Multisystem Inflammatory Syndrome in Children—United States, February 2020–July 2021

Abstract

Background

Multisystem inflammatory syndrome in children (MIS-C) is a severe hyperinflammatory condition in persons aged <21 years associated with antecedent severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection. Our objective was to describe MIS-C cases reported to Centers for Disease Control and Prevention’s (CDC’s) national surveillance since the coronavirus disease 2019 (COVID-19) pandemic began.

Methods

We included patients meeting the MIS-C case definition with onset date from 19 February 2020 through 31 July 2021, using CDC’s MIS-C case report form, which collects information on demographics, clinical presentation, and laboratory results. Trends over time across 3 MIS-C pandemic waves were assessed using Cochran-Armitage test for categorical and Jonckheere-Terpstra test for continuous variables.

Results

Of 4901 reported cases, 4470 met inclusion criteria. Median patient age increased over time (P < .001), with a median of 9 years (interquartile range, 5–13 years) during the most recent (third) wave. Male predominance also increased (62% in third wave, P < .001). A significant (P < .001) increase in severe hematologic and gastrointestinal involvement was observed across the study period. Frequency of several cardiovascular complications (ie, cardiac dysfunction, myocarditis, and shock/vasopressor receipt) and renal failure declined (P < .001). Provision of critical care including mechanical ventilation (P < .001) and extracorporeal membrane oxygenation (ECMO; P = .046) decreased, as did duration of hospitalization and mortality (each P < .001).

Conclusions

Over the first 3 pandemic waves of MIS-C in the United States, cardiovascular complications and clinical outcomes including length of hospitalization, receipt of ECMO, and death decreased over time. These data serve as a baseline for monitoring future trends associated with SARS-CoV-2 B.1.617.2 (Delta) or other variants and increased COVID-19 vaccination among children.

multisystem inflammatory syndrome in children, COVID-19, child, epidemiology

Multisystem inflammatory syndrome in children (MIS-C) is a severe hyperinflammatory condition in children and adolescents associated with antecedent severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection, characterized by fever, systemic inflammation, and multisystem organ involvement [1–6]. Clinicians in the United Kingdom first described severe inflammation in previously healthy children after acute SARS-CoV-2 infection in April 2020, and this illness was soon recognized elsewhere including the United States [5–10]. In May 2020, the US Centers for Disease Control and Prevention (CDC) issued a health advisory, outlined a MIS-C case definition, and asked clinicians to report suspected cases to local and state health departments [10]. The CDC established a national reporting platform to systematically collect data on suspected cases of MIS-C from health departments. MIS-C incidence in 7 jurisdictions from April to June 2020 was 1 case of MIS-C per approximately 3200 SARS-CoV-2 infections among persons aged <21 years [11].

MIS-C represents a severe complication of COVID-19 in children, although initial SARS-CoV-2 infection in most persons with MIS-C is mild or asymptomatic [12]. MIS-C generally occurs 2–6 weeks after SARS-CoV-2 infection, and higher MIS-C incidence closely follows peaks of reported SARS-CoV-2 circulation [2, 5, 13]. Following issue of Emergency Use Authorizations for the Pfizer-BioNTech COVID-19 vaccine, the Advisory Committee on Immunization Practices issued interim recommendations for use of the vaccine in persons ≥16 years on 12 December 2020 [14]; for adolescents aged ≥12 years on 12 May 2021 [15]; and for children aged 5–11 years, the group comprising the largest burden of MIS-C, on 2 November 2021 [11, 13, 16]. Variable adherence to recommended preventive strategies, including consistent and correct mask use, physical distancing, and vaccination among eligible persons, places unvaccinated children at risk for SARS-CoV-2 infection and subsequent development of MIS-C [17, 18]. This was especially important as the United States experienced a rise in circulation of highly transmissible SARS-CoV-2 variants including the B.1.617.2 (Delta) variant, which comprised <15% of circulating variants as of 5 June 2021, and rose to >95% by 31 July 2021 [19, 20].

In this analysis we summarize over 4000 MIS-C cases reported to CDC’s national surveillance since the start of the COVID-19 pandemic; cases reported through January 2021 have been summarized previously [13]. This period represents surveillance prior to and including the first months of authorization and recommendation for vaccination for persons aged ≥12 years. We present patient characteristics, detailed clinical and radiologic features, illness management, and clinical outcomes by each of the 3 waves of MIS-C over the study period.

METHODS

Local, state, and territorial health departments reported cases using a standardized case report form based on medical chart abstractions performed by clinicians, hospital staff, or health department staff (Supplement 1). Patients’ illnesses were evaluated to confirm they met the CDC MIS-C case definition: (1) clinically severe illness requiring hospitalization in persons aged <21 years, (2) fever ≥38°C for ≥24 hours or report of subjective fever for ≥24 hours, (3) laboratory evidence of inflammation, (4) multisystem (≥2) organ involvement, (5) laboratory evidence of acute or previous SARS-CoV-2 infection by reverse transcription polymerase chain reaction (RT-PCR), serology, or antigen test, or known COVID-19 exposure within 4 weeks of symptom onset, and (6) no alternative plausible diagnosis [21]. Information collected included patient demographics, clinical manifestations, complications, illness management, imaging studies, laboratory test results, outcomes, and vaccination status (added to case report form 21 May 2021).

To account for delays in reporting and maximize data completeness, we included patients with an MIS-C onset date on or before 31 July 2021 who were reported on or before 18 August 2021. Patients with known SARS-CoV-2 exposure without subsequent laboratory confirmation of infection were excluded from this analysis because of concerns about potential misclassification. We performed clinical review of free-text responses and supplemented patient comorbidities by classifying into available categorical variables or, for obesity, calculated body mass index using national reference standards for those with available height, weight, sex, and age information [22]. Race and ethnicity data were obtained from medical records as documented at time of hospitalization and categorized in accordance with previously established methods as missing race and ethnicity data accounted for 5.8% of patients in our cohort [23].

