Racial/Ethnic Differences in Long-COVID-Associated Symptoms among Pediatrics Population: Findings from Difference-in-differences Analyses in RECOVER Program

Racial/ethnic differences are associated with the potential symptoms and conditions of post-acute sequelae SARS-CoV-2 infection (PASC) in adults. These differences may exist among children and warrant further exploration. We conducted a retrospective cohort study for children and adolescents under the age of 21 from the thirteen institutions in the RECOVER Initiative. The cohort is 225,723 patients with SARS-CoV-2 infection or COVID-19 diagnosis and 677,448 patients without SARS-CoV-2 infection or COVID-19 diagnosis between March 2020 and October 2022. The study compared minor racial/ethnic groups to Non-Hispanic White (NHW) individuals, stratified by severity during the acute phase of COVID-19. Within the severe group, Asian American/Pacific Islanders (AAPI) had a higher prevalence of fever/chills and respiratory symptoms, Hispanic patients showed greater hair loss prevalence in severe COVID-19 cases, while Non-Hispanic Black (NHB) patients had fewer skin symptoms in comparison to NHW patients. Within the non-severe group, AAPI patients had increased POTS/dysautonomia and respiratory symptoms, and NHB patients showed more cognitive symptoms than NHW patients. In conclusion, racial/ethnic differences related to COVID-19 exist among specific PASC symptoms and conditions in pediatrics, and these differences are associated with the severity of illness during acute COVID-19.


INTRODUCTION
The post-acute sequelae SARS-CoV-2 infection (PASC) has emerged as a signi cant concern [1][2][3] , particularly among young individuals with a previous diagnosis of COVID-19 [4][5][6][7] .De ned by the World Health Organization (WHO) as the persistence of at least one physical symptom for 12 weeks following initial testing without an alternative diagnosis and expanded by the National Institutes of Health (NIH) to include ongoing, relapsing, or new symptoms four or more weeks post-acute infection, PASC potentially affects a signi cant proportion of COVID-19 survivors [8][9][10] .In pediatric populations, recent studies show that potential PASC symptoms and conditions tend to be systematic and/or syndromic, with higher incidence conditions such as loss of taste/smell, myocarditis, and symptoms associated with cold-like illness occurring in patients after the acute phase of COVID-19 11 .The estimated prevalence of potential PASC symptoms and conditions in the pediatric population ranges from 23-45% among those previously infected by SARS-CoV-2 [12][13][14] , depending on study designs and PASC de nitions.These ndings highlight the urgent need for further research and comprehensive support to address the prevalence of PASC in pediatric populations.
Prior investigations into potential racial/ethnic differences in PASC among adults have unearthed important ndings.For example, the Centers for Disease Control and Prevention (CDC) have shown variations in PASC's impact based on race/ethnicity 15 .A study by Khullar et al. 16 reported that Non-Hispanic Black (NHB) individuals exhibited a higher incidence of new PASC symptoms and conditions compared to Non-Hispanic White (NHW) patients, a difference more pronounced in hospitalized than in non-hospitalized patients.These ndings suggest the existence of potential racial/ethnic differences in PASC among adults.Importantly, it is crucial to note that race and ethnicity are social constructs rather than biological ones 17,18 .Concurrently, research indicates children's likelihood of testing positive for COVID-19 correlates with their race/ethnicity [19][20][21][22][23] .NHB, Hispanic, and multi-racial children exhibited higher rates of COVID-19 positivity compared to their NHW counterparts, indicating differences in infection rates across different racial/ethnic groups 24 .However, limited research to date has addressed potential racial/ethnic differences in PASC among children and adolescents, making it a pressing area of study.Therefore, our study aimed to quantify such racial/ethnic differences by conducting an association study involving children and adolescents, to determine if the observed patterns are consistent with the ndings from studies conducted among adults.
Examining health outcomes through the lens of racial/ethnic differences in the context of COVID-19 runs the risk of pre-existing racial/ethnic differences being either overshadowed or underestimated.To address this, we employed a difference-in-differences approach to discern the shifts in racial/ethnic differences before and after COVID-19.As existing research has focused predominantly on adult populations, the clinical evidence regarding these differences among the pediatric population also remains uncertain.Our investigation centered on a pediatric cohort from the RECOVER electronic health records (EHR) database across thirteen institutions.To the best of our knowledge, this is the rst and the largest study for the pediatric population with the longest follow-up that investigates the racial difference in PASC symptoms and conditions attributable to SARS-CoV-2 infection.This study also included the Asian American/Paci c Islanders (AAPI) population, which was notably absent in prior research.Our study aims to answer several key unaddressed questions.First, we sought to compare the incidence of potential PASC symptoms and conditions among COVID-19-positive patients with the incidence among those without documented SARS-CoV-2 infection.Second, we aimed to ascertain whether COVID-19 status correlates with any racial/ethnic differences across potential PASC symptoms and conditions.More importantly, we quanti ed the racial/ethnic differences linked with COVID-19 by carefully accounting for pre-infection disparities through the application of difference-in-differences analyses.We also applied a novel negative control method to adjust for the impacts of potential unmeasured confounders.By addressing these questions, our objective is to shed light on the relationship between COVID-19, racial/ethnic differences, and potential PASC symptoms and conditions.

