologic disease, obesity, and pulmonary disease (excluding mild asthma) were associated with hospitalization between Day 1-28 (Figure 2 , Table IV (Table V) .

In the interaction analysis, the interaction term for obesity and immunocompromising condition was significant in screening analysis (P=0.042). This interaction term remained significant in the multivariable interaction model, though the effect was negative (aOR 0.04 [95%CI 0.005-0.35]), indicating that the risk conferred by these two comorbidities was less than additive for the 5 patients who had both comorbidities. The observed hospitalization rate in this group was 40%, which was similar to the observed hospitalization rate in immunocompromised patients overall (50%) and patients with obesity overall (38%). Because the interaction finding was not intuitive, we examined the 5 patients with obesity and immunocompromising conditions in more detail and noted that 4 of 5 patients had other pre-specified comorbidities. We therefore interpret the interaction finding as suggesting that these patients' risk of hospitalization was likely accounted for by the contributions of other comorbidities, rather than just the combination of obesity and immunocompromising conditions.

We used a national registry of 1877 adolescents with symptomatic acute COVID-19 to investigate risk factors for COVID-19-related hospitalization within 28 days of diagnosis.

Adjusting for race/ethnicity, we found that SCD, immunocompromising conditions, obesity, diabetes, neurologic disease, and pulmonary disease (excluding mild asthma) were associated with hospitalization at or following COVID-19 diagnosis. The same conditions, with the addition of CKD, were associated with progression to hospitalization between Day 1 -28.

Adolescents with these comorbidities should be prioritized to receive interventions to avert disease progression. study, as in ours, obesity was likely under-reported, which is a limitation to our findings. In our study, only 5.6% of subjects were reported to have obesity, compared with an estimated 21.2% prevalence among US adolescents aged 12-19 years. (37) Diabetes and CKD were each associated with progression to hospitalization after day 0, but analyses were limited by small numbers of patients in each category. Diabetes was additionally associated with hospitalization, but CKD was not. Despite sample limitations, our results contribute to accumulating evidence that both type 1 and type 2 diabetes are risk factors for severe COVID-19 in children and adults. (12, 38) One pediatric study found that subjects with either type 1 or type 2 diabetes and COVID-19 were at elevated risk for hospitalization although the magnitude of risk differed between types.(10) In addition, COVID-19 has led to increased rates of diabetic ketoacidosis and disruptions to long-term diabetes care. (39) Diabetes was reported to the Pediatric COVID-19 U.S. registry without differentiation between types 1 and 2, which may confer risk for severe disease through different pathways. Large registries have identified CKD as a risk factor for COVID-19 severity and mortality, though no clear mechanism has been determined. (40, 41) Challenges with studying CKD-associated risk include the rarity of this condition among children and its association with a variety of comorbidities that may also predispose to severe illness. In our cohort, only 13 patients had CKD.

Our results concur with prior research showing that neurodevelopmental disorders are prevalent among US adolescents hospitalized with COVID-19(6) and are a risk factor for hospitalization and progression. (10, 42) The mechanisms underlying this association are not known. As in our study, Kompaniyets et al identified neurodevelopmental disorders as a risk for hospitalization, but these disorders were not associated with severe illness among hospitalized adolescents. (10) In that study, neurodevelopmental disorders mostly consisted of mental health disorders, and the identified association may reflect the COVID-19 and mental health syndemics, rather than physiologic predisposition to severe COVID-19.(10) Neurologic disease is also often a marker of medical complexity, and our findings may reflect increased need for hospital-based supportive care among adolescents with complex conditions. (9, 43) We found that pulmonary disease (excluding mild asthma) was associated with hospitalization, ICU admission, and hospitalization after Day 0. Prior studies have conflicted regarding risk of severe COVID-19 among patients with asthma and non-asthma pulmonary disease. Large observational studies have found asthma to be associated with hospitalization risk, (6, 10) although smaller retrospective studies have found asthma to be protective against hospitalization (21) . Little data exist to guide risk assessments among adolescents with nonasthma pulmonary disease.