To assess case characteristics and outcomes over time, we reviewed the epidemic curve of MIS-C illness onset dates. If onset date was unavailable, fever onset or hospitalization admission date was used as a proxy. We identified 3 peaks of reported MIS-C cases and two nadirs (Figure 1). We defined the three time periods or “waves” of MIS-C activity in the pandemic with respect to timing of MIS-C symptom onset dates among cases to (1) February 19 through 28 June 2020, (2) June 29 through 17 October 2020, and (3) 18 October 2020 through 31 July 2021. Geographic regions were categorized in accordance with the four US census regions [24]. Intensive care unit (ICU) admission was defined as having a documented date of ICU admission or known length of ICU stay or having received ICU-level care, including mechanical ventilation, vasopressor support, or extracorporeal membranous oxygenation (ECMO). We defined values for lymphopenia and thrombocytopenia using age standards [25]. We adapted previously established frameworks to describe severe organ-system involvement (Supplement 2) [1, 26]. We performed clinical review and supplemented radiographic findings using free-text responses when available. Analyses of laboratory markers of inflammation were performed only on those with available data collected in section 6 of the case report form (Supplement 1).

Figure 1.

MIS-C Cases by U.S. Census Geographic Region with 3 MIS-C Waves and Million COVID-19 Cases by Week, United States, February 2020 to July 2021. The 3 MIS-C waves were (1) 19 February 2020 through 28 June 2020, (2) 29 June 2020 through 17 October 2020, and (3) 18 October 2020 through 31 July 2021. Gray dotted lines denote divider for each of the 3 MIS-C waves. Scale for the Y-axis differs on the left and the right sides of the figure. Left Y-axis marks the number of MIS-C cases in units of 50 with a scale of 0 to 250; right Y-axis marks the number of COVID-19 cases in units of 500 000 with a scale from 0 to 1 500 000. Each bar on the X-axis represents 1 week. Abbreviations: COVID-19, coronavirus disease 2019; MIS-C, multisystem inflammatory syndrome in children.

Using SAS version 9.4 (SAS Institute, Cary, North Carolina, USA) we calculated the frequency of clinical features, relevant laboratory findings, and treatments among patients stratified by wave of MIS-C onset. Continuous variables were expressed as medians and interquartile ranges (IQRs); trends in median values over time were assessed through the Jonckheere-Terpstra test. Categorical variables were expressed as counts and percentages; trends over time were assessed through the Cochran-Armitage test. Two-sided P values were considered significant at α <.05.

This activity was reviewed by CDC, was determined to meet the requirements of public health surveillance and was conducted in consistence with applicable federal law and CDC policy (45 C.F.R. part 46.102(l)(2), 21 C.F.R. part 56; 42 U.S.C. §241(d); 5 U.S.C. §552a; 44 U.S.C. §3501 et seq) [27, 28].

RESULTS

Clinical Characteristics

Out of 4901 total patients reported with suspected MIS-C, we excluded 249 patients who did not meet the case definition, 87 with missing illness onset date or onset after the study period, and 95 who otherwise met clinical criteria of the CDC case definition but who lacked laboratory evidence of SARS-CoV-2 infection (Supplements 3 and 4). The 4470 included cases were reported from 49 state health departments, the District of Columbia, New York City, and Puerto Rico. Median patient age for the overall cohort of MIS-C patients was 9 years (IQR: 5–13 years); 59.9% were male (Table 1). Of patients with complete race and ethnicity information, 31.1% were non-Hispanic Black, 30.6% Hispanic/Latino, and 28.9% non-Hispanic White. Underlying medical conditions were reported for 37.6% of MIS-C patients; obesity (25.1%) and chronic lung disease including asthma (9.6%) were most frequent among all MIS-C patients.

Table 1.

Characteristics of Patients With MIS-C by Wave, United States, February 2020 to July 2021.

		MIS-C Onset, no. (%)
Characteristic	All MIS-C Cases (n = 4470), no. (%)	Wave 1 (n = 649)	Wave 2 (n = 769)	Wave 3 (n = 3052)	Test for Trend P valuea
Age group, yearsb
<1	139 (3.1)	27 (4.2)	28 (3.6)	84 (2.8)	.038
1–4	946 (21.2)	158 (24.3)	164 (21.4)	624 (20.5)	.035
5–11	2016 (45.1)	281 (43.1)	349 (45.4)	1386 (45.5)	.376
12–15	888 (19.9)	106 (16.3)	144 (18.8)	638 (20.9)	.005
16–20	477 (10.7)	77 (11.9)	83 (10.8)	317 (10.4)	.281
Age, years, median (IQR)	9 (5–13)	8 (4–13)	9 (5–13)	9 (5–13)	<.001
Sex, no. (%)c
Male	2674 (59.9)	355 (54.7)	434 (56.4)	1885 (61.9)	<.001
Female	1790 (40.1)	294 (45.3)	335 (43.6)	1161 (38.1)	<.001
Race/Ethnicity, no. (%)d
Hispanic	1288 (30.6)	215 (39.0)	292 (39.9)	781 (26.7)	<.001
Non-Hispanic Black	1307 (31.1)	194 (35.2)	269 (36.8)	844 (28.9)	<.001
Non-Hispanic White	1217 (28.9)	73 (13.2)	118 (16.1)	1026 (35.1)	<.001
Non-Hispanic Asian	103 (2.4)	14 (2.5)	7 (1.0)	82 (2.8)	.167
American Indian or Alaska Native	37 (0.9)	9 (1.6)	6 (0.8)	22 (0.8)	.068
Native Hawaiian or Other Pacific Islander	33 (0.8)	8 (1.5)	9 (1.2)	16 (0.5)	.009
Non-Hispanic Other	154 (3.7)	25 (4.5)	21 (2.9)	108 (3.7)	.653
Non-Hispanic Multiple	68 (1.6)	13 (2.4)	9 (1.2)	46 (1.6)	.357
Regione
Northeast	811 (18.2)	339 (52.2)	48 (6.3)	424 (13.9)	<.001
Midwest	986 (22.1)	109 (16.8)	87 (11.3)	790 (26.0)	<.001
South	1558 (35.0)	144 (22.2)	430 (56.1)	984 (32.4)	.864
West	1101 (24.7)	57 (8.8)	202 (26.3)	842 (27.7)	<.001
Any comorbidity reported	1688 (37.8)	239 (36.8)	293 (38.1)	1156 (37.9)	.688
Obesityf	1121 (25.1)	148 (22.8)	189 (24.6)	784 (25.7)	.117
Chronic lung disease including asthma	428 (9.6)	75 (11.6)	81 (10.5)	272 (8.9)	.022
Other congenital malformations	144 (3.2)	20 (3.1)	27 (3.5)	97 (3.2)	.962
Neurologic/neuromuscular disorderg	108 (2.4)	17 (2.6)	24 (3.1)	67 (2.2)	.285
Congenital heart disease	94 (2.1)	11 (1.7)	22 (2.9)	61 (2.0)	.940
Seizures	91 (2.0)	16 (2.5)	19 (2.5)	56 (1.8)	.198
Diabetesh	39 (0.9)	5 (0.8)	6 (0.8)	28 (0.9)	.656
Immunosuppressive disorder/malignancy	36 (0.8)	4 (0.6)	8 (1.0)	24 (0.8)	n/ai
Sickle cell disease	24 (0.5)	5 (0.8)	4 (0.5)	15 (0.5)	n/ai
Other conditionsj	112 (2.5)	9 (1.4)	31 (4.0)	72 (2.4)	.716