RESULTS
The study involved Incidence of PASC symptoms and conditions for COVID-19 positive and negative patients Table 2 presents the incidence for 24 potential PASC symptoms and conditions in the COVID-19 positive cohort compared with the COVID-19 negative cohort, strati ed by acute COVID-19 severity status.The data in Table 2 reveal that the incidence rates of all listed PASC symptoms and conditions were signi cantly increased in COVID-19-positive patients as compared with the COVID-19-negative group during the follow-up period.For example, the incidence of respiratory signs and symptoms for COVID-19positive patients was 9.68% while the incidence was 7.25% for the COVID-19-negative group (P < 0.001).Moreover, the incidence of the potential PASC symptoms and conditions in the severe group was increased compared to the incidence of these symptoms and conditions within the non-severe COVID-19 patient group.Racial/ethnic differences in PASC symptoms and conditions After achieving the balance of SMD (Section S8), Fig. 3 shows the racial/ethnic difference attributable to COVID-19 in potential PASC symptoms and conditions by the severity of COVID-19.Overall, we found moderate evidence of an increase in composite outcomes, i.e., at least one condition and any of the syndromic conditions, after SARS-CoV-2 infection among the AAPI group in both severe and non-severe COVID-19 group, but there was no strong evidence of increased racial differences among Hispanic and Non-Hispanic Black groups.
[Insert Fig. 3 Here] For patients with severe COVID-19, AAPI patients showed a higher increase in any of the conditions (RR 1.24, 95% con dence interval (CI) 1.04 to 1.49, P = 0.019) and any of syndromic conditions (RR 1.22, 95% CI 1.01 to 1.47, P = 0.042) compared to NHW after SARS-CoV-2 infection.Hispanic patients showed no increase in any of the conditions (RR 0.99, 95% CI 0.91 to 1.08, P = 0.804), and NHB patients showed a minor decrease in any of the conditions (RR 0.93, 95% CI 0.85 to 1.24, P = 0.147) as compared to NHW patients.For patients with non-severe COVID-19, AAPI patients showed a higher increase in any of the conditions (RR 1.08, 95% CI 1.01 to 1.14, P = 0.015) and any of syndromic conditions (RR 1.08, 95% CI 1.01 to 1.08, P = 0.017) compared to NHW.Hispanic patients showed almost no increase in any of the conditions (RR 1.01, 95% CI 0.98 to 1.04, P = 0.498), and NHB patients also showed almost no decrease in any of the conditions (RR 0.99, 95% CI 0.89 to 1.11, P = 0.915) as compared to NHW patients.
However, there exist statistically signi cant differences among all minority groups across several PASC symptoms and conditions after SARS-CoV-2 infection.For example, for patients with severe COVID-19, the increased prevalence of hair loss among Hispanic patients was greater (RR 2.62, 95% CI 1.06 to 6.49, P = 0.038) than the increased prevalence among NHW patients.The corresponding increase in the prevalence of fever and chills among AAPI was greater (RR Furthermore, we observed a differential increase by racial/ethnic group within both severe and non-severe groups.These racial/ethnic differences varied depending upon the severity of the acute phase of COVID-19 as well as the speci c potential PASC symptoms and conditions being analyzed.For example, among the severe group, the differential increase in abdominal pain was more pronounced for all three minority groups compared to those in the non-severe category.