We found no association between heart disease and hospitalization, although patients with heart disease were at increased risk of ICU admission. Consistent with the current EUAs, our study included adolescents with hypertension as well as congenital cardiac structural and muscular diseases.(17,18) Hypertension has been identified as a risk factor for severe disease among adults with COVID-19,(44) whereas congenital heart disease has been associated with increased risk among young children. showing that race/ethnicity and specific comorbidities are independently associated with hospitalization. (10, 45) In our interaction model, we found that the interaction between obesity and immunocompromising conditions was statistically significant, though the result did not show additive risk. Four of 5 patients with obesity and immunocompromising conditions had at least one other pre-specified comorbidity, and their hospitalization risk was likely accounted for by the other condition(s). Interpretation of the interaction analysis was also limited by small sample sizes. We found that a higher total number of comorbidities was associated with increased odds of oxygen need between Day 1-28, as well as odds of hospitalization between Day 1-28 J o u r n a l P r e -p r o o f compared with on Day 0. Additional investigation into the combined effects of specific comorbidities is needed to better risk stratify adolescents with medical complexity.

Our study has several limitations. The Pediatric U.S. COVID-19 registry is a voluntary surveillance study using secondary data abstracted from medical records. Although the registry captured need for hospitalization due to COVID-19, the specific reasons for hospitalization were not available. Some hospitalizations may have reflected policies for treatment of high-risk patients rather than disease severity. Additionally, the infrequency of hypoxemia precluded analysis of associations between this outcome and specific comorbidities. Further, because the registry did not capture granular data on comorbidity characteristics, we were unable to evaluate subgroups of patients with specified comorbidities (e.g. type 1 vs type 2 diabetes). Some comorbidities may have been underreported, including obesity. Additionally, race and ethnicity were based on information available in medical records, which has substantial limitations, including missing data, discrepant categories across electronic record systems, and lack of self report.(46) Furthermore, we were unable to measure markers of socioeconomic status, which affect the association between race/ethnicity and health outcomes. Because pediatric hospitals were probably more likely to contribute cases than non-pediatric centers, registry data may be enriched for more severe cases, as reflected by a hospitalization rate of 15% in our cohort. (11) Finally, case submission stopped in April 2021, prior to the surges associated with the Delta and Omicron variants, and prior to widespread vaccination of adolescents. Risk of hospitalization for adolescents with COVID-19 will continue to change as new variants emerge, and as vaccine policy and access evolves. Nonetheless, as underscored by the recent shortages in mAbs and antivirals effective against the Omicron variant, it remains important to understand which adolescents are most likely to experience severe illness and disease progression.

Our study also has key strengths. Data derive from a national, multicenter, pediatricfocused registry gleaned from detailed chart review that captured information on comorbidities.

The database size enabled evaluation of adolescents specifically and assessment of risk conferred by rare conditions, such as CKD, that have not been reported in observational or surveillance studies. The registry prospectively captured COVID-19-related hospitalization within 28 days of diagnosis, enabling analysis of progression to hospitalization (Model 3 and Model 5), which is the outcome that mAbs have been shown to prevent. Last, we specifically selected conditions included in current mAb guidance, making our results directly relevant to decisions that clinicians currently face. 2 For analysis of patients initiating oxygen between Day 1-28, the total N is 1,771 after excluding 106 patients requiring oxygen on Day 0. 3 The percentage in the first column refers to the percentage of the total cohort with the characteristic (i.e., the column percentage). The percentage in in the second through sixth columns refers to the percentage of patients with the specified characteristic who experienced the specified outcome (i.e., the row percentage). 4 Excludes patients with technology dependence, for consistency with regression analyses. J o u r n a l P r e -p r o o f J o u r n a l P r e -p r o o f 

Children's Hospital Association. Children and COVID-19: state data report

Hospitalizations associated with COVID-19 among children and adolescents -COVID-NET, 14 states

FDA authorizes Pfizer-BioNTech COVID-19 Vaccine for emergency use in children 5 through 11 years of age