The 3 time periods were (1) 19 February 2020 through 28 June 2020, (2) 29 June 2020 through 17 October 2020, and (3) 18 October 2020 through 31 July 2021.

Abbreviations: IQR, interquartile range; MIS-C, multisystem inflammatory syndrome in children.

a

P values from Cochran-Armitage test for trend for categorical variables and Jonckheere-Terpstra trend test for continuous variables.

b

Percentages calculated among 4466 persons with known age.

c

Percentages calculated among 4464 persons with known sex.

d

Percentages calculated among 4207 persons with known race and ethnicity data. Racial and ethnic classifications followed CDC’s Office of Minority Health and Health Equity (OMHHE) guidance. Non-Hispanic ethnicity was assumed if Hispanic ethnicity was not noted. Hispanic ethnicity was top-coded over White, Black, and Asian race. Because of small patient sizes, Native Hawaiian/Pacific Islander and American Indian/Alaskan Native populations were reported as such, regardless of ethnicity.

e

Percentages calculated among 4456 persons; territories not included in regional count.

f

By either clinician diagnosis of obesity or body mass index-based obesity; calculated only in children > 2 years.

g

Includes cerebral palsy, developmental disabilities, autism, and other neurologic conditions.

h

Includes type 1 diabetes, type 2 diabetes, and unspecified diabetes.

i

Unable to assess significance because of small cell size.

j

Includes chronic kidney disease, Down syndrome, failure to thrive, obstructive sleep apnea, prematurity, and other reported conditions.

All patients with MIS-C had reported fever as indicated by the case definition. The median reported duration of fever was 5 days (IQR: 4–7 days). The most common additional signs and symptoms included abdominal pain (68.5%), vomiting (66.6%), rash (55.2%), conjunctival injection (55.4%), diarrhea (53.8%), and hypotension (51.7%) (Table 2). Seventy-four percent of MIS-C patients reported mucocutaneous involvement (eg, skin rash, mucocutaneous lesions, and/or conjunctivitis).

Table 2.

Clinical Features of Patients with MIS-C by Wave, United States, February 2020 to July 2021.