Sensitivity analysis
Figure S2 showed the results of the negative control outcome experiments and estimated systematic error, such as the unmeasured confounder bias.Figure S3 showed the racial/ethnic differences after SARS-CoV-2 infection strati ed by severity of COVID-19, using standard regression models.Among COVID-19 patients within the severe group, NHB patients showed a greater incidence in any of the conditions (RR 1.16, 95% CI 1.02 to 1.32, P = 0.024) and any syndromic conditions (RR 1.14, 95% CI 1.00 to 1.30, P = 0.042) as compared to NHW patients.Hispanic patients also showed a greater incidence in any of the conditions (RR 1.12, 95% CI 0.99 to 1.27, P = 0.075) as compared to NHW patients.These ndings revealed that our difference-in-differences approach identi ed fewer racial/ethnic differences compared to standard regression models.It is worth noting that the difference-in-differences approach adjusted for the baseline racial/ethnic difference before the SARS-CoV-2 infection, a step that a standard regression analysis failed to take into consideration.Consequently, some of the observed racial/ethnic differences with prior work might not be attributed to COVID-19.Nevertheless, given its adjustment for baseline racial/ethnic differences, the difference-in-differences approach holds greater robustness.
In the analysis including only those patients identi ed based on positive SARS-CoV-2 PCR or antigen testing (Section S4), differences among severe patients were diminished among some potential PASC symptoms and conditions, while among the non-severe patients, the differences that we identi ed were consistent in both sets of analyses.To account for the potential bias stemming from limited hospital capacity during the initial COVID-19 period, we performed a secondary analysis excluding COVID-19 patients from the rst wave of the pandemic (March to May 2020).This exclusion did not signi cantly alter the results, as demonstrated in Section S5.Section S6 shows the results of subgroup analysis by age group.Section S7 shows the results of strati ed analysis by virus variants.

DISCUSSION
We examined racial/ethnic variations in long-term consequences of documented SARS-CoV-2 infection across thirteen health institutions in the RECOVER study.Our analysis revealed a higher incidence of potential Post-Acute Sequelae of SARS-CoV-2 (PASC) symptoms in COVID-19-positive patients, with differences attributed to severity.Notably, NHB patients showed a smaller increase in skin symptoms compared to NHW patients, consistent with previous ndings in adults.This observation aligns with the ndings reported for the adult population by Khullar et al. 30 .After accounting for pre-existing differences and confounder bias, moderate evidence suggested greater differences attributable to COVID-19 for AAPI compared to NHW, while no strong evidence indicated disparities in composite outcomes for NHB or Hispanic populations compared to NHW.
Our study has multiple strengths.First, we used propensity score matching methods instead of linear regression models in our adjustment of the confounders, which helped us reduce the non-linear effects of the confounders 31 .Second, we accounted for the pre-infection racial/ethnic difference in long-term COVID-19-related symptoms and conditions.This enabled us to quantify racial/ethnic differences attributable to COVID-19 in the PASC symptoms and conditions and to control any pre-infection differences in these health issues.Third, we used negative control outcomes to calibrate the systematic bias, which is powerful in controlling the unmeasured confounders.
Our study has several research directions that warrant future investigations.First, socioeconomic differences due to race/ethnicity may exacerbate racial/ethnic differences in potential PASC symptoms and conditions, thereby acting as a mediator effect in the causal pathway between race/ethnicity and clinical outcomes.Such in uences have been suggested as risk factors for acute COVID-19 by Chisolm and colleagues in a RECOVER EHR study 32 .Future research on PASC outcomes is of interest to study such mediation effects.
Secondly, health-seeking behavior or healthcare access is an important consideration 33 .It is possible that certain minority racial groups have more limited access to care and associated medical records and that this contributes to potential bias in the observed racial/ethnic differences.Related issues were recently described by Nasir et al. 34 for the ascertainment of PASC symptoms and conditions in adult populations through EHR 34 .
Thirdly, confounding poses a signi cant bias threat in observational studies.To address this, we extensively adjusted for potential confounders using a propensity-score-based matching method and difference-in-differences analyses.We employed negative control outcomes to reduce the residual bias, such as unmeasured confounder bias.Additionally, EHR data completeness issues may lead to misclassi cation and loss-to-follow-up bias.Some attempts have been made to mitigate the impacts of these biases [35][36][37][38] .The analysis used a combined set of patients, but potential race/ethnicity bias may vary between outpatients and inpatients.Addressing these issues can help improve the reliability of evidence generated from these investigations.
In summary, we rigorously quanti ed the racial/ethnic differences in potential PASC symptoms and conditions and the impact of SARS-CoV-2 infection on these differences.The impact of COVID-19 varied across racial/ethnic groups, severity of acute COVID-19, and different PASC symptoms and conditions.