Coronavirus (COVID-19) Update: FDA Authorizes Pfizer-BioNTech COVID-19 Vaccine for Emergency Use in Adolescents in Another Important Action in Fight Against Pandemic

CDC COVID-19 Response Team. Coronavirus disease 2019 in children -United States

Hospitalization of adolescents aged 12-17 years with laboratory-confirmed COVID-19 -covid-net, 14 states

Deaths in children and adolescents associated with COVID-19 and MIS-C in the United States

COVID-19 disease severity risk factors for pediatric patients in italy

Characteristics and outcomes of children with coronavirus disease 2019 (COVID-19) infection admitted to US and Canadian pediatric intensive care units

Underlying medical conditions associated with severe COVID-19 illness among children

Factors associated with COVID-19 disease severity in US children and adolescents

Online First

Risk factors for severe COVID-19 in children

Fact sheet for healthcare providers emergency use authorization (EUA) of sotrovimab

Early treatment for covid-19 with SARS-CoV-2 neutralizing antibody sotrovimab

Fact sheet for healthcare providers: emergency use authorization for paxlovid

Food and Drug Administration. Fact sheet for health care providers emergency use authorization (EUA) of bamlanivimab and etesevimab

Fact sheet for health care providers: emergency use authorization (EUA) of casirivimab and imdevimab

Initial guidance on use of monoclonal antibody therapy for treatment of COVID-19 in children and adolescents

Prevalence of asthma in hospitalized and non-hospitalized children with COVID-19

An overview of asthma and COVID-19: protective factors against SARS-COV-2 in pediatric patients. Front Pediatr

Acute chest syndrome in the setting of SARS-COV-2 infections-A case series at an urban medical center in the Bronx. Pediatr Blood Cancer

Sickle cell disease and COVID-19: susceptibility and severity

Secure-SCD registry: surveillance epidemiology of coronavirus (COVID-19) under research exclusion

Anti-SARS-CoV-2 antibody responses are attenuated in patients with IBD treated with infliximab

How is immunosuppressive status affecting children and adults in SARS-CoV-2 infection? A systematic review

The pediatric solid organ transplant experience with COVID-19: An initial multi-center, multi-organ case series

COVID-19 in children with cancer in New York City

COVID-19 in pediatric hematopoietic stem cell transplantation: the experience of Spanish Group of Transplant (GETMON/GETH). Pediatr Blood Cancer

The severity of COVID-19 in children on immunosuppressive medication

Impact of SARS-CoV-2 pandemic on patients with primary immunodeficiency

Obesity is a risk factor for severe COVID-19 infection: multiple potential mechanisms. Circulation

The role of adipocytes and adipocyte-like cells in the severity of COVID-19 infections. Obes Silver Spring Md

Severe COVID-19 infection and pediatric comorbidities: A systematic review and meta-analysis

COVID-19 and obesity in childhood and adolescence: a clinical review

Prevalence of overweight, obesity, and severe obesity among children and adolescents aged 2-19 years: united states

Risk factors for intensive care unit admission and in-hospital mortality among hospitalized adults identified through the US coronavirus disease 2019 (COVID-19)-associated hospitalization surveillance network (COVID-NET)

Incidence of COVID-19 and risk of diabetic ketoacidosis in new-onset type 1 diabetes

Chronic kidney disease is a key risk factor for severe COVID-19: a call to action by the ERA-EDTA

Outcomes of COVID-19 in CKD patients: a multicenter electronic medical record cohort study

Severe coronavirus disease-2019 in children and young adults in the Washington, DC, metropolitan region

Scoping review of prevalence of neurologic comorbidities in patients hospitalized for COVID-19

COVID-19 and hypertension: risks and management. A scientific statement on behalf of the British and Irish Hypertension Society

Demographic predictors of hospitalization and mortality in US children with COVID-19

Use of race and ethnicity in biomedical publication

We would like to acknowledge the efforts of Dr Robert Husson at Boston Children's