Characteristic	All MIS-C cases (n = 4470), no. (%)	MIS-C Onset, no. (%)
		Wave 1 (n = 649)	Wave 2 (n = 769)	Wave 3 (n = 3052)	Test for Trend P-Value a
Signs and symptoms
Fever (>38°C or subjective)	4470 (100.0)	649 (100.0)	769 (100.0)	3052 (100.0)	N/A
Abdominal pain	3061 (68.5)	407 (62.7)	517 (67.2)	2137 (70.0)	<.001
Vomiting	2975 (66.6)	395 (60.9)	512 (66.6)	2068 (67.8)	.002
Conjunctival injection	2476 (55.4)	334 (51.5)	437 (56.8)	1705 (55.9)	.102
Rash	2467 (55.2)	360 (55.5)	423 (55.0)	1684 (55.2)	.928
Diarrhea	2463 (53.8)	349 (53.8)	415 (54.0)	1699 (55.7)	.288
Hypotension	2313 (51.7)	329 (50.7)	411 (53.4)	1573 (51.5)	.982
Headache	1993 (44.6)	222 (34.2)	331 (43.0)	1440 (47.2)	<.001
Myalgia	1332 (29.8)	150 (23.1)	225 (29.3)	957 (31.4)	<.001
Cough	1313 (29.4)	191 (29.4)	229 (29.8)	893 (29.3)	.863
Shortness of breath	1252 (28.0)	172 (26.5)	239 (31.1)	841 (27.6)	.854
Mucocutaneous lesions	1008 (22.6)	197 (30.4)	165 (21.5)	646 (21.2)	<.001
Neck pain	985 (22.0)	85 (13.1)	158 (20.5)	742 (24.3)	<.001
Chest pain	631 (14.1)	88 (13.6)	108 (14.0)	435 (14.3)	.649
Periorbital edema	533 (11.9)	47 (7.2)	94 (12.2)	392 (12.8)	<.001
Cervical lymphadenopathy	503 (11.3)	87 (13.4)	82 (10.7)	334 (10.9)	.131
Altered mental status	497 (11.1)	65 (10.0)	100 (13.0)	332 (10.9)	.991
Syncope	216 (4.8)	28 (4.3)	35 (4.6)	153 (5.0)	.399
Severe cardiovascular involvement	3556 (79.6)	503 (77.5)	615 (80.0)	2438 (79.9)	.238
Elevated troponin	2356 (52.7)	279 (43.0)	431 (56.0)	1646 (53.9)	<.001
Shock/receipt of vasopressors	2018 (45.1)	312 (48.1)	364 (47.3)	1342 (44.0)	.025
BNP or NT-pro BNP ≥ 1000 pg/mL	1617 (36.2)	224 (34.5)	259 (33.7)	1134 (37.2)	.084
Cardiac dysfunctionb	1296 (30.9)	238 (40.1)	191 (27.4)	867(29.8)	<.001
Pericardial effusion/pericarditis	989 (22.1)	159 (24.5)	171 (22.2)	659 (21.6)	.120
Coronary artery aneurysm/dilatationc	700 (16.7)	110 (18.5)	114 (16.4)	476 (16.4)	.263
Myocarditisd	665 (14.6)	139 (21.4)	115 (15.0)	411 (13.5)	<.001
Arrhythmiae	331 (7.4)	49 (7.6)	71 (9.2)	211 (6.9)	.215
Congestive heart failure	240 (5.4)	45 (6.9)	29 (3.8)	166 (5.4)	.469
ECMO	69 (1.5)	17 (2.6)	10 (1.3)	42 (1.4)	.046
Severe hematologic involvement	2663 (59.6)	290 (44.7)	464 (60.3)	1909 (62.5)	<.001
Thrombocytopeniaf	1890 (42.3)	214 (33.0)	352 (45.8)	1324 (43.4)	<.001
Lymphopeniag	1580 (35.3)	150 (23.1)	233 (30.3)	1197 (39.2)	<.001
Neutropeniah	47 (1.1)	4 (0.6)	8 (1.0)	35 (1.1)	n/ai
Deep vein thrombosis/pulmonary embolism	38 (0.9)	6 (0.9)	5 (0.7)	27 (0.9)	.899
Severe respiratory involvement	1962 (43.9)	295 (45.5)	412 (53.6)	1255 (41.1)	<.001
Pneumoniaj	1044 (23.4)	172 (26.5)	236 (30.7)	636 (20.8)	<.001
Pleural effusionk	954 (21.3)	121 (18.6)	202 (26.3)	631 (20.7)	.887
Oxygen, high flow nasal cannula	753 (16.8)	107 (16.5)	162 (21.1)	484 (15.9)	.130
Invasive mechanical ventilation (intubation)	419 (9.4)	82 (12.6)	96 (12.5)	241 (7.9)	<.001
Noninvasive mechanical ventilation	368 (8.2)	59 (9.1)	75 (9.8)	234 (7.7)	.091
Acute respiratory distress syndrome	261 (5.8)	45 (6.9)	64 (8.3)	152 (5.0)	.004
Severe gastrointestinal involvementl	1133 (25.3)	133 (20.5)	176 (22.9)	824 (27.0)	<.001
Mesenteric adenitis	540 (28.9)	45 (21.0)	70 (21.7)	425 (31.9)	<.001
Free fluid	460 (24.6)	55 (25.7)	80 (24.8)	325 (24.4)	.677
Hepatomegaly/splenomegalym	198 (10.6)	26 (12.1)	35 (10.9)	137 (10.3)	.410
Colitis/enteritis	191 (10.2)	27 (12.6)	33 (10.2)	131 (9.8)	.250
Cholecystitis/gallbladder abnormalities	137 (7.3)	15 (7.0)	29 (9.0)	93 (7.0)	.605
Appendicitis/appendiceal changes	77 (4.1)	5 (2.3)	17 (5.3)	55 (4.1)	.506
Severe renal involvement	908 (20.3)	133 (20.5)	171 (22.2)	604 (19.8)	.386
Acute kidney injury	849 (19.0)	106 (16.3)	166 (21.6)	577 (18.9)	.449
Renal failure	148 (3.3)	39 (6.0)	27 (3.5)	82 (2.7)	<.001
Dialysis	42 (0.9)	5 (0.8)	4 (0.5)	33 (1.1)	n/ai
Severe neurologic involvement	382 (8.5)	45 (6.9)	66 (8.6)	271 (8.9)	.133
Meningitis	238 (5.3)	26 (4.0)	45 (5.9)	167 (5.5)	.236
Encephalopathy	168 (4.0)	23 (4.8)	287 (3.8)	117 (4.0)	.524
Stroke	29 (0.6)	8 (1.2)	2 (0.3)	19 (0.6)	n/ai
Any mucocutaneous involvement	3323 (74.3)	472 (72.7)	575 (74.8)	2276 (74.6)	.411
Laboratory markers of inflammation
Elevated C-reactive proteinn	4033 (99.3)	563 (99.5)	669 (99.3)	2801 (99.3)	.788
Elevated d-dimer	4125 (92.3)	572 (88.1)	717 (93.2)	2836 (92.9)	<.001
Elevated ferritino	2982 (87.2)	406 (88.3)	527 (90.5)	2049 (86.2)	.041
Elevated IL-6p	692 (80.8)	147 (96.1)	145 (84.8)	400 (75.2)	<.001
Elevated fibrinogenq	2761 (80.3)	368 (78.0)	456 (80.4)	1937 (80.7)	.202
SARS-CoV-2 results
Serologyr positive	3905 (97.7)	512 (97.3)	626 (96.5)	2767 (98.1)	.079
RT-PCRs positive	2225 (52.6)	331 (52.0)	446 (60.1)	1448 (50.7)	.040
Length of stay
Total days in hospital, days, median (IQR)t	5 (4–8)	6 (4–10)	6 (4–8)	5 (4–8)	<.001
ICU admission, days, median (IQR)u	4 (2–6)	5 (3–7)	4 (2–6)	3 (2–5)	<.001

The 3 time periods were (1) 19 February 2020 through 28 June 2020, (2) 29 June 2020 through 17 October 2020, and (3) 18 October 2020 through 31 July 2021.