COHORT CONSTRUCTION
We conducted a retrospective study from March 1, 2020, to October 3, 2022, with at least 6 months of follow-up time.We included patients under the age of 21 who had at least one visit within 18 months to 7 days before the index date (de ned as the baseline period) and at least one encounter within 28 days and 179 days after the index date (de ned as the follow-up period).For COVID-19-positive patients, we included the patients who had positive polymerase-chain-reaction (PCR), serology, or antigen tests or diagnoses of COVID-19, or post-acute sequelae of SARS-CoV-2 (PASC), which we de ned as documented SARS-CoV-2 infection.The index date for these patients was de ned as the rst time of SARS-CoV-2 infection.For COVID-19-negative patients, we included patients who had neither a documented SARS-CoV-2 infection nor a diagnosis of multisystem in ammatory syndrome in children (MISC) within the same study period, and who had at least one negative COVID-19 test result.A random negative test was chosen as the index date for COVID-19-negative patients.The selection of participants for both COVID-19 positive and negative patients in real-world data is summarized in Fig. 1.

PATIENT CHARACTERISTICS
The primary exposure was race/ethnicity, categorized into NHW, NHB, Hispanic, and AAPI.
Other/unknown racial/ethnic groups were classi ed as missing or race/ethnicity not listed above and were excluded due to small sample sizes.Various patient characteristics were considered as confounders, such as age at the cohort entry date (< 5, 5-12, 12-21), gender (female, male), cohort entry month (from March 2020 to October 2022), site indicators, obesity (obese, non-obese), a chronic condition indicator as de ned by the Pediatric Medical Complexity Algorithm (PMCA, no chronic condition, non-complex chronic condition, and complex chronic condition)25, healthcare visits (inpatient, outpatient, and emergency department visits), medications (0, 1, 2, ≥ 3), negative tests (0, 1, 2, ≥ 3), vaccine doses (0, 1, ≥ 2), and immunization duration during the baseline period (no vaccine, < 4 months, ≥ 4 months).The severity of COVID-19 at the cohort entry date was strati ed into the following categories: asymptomatic, mild (symptomatic), moderate (involving moderately severe COVID-19-related conditions like gastroenteritis, dehydration, and pneumonia), and severe (comprising unstable COVID-19related conditions, ICU admission, or mechanical ventilation)26.In this paper, we categorized patients exhibiting either asymptomatic or mild symptoms as belonging to the "non-severe" group, while all other patients were classi ed as part of the "severe" group.

STATISTICAL METHODS
We calculated the incidence of potential PASC symptoms and conditions in both COVID-19 positive and negative cohorts strati ed by severity.For each PASC symptom or condition, we calculated its incidence by dividing the number of patients who experienced the symptom or condition during the follow-up period but not at baseline by the total number of patients.To quantify the racial/ethnic differences in the potential PASC symptoms and conditions, we use relative risk (RR) as the comparative measure.The RR is known to be a collapsible measure, where collapsibility 27 refers to the measure of association conditional on some factors that remain consistent with the marginal measure collapsed over strata 28 .
To eliminate the impact of potential measured confounders, we used a propensity score matching technique with the covariates detailed in the patient characteristics section.The propensity score is calculated by the logistic regression model tted by regressing the racial/ethnic groups on the covariates.We performed this matching separately for minority racial/ethnic groups (NHB, Hispanic, and AAPI), each strati ed by severity status, compared with the NHW group.After performing the matching, we assessed the standardized mean difference (SMD) between each covariate value for different racial/ethnic groups, with a difference of 0.1 or less indicating an acceptable balance 29 .Subsequently, we quanti ed the differential increase in the prevalence of potential PASC symptoms and conditions across different racial/ethnic groups by the difference-in-difference method.A Poisson regression model was tted by regressing the potential PASC symptoms and conditions on racial/ethnic groups, SARS-CoV-2 infection status, and their interaction terms.Figure 2 provides a visual representation of the difference-in-difference method used to estimate racial/ethnic differences in the increased prevalence of potential PASC symptoms and conditions related to COVID-19.