Abbreviations: BNP, brain natriuretic peptide; ECMO, extracorporeal membrane oxygenation; ICU, intensive care unit; IL-6, interleukin 6; IQR, interquartile range; MIS-C, multisystem inflammatory syndrome in children; NT-proBNP, N-terminal pro b-type natriuretic peptide; RT-PCR, reverse transcription polymerase chain reaction; SARS-CoV-2, severe acute respiratory syndrome coronavirus 2.

P values from Cochran-Armitage test for trend for categorical variables and Jonckheere-Terpstra trend test for continuous variables.

Includes specified left ventricular dysfunction (n = 1151) and right ventricular dysfunction (n = 304); percentages calculated among 4198 persons with an echocardiogram performed (Wave 1 n = 594; Wave 2 n = 697; Wave 3 n = 2907).

Percentages calculated among 4198 persons with an echocardiogram performed (Wave 1 n = 594; Wave 2 n = 697; Wave 3 n = 2907).

Indicated on case report form.

Includes ventricular arrhythmia (n = 116), supraventricular arrhythmia (n = 107), atrioventricular block (n = 23), and other electrocardiogram changes (n = 86).

Thrombocytopenia was collected under signs and symptoms or calculated from laboratory results as platelets <150/µL.

Lymphopenia was defined as lymphocyte count <4500 cells/µl if age <8 months or <1500 cells/µl if age ≥8 months.

Neutropenia was defined as absolute neutrophil count <500 cells/μL.

Unable to assess significance because of small cell size.

Information about pneumonia was collected on the case report form under signs and symptoms, complications, or chest imaging.

Percentages calculated among 3522 persons with chest imaging performed (Wave 1 n = 444; Wave 2 n = 664; Wave 3 n = 2414).

Percentages calculated among 1868 persons with an abdominal imaging performed (Wave 1 n = 214; Wave 2 n = 322; Wave 3 n = 1332).

Includes hepatosplenomegaly.

Defined by C-reactive protein ≥0.6 mg/dL. Percentages calculated among 4060 persons with C-reactive protein testing results available (Wave 1 n = 566; Wave 2 n = 674; Wave 3 n = 2820).

Defined by ferritin ≥300 ng/mL. Percentages calculated among 3418 persons with ferritin testing results available (Wave 1 n = 460; Wave 2 n = 582; Wave 3 n = 2049).

Defined by IL-6 >1.8 pg/mL. Percentages calculated among 856 persons with IL-6 testing results available (Wave 1 n = 153; Wave 2 n = 171; Wave 3 n = 532).

Defined by fibrinogen >400 mg/dL. Percentages calculated among 3438 persons with fibrinogen testing results available (Wave 1 n = 472; Wave 2 n = 567; Wave 3 n = 2399).

Percentages calculated among 3997 patients with serology testing.

Percentages calculated among 4233 patients with PCR testing.

Calculated based on hospitalization duration available for 3988 patients.

Calculated based on ICU duration available for 1810 patients.

Many MIS-C patients had severe cardiovascular involvement (79.6%) including cardiac dysfunction (30.9%), particularly left ventricular dysfunction (27.4%), coronary artery aneurysm or dilatation (16.7%), myocarditis (14.6%), and congestive heart failure (5.4%) (Table 2). Other severe cardiovascular manifestations included shock or receipt of vasopressor medication (45.1%), elevated cardiac laboratory tests (ie, troponin [52.7%] or brain natriuretic peptide ≥1000 pg/mL [36.2%]) and receipt of ECMO support (1.5%). Sixty percent of MIS-C patients experienced severe hematologic involvement, including thrombocytopenia (42.3%) and lymphopenia (35.3%). Severe respiratory involvement (43.9%), including pneumonia (23.4%), pleural effusion (21.3%), and acute respiratory distress syndrome (5.8%), was reported less frequently. Receipt of supplemental oxygen, including high-flow nasal cannula (16.8%), noninvasive mechanical ventilation (8.2%), and invasive mechanical ventilation (9.4%) were also less common. One quarter of MIS-C patients had signs of severe gastrointestinal involvement (25.3%), with abnormal findings on abdominal imaging including free fluid (24.3%), colitis/enteritis (10.2%), mesenteric adenitis (28.9%), hepatomegaly/splenomegaly (10.6%), and inflammation of the gallbladder (7.3%), and appendix (4.1%). Severe renal complications (20.3%), including acute kidney injury (19.0%), renal failure (3.3%) and receipt of dialysis (0.9%), and severe neurologic complications (8.5%), including meningitis (5.3%), encephalopathy (4.0%), and stroke (0.6%), were rare (Table 2).

Most MIS-C patients underwent radiography including echocardiography (94%) and chest X-ray or computed tomography (71%), and 42% underwent abdominal imaging. Over half of those for whom these radiographic studies were performed had abnormal findings, including 59% of those who underwent echocardiography, 55% of those who underwent chest imaging, and 61% of those who underwent abdominal imaging. Among laboratory markers of inflammation, C-reactive protein (CRP) (99.3%) and ferritin (87.2%) were frequently elevated among patients tested. Nearly all of the 3997 MIS-C patients who had serologic testing for SARS-CoV-2 antibodies performed tested positive (97.7%); 52.6% of the 4233 MIS-C patients who had SARS-CoV-2 RT-PCR performed tested positive (Table 2). Vaccination status was reported for 4% of MIS-C patients, only 0.4% of whom were vaccinated.

Characteristics and Outcomes Over Time

There were three defined waves of MIS-C illness, with most cases (68.3%) occurring during the third wave, compared with the first (17.2%) and second (14.5%) (Table 1; Figure 1). By region, the Northeast had the highest proportion of cases in the first wave (52.2%) followed by a significant decrease over the remaining waves (P < .001) (Table 1). By the third wave, cases were spread similarly across the Midwest (26%), South (32%), and West (27%), whereas the Northeast reported only 14% of cases.

The proportion of patients aged 12–15 years significantly increased over time (P=.005) (Table 1). The proportion of males and of patients who were non-Hispanic White increased significantly with successive waves (P < .001 for each) and were 62% and 35%, respectively, by the third wave. The proportion of MIS-C patients with each reported underlying medical condition did not differ significantly over time.

Figure 2 illustrates organ system involvement (defined in Supplement 2) over the three MIS-C waves. The proportion of patients reported to have severe hematologic and gastrointestinal involvement increased over the 3 waves (P < .001 each). There were no other consistent trends for the remaining organ systems. Symptoms reported more frequently over time included abdominal pain, headache, myalgia, and neck pain (P < .001) (Table 2, Supplement 5). Some cardiovascular complications, including cardiac dysfunction, myocarditis, and shock/use of vasopressor medications were significantly less frequent with successive waves. The proportion of patients receiving ECMO support decreased over time (P < .001) (Table 2).

Figure 2.

Proportion of MIS-C Patients by Organ System Involvement over 3 Time Periods, United States, February 2020–July 2021. The 3 MIS-C waves were (1) 19 February 2020 through 28 June 2020, (2) 29 June 2020 through 17 October 2020, and (3) 18 October 2020 through 31 July 2021. ∗Represents significance using Cochran-Armitage test for trend across the 3 waves (P < .001). P-values for severe cardiovascular (P = .238), any mucocutaneous (P = .411), severe renal (P = .386), and severe neurologic (P = .133) were not significant using Cochran-Armitage test for trend across the 3 waves. Abbreviations: MIS-C, multisystem inflammatory syndrome in children.

The proportion of those with severe respiratory complications peaked during the second wave, with pneumonia, pleural effusion, and acute respiratory distress syndrome the highest in the second wave (31%, 26%, and 8%, respectively). However, the proportion of patients receiving invasive mechanical ventilation trended downward over the study period (P < .001) (Table 2).

The proportion of MIS-C patients with a reported elevated troponin, d-dimer and presence of lymphopenia and thrombocytopenia increased over time (P < .001). The proportion of patients who had elevated CRP and elevated fibrinogen among those tested remained similarly high across the MIS-C pandemic waves. The number of MIS-C patients testing positive for SARS-CoV-2 on RT-PCR was highest during the second wave while positive serologic testing for SARS-CoV-2 antibody remained high across all 3 waves (Table 2).

Overall, MIS-C patients were hospitalized for a median of 5 days (IQR: 4–8 days). Median ICU stay was 4 days (IQR: 2–6 days). Duration of hospitalization and ICU stay decreased over time (each P < .001). Intravenous immunoglobulin (IVIG) (84.4%), steroids (76.7%), and immune modulators (20.9%) were commonly administered to patients during hospitalization. The proportion of cases that received IVIG and steroids increased significantly (each P < .001) (Figure 3). Sixty-three percent of MIS-C patients were admitted to the ICU or received ICU-level care. Overall, 37 (0.8%) patients died, and the case fatality ratio among MIS-C patients decreased over the three waves (P < .001).

Figure 3.

Proportion of MIS-C Patients by Treatment and Outcomes over 3 Time Periods, United States, February 2020 to July 2021. The 3 MIS-C waves were (1) 19 February 2020 through 28 June 2020, (2) 29 June 2020 through 17 October 2020, and (3) 18 October 2020 through 31 July 2021.∗Represents significance using Cochran-Armitage test for trend across the 3 waves (P < .001). P-Values for immune modulators (P = .513) and ICU admission (P = .056) were not significant using Cochran-Armitage test for trend across the 3 waves. Abbreviations: ICU, intensive care unit; IVIG, intravenous immune globulin; MIS-C, multisystem inflammatory syndrome in children.

DISCUSSION

This study describes the largest cohort of MIS-C patients to date, including patients from nearly all reporting jurisdictions within the United States since MIS-C national surveillance began. MIS-C is an important complication of COVID-19 in children, with half of reported cases occurring among children aged 5–13 years. Over half of MIS-C patients were of Hispanic ethnicity or Black race, particularly in the first 2 MIS-C waves, similar to findings from previous studies [11, 29–31]. The proportion of non-Hispanic White patients significantly increased over the course of the pandemic; similar racial and ethnic trends have been described in studies of hospitalized COVID-19 patients, which suggest that increased COVID-19 incidence among White persons and regional and temporal patterns may be contributors [32, 33]. Further investigation into the impact of spatiotemporal trends on the racial and ethnic distribution of MIS-C patients over time is warranted given these findings.

The MIS-C patient population had a similar proportion of children with preexisting underlying medical conditions across the 3 MIS-C waves. There were increasing trends reported in severe hematologic and severe gastrointestinal involvement; other organ system involvement did not differ substantially. These increasing trends may be the result of changes in reporting completeness (eg, lymphopenia, thrombocytopenia) or changing testing practices for case identification (eg, increase in receipt of abdominal imaging) [34]. For example, the proportion of patients who underwent abdominal imaging increased from 33.0% to 41.9% to 43.6% over the 3 waves (P < .001). Reporting of other clinical findings may reflect improved awareness of MIS-C clinical features (eg, neck pain) [35]. Thus, increasing trends may or may not reflect an actual increase in the proportion of children experiencing each sign or symptom.

The proportion of cases with cardiac dysfunction and myocarditis, conditions associated with severe outcomes of MIS-C [3], declined after the first wave. The duration of hospitalization and death also significantly decreased over time; these decreases in case fatality were previously described [26]. These observed decreases could be associated with several factors including reporting of milder MIS-C cases or earlier case identification as awareness of the condition increased, improvements in clinical management of MIS-C, changes in the SARS-CoV-2 virus, or some other combination of factors. For example, a previous analysis showed that treatment of MIS-C with IVIG plus glucocorticoids was associated with a lower risk of new or persistent cardiovascular dysfunction than treatment with IVIG alone [36]. We did find increases in the use of medications including IVIG and steroids over the surveillance period. However, MIS-C national surveillance does not collect information related to the timing of treatments relative to clinical findings; therefore, it is not possible to estimate therapeutic effectiveness from these observational data.