SENSITIVITY ANALYSES
We conducted a list of sensitivity analyses to examine the robustness of our ndings.First, to evaluate how different statistical methods might in uence the analytical results, we used an alternative approach, multivariate regression analyses, with RR as the comparative measure.Speci cally, we considered the incidence of potential PASC symptoms and conditions as outcomes, while controlling for the same confounders that were used in the matching process within the difference-in-differences analysis.

ADJUSTMENT FOR UNMEASURED CONFOUNDERS
While we used propensity score matching to account for the measured confounders and difference-indifferences analyses to address pre-infection racial/ethnic differences, the results can still be impacted by unmeasured confounder bias.To mitigate such bias, we collected 31 negative control outcomes, as prespeci ed by pediatric physicians, that should not exhibit racial/ethnic differences due to COVID-19.By using these negative control outcomes, the study was able to calibrate the residual bias from unmeasured and systematic sources.A comprehensive explanation of our statistical methods can be found in Section S1 of the Supplementary Materials.Illustration of difference-in-differences analysis for disentangling racial/ethnic differences related to COVID-19 infections in potential PASC symptoms and conditions from the pre-infection observed racial/ethnic differences.

Figure 1
Figures

Table 1
Baseline characteristics of COVID-19 positive patients, by race/ethnicity and severity status.

Table 2
Raw incidence (%) of potential PASC symptoms and conditions comparing COVID-19 positive and negative patients.
This retrospective cohort study is part of the NIH Researching COVID-19 to Enhance Recovery (RECOVER) Initiative (https://recovercovid.org/),which aims to learn about the long-term effects of COVID-19.The data were contributed by thirteen sites.Participating institutions in this study included: Children's Hospital of Philadelphia, Cincinnati Children's Hospital Medical Center, Children's Hospital of Colorado, Ann & Robert H. Lurie Children's Hospital of Chicago, Nationwide Children's Hospital, Nemours Children's Health System (in Delaware and Florida), Duke University, University of Iowa Healthcare, University of Michigan, University of Missouri, OCHIN, University of California, San Francisco, and Vanderbilt University Medical Center.
11Here] DEFINING OUTCOMES Our de nition of potential PASC symptoms and conditions included 24 symptoms and conditions as shown in Rao et al.11, including abdominal pain, abnormal liver enzyme, acute kidney injury, acute respiratory distress syndrome, arrhythmias, cardiovascular signs and symptoms, changes in taste and smell, chest pain, cognitive functions, fatigue and malaise, fever and chills, uid and electrolyte, generalized pain, hair loss, headache, heart disease, mental health disorders, musculoskeletal pain, myocarditis, myositis, Postural Orthostatic Tachycardia Syndrome (POTS) or dysautonomia, respiratory signs and symptoms, skin symptoms, and thrombophlebitis and thromboembolism.Systematic and syndromic conditions related to PASC were grouped by the 24 potential PASC symptoms and conditions.
Second, we conducted analyses for COVID-19 patients identi ed only by positive SARS-CoV-2 PCR or antigen tests, because the recorded date of COVID-19 diagnosis may not accurately re ect the actual infection date.Third, we conducted analyses excluding patients whose index dates fell within the rst wave of COVID-19 (March to May 2020) due to limited SARS-CoV-2 testing availability during this period.Additionally, our sensitivity analysis featured strati cation by a set of age group strata (< 5, 5-12, 12-21), differing from the ones previously speci ed, and by estimated time frames corresponding to dominant virus variants (pre-Delta, Delta, Omicron).