Previous studies have described a respiratory subgroup of MIS-C cases phenotypically similar to patients with acute COVID-19 [4, 37]. This study found that the proportion of MIS-C patients with severe respiratory involvement and with a positive SARS-CoV-2 RT-PCR test followed similar patterns and were highest during the second wave of MIS-C illness. These findings may support the presence of a distinct MIS-C respiratory phenotype or the unintentional misclassification of patients with acute COVID-19 because of an overly sensitive case definition. It is not clear why this phenotype would increase in the second wave and then decrease in the third, although this may be influenced by changing SARS-CoV-2 testing and MIS-C reporting practices.

This study has several limitations. Not all MIS-C cases are reported to national surveillance, resulting in under-ascertainment. Data from April through June 2020 suggest that national surveillance detected approximately two-thirds of MIS-C cases in 7 jurisdictions when data were compared with an active hospital-based surveillance network [11]. Case identification, including diagnosis, testing, and management of patients can vary by jurisdiction. MIS-C case report forms require medical chart abstractions which are resource- and time-intensive and can contribute to delays or gaps in reporting, particularly during times of public health emergency and competing demands, and availability of medical records can vary by jurisdiction. The case report form does not include specific definitions of comorbidities, clinical signs/symptoms, or complications, which may result in subjective interpretation of questions and contribute to misclassification. Additionally, the staff abstracting medical records (eg, health department staff, clinicians) could vary in training and experience by jurisdiction and result in lack of standardized reporting practices. MIS-C is not a nationally notifiable illness; reporting of cases is voluntary, and participation varies by jurisdiction and over time. Inconsistency in completion of case reporting forms might have affected data completeness. Vaccination status was not routinely collected until 21 May 2021 and was not analyzed because of insufficient data. Access to SARS-CoV-2 testing at time of onset might have varied by region, by hospital, and over time. Finally, the CDC MIS-C case definition is broad, which might have led to the unintentional inclusion of patients with a history of COVID-19 experiencing other acute inflammatory illnesses such as severe acute COVID-19, Kawasaki disease, and toxic shock syndrome.

In conclusion, MIS-C is a rare but severe complication of COVID-19 in children and adolescents, with a clinical presentation that has remained similar over the first year and a half of national MIS-C surveillance and outcomes such as length of hospitalization and death that have shown improvement over time. The ongoing transmission of SARS-CoV-2 and the emergence of potentially more severe and highly transmissible variants, such as the Delta variant, are likely to contribute to increased incidence of MIS-C following increased SARS-CoV-2 transmission in the United States. This US surveillance summary describes clinical characteristics of MIS-C patients prior to widespread Delta variant circulation, and before vaccination was authorized and widespread among children and adolescents. These national surveillance data provide a baseline for future comparisons of the characteristics of MIS-C patients infected with B.1.617.2 (Delta) or other SARS-CoV-2 variants and for measuring the impact of increased COVID-19 vaccination coverage among children and adolescents.

Supplementary Data

Supplementary materials are available at Clinical Infectious Diseases online. Consisting of data provided by the authors to benefit the reader, the posted materials are not copyedited and are the sole responsibility of the authors, so questions or comments should be addressed to the corresponding author.

Notes

Acknowledgments. The authors thank Sherri L. Davidson, Rachel Tulibagenyi, Melanie C. Roderick (Alabama Department of Public Health), Jessica M. de Jarnette (California Department of Public Health), Moon Kim, Lauren E. Finn (Los Angeles County Department of Public Health), Lynn E Sosa, Joanne G. Colletti (Connecticut Department of Public Health), Kossia Y. M. Dassie, Nkembi L. Bianda (D.C. Health), Monika N. Bray, Mary B. Fukushima (Centers for Disease Control and Prevention; Booz Allen Hamilton), Nicholas R. Patsy (Centers for Disease Control and Prevention; Shine Systems), Grace A. Collins (Florida Department of Health), Shaunta N. Rutherford, Carolyn M. Adam (Georgia Department of Public Health), Luis Vela, Kathryn A. Turner, Scott C. Hutton (Idaho Department of Health and Welfare), Madison A. Asbell, Andzelika E. Rzucidlo, Mary-Elizabeth Steppig (Indiana Department of Health), Oluwakemi O. Oni, Caitlin Pedati (Iowa Department of Public Health), Justin L. Blanding, Sujata Mallik (Kansas Department of Health and Environment), Kevin B. Spicer (Centers for Diseases Control and Prevention; Kentucky Department for Public Health), Stacy L. Davidson (Kentucky Department for Public Health), Anna Krueger, Chloe S. Manchester (Maine Center for Disease Control and Prevention), Bryce L. Spiker, Justin Henderson (Michigan Department of Health and Human Services), Kathryn J. Como-Sabetti (Minnesota Department of Health), Robin M. Williams (University of Nebraska-Lincoln; Nebraska Department of Health and Human Services), Jessica Pahwa (Nebraska Department of Health and Human Services), Ali Garcia (Nevada Department of Health and Human Services), Tara Fulton, Stella Tsai (New Jersey Department of Health), Eirian Coronado (New Mexico Department of Health), Jessica A. Kumar, Bridget J. Anderson, Faud Ishaq (New York State Department of Health), Kathleen Heather Reilly, Maura K. Lash (New York City Department of Health and Mental Hygiene), Levi Schlosser (North Dakota Department of Health), Melissa Sutton (Oregon Health Authority), Nottasorn Plipat, Allison H. Longenberger (Pennsylvania Department of Health). Abby L. Berns, Karen Luther (Rhode Island Department of Health), Sanet Torres-Torres, Mónica M. Allende-Quirós (Puerto Rico Department of Health), Hani M. Mohamed, Rachel A. Radcliffe (South Carolina Department of Health & Environmental Control), Sara B. Bowman (South Dakota Department of Health), Amanda L. Hartley (Tennessee Department of Health), Ariel B. Morales (Texas Department of State Health Services), Bree Barbeau, Karen E. James, Erin M. Treemarcki (Utah Department of Health), Sabine A. Pierre-Louis (Virginia Department of Health), Marisa D’Angeli, Amanda Dodd, Kimberly Carlson (Washington State Department of Health), Lindsey J. Mason, Lesley A. Roush (West Virginia Department of Health and Human Resources), Thomas E. Haupt (Wisconsin Department of Health Services), Elizabeth A. Walker-Short (Wyoming Department of Health).

Disclaimer. The findings and conclusions in this report are those of the authors and do not necessarily represent the views of the US Centers for Disease Control and Prevention.

Financial support. All participating jurisdictions received financial support from the Centers for Disease Control and Prevention through the Epidemiology and Laboratory Capacity for Prevention and Control of Emerging Infectious Diseases cooperative agreement.

Potential conflicts of interest. A. R. L. reports grants through CDC Foundation during this study. J. R.C. S. received support through an appointment to the Applied Epidemiology Fellowship Program administered by the Council of State and Territorial Epidemiologists (CSTE) and funded by the Centers for Disease Control and Prevention (CDC) Cooperative Agreement Number 1NU38OT000297-03-00. KLH reports Project O Grant for the present manuscript. C. P. B. reports CDC Epidemiology and Laboratory Capacity Grant CDC-RFA-CK19-1904 outside of the submitted work. M. E. O. reports payment made to the institution where they have clinical responsibilities from the National Institutes of Health (NIH) (MUSIC Study) outside of the submitted work. R. K. reports being funded through the CDC ELC Core Cooperative Agreement as well as CDC PHEP grant during and outside of the conduct of the study. All other authors report no potential conflicts. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.

aMIS-C Surveillance Authorship Group. Gloria E. Anyalechi, MD, MPH¹; Anna Bowen, MD, MPH¹; Tuyen Do¹; Paul A. Gastañaduy, MD, MPH¹; Katherine Lindsey, MPH¹; Sancta B. St. Cyr, MD¹; Ramandeep Kaur, PhD, MPH, BSN²; Xandy Peterson Pompa, MPH³; Chloe E. Le Marchand, MD, MSc⁴; Jason Robert C. Singson, MPH^4,5; Shannon C. O’Brien, MD, MPH⁶; Ann M. Schmitz, DVM,^7,8; Carola I. Torres Díaz, MPH⁷; Walaa M. Elbedewy, MBBCh, MPH, MPA⁹; Melissa J. Tobin-D’Angelo, MD, MPH⁹; Heather D. Reid, BS, CHES¹⁰; Marielle J. Fricchione, MD¹¹; Sara J. Hallyburton, MPH¹²; Gillian Richardson, MPH¹³; Julie P. Hand, MSPH¹³; Dylan H. Leach, MPH¹⁴; Cole P. Burkholder, MPH¹⁵; Sarah Lim, MBBCh, MPH¹⁶; Deepam Thomas, MPH¹⁷; Donna L. Gowie, AAS¹⁸; Elizabeth M. Dufort, MD¹⁸; Ellen H. Lee, MD,MPH¹⁹; Ayotola A. Falodun, MD, MPH²⁰; Courtney M. Dewart, PhD, MPH, RN^21,22; Zachary J. Colles, MPH²¹; Jennifer L. Wallace, MD, MS²³; LaKita D. Johnson, MPH²⁴; Kristina L. Herring, BS, ASN²⁵; Andrea R. Liptack, MSN^26,27

¹CDC COVID-19 Response Team, Centers for Disease Control and Prevention, Atlanta, GA, USA;²Alabama Department of Public Health, Montgomery, Alabama, USA; ³ Arizona Department of Health Services, Phoenix, Arizona, USA; ⁴California Department of Health, Richmond, California, USA; ⁵Council of State and Territorial Epidemiologists, Atlanta, Georgia, USA; ⁶Colorado Department of Public Health and Environment, Denver, Colorado, USA; ⁷Florida Department of Health, Tallahassee, Florida, USA; ⁸Division of State and Local Readiness, Center for Preparedness and Response, Centers for Disease Control and Prevention, Atlanta, Georgia, USA; ⁹Georgia Department of Public Health, Atlanta, GA, USA; ¹⁰Illinois Department of Public Health, Springfield, IL, USA; ¹¹Chicago Department of Public Health, Chicago, Illinois, USA; ¹²Indiana Department of Health, Indianapolis, Indiana, USA; ¹³Louisiana Department of Health, New Orleans, Louisiana, USA; ¹⁴Massachusetts Department of Public Health, Boston, Massachusetts, USA; ¹⁵Michigan Department of Health and Human Services, Lansing, Michigan, USA; ¹⁶Minnesota Department of Health, St Paul, Minnesota, USA; ¹⁷New Jersey Department of Health, Trenton, New Jersey, USA; ¹⁸New York State Department of Health, Albany, New York, USA; ¹⁹New York City Department of Health and Mental Hygiene, Long Island, New York, USA; ²⁰ North Carolina Department of Health and Human Services, Raleigh, NC, USA; ²¹Ohio Department of Health, Columbus, Ohio, USA; ²²Epidemic Intelligence Service, Centers for Disease Control and Prevention, Atlanta, Georgia, USA; ²³Pennsylvania Department of Health, Harrisburg, Pennsylvania, USA; ²⁴South Carolina Department of Health and Environmental Control, Columbia, South Carolina, USA; ²⁵Tennessee Department of Health, Nashville, Tennessee, USA; ²⁶Wisconsin Department of Health Services, Madison, Wisconsin, USA; ²⁷CDC Foundation, Atlanta, Georgia, USA;

Author notes

MIS-C Surveillance Authorship Group listed in the Acknowledgements.