exclude: true --- class: top left hide-count background-image: url(img/sarscov2_cdc_blur.png) background-size: cover .fade-in[ # <span style='color:white'>Pediatric SARS-CoV-2 Seroprevalence</span> ] .fade-in[ .callout-url-leftcorner[ <span style='font-size:0.75em;font-weight:bold;color:white'> Brendan J. Kelly, MD, MS Infectious Diseases & Epidemiology University of Pennsylvania </span> <span style='font-size:0.5em;font-weight:bold;color:white;'> 31 March 2021 </span> ] ] --- background-image: url(img/tonshoff_title.png) background-size: 80% background-position: 50% 40% --- ## What to make of SARS-CoV-2 antibodies in kids? .pad-left[ - Seroprevalence studies are important & and <u>hard to do in kids</u> - asymptomatic disease is more common among kids than adults - systematic under-reporting of pediatric cases - implications for 🔥hotly🔥 debated policy decisions 🏫 ] --- ## What to make of SARS-CoV-2 antibodies in kids? .pad-left[ - Context is everything: - rapidly evolving & spatially heterogeneous pandemic - school policies have large effects on kids' mobility - **Explicit causal/process models** are essential: - how do we interpret conflicting studies of pediatric populations? - how might different causal models account for observations? ] --- class: full-screen hide-count <div class="grid-3-1"> <div class="extension-tile gridset-a animated flipInY"><div>Tönshoff B<br>et al<br>Findings</div></div> </div> --- layout: false class: full-screen hide-count <div class="grid-3-1"> <div class="extension-tile gridset-a"><div>Tönshoff B<br>et al<br>Findings</div></div> <div class="extension-tile gridset-b animated flipInX"><div>Parallel<br>Studies</div></div> </div> --- layout: false class: full-screen hide-count <div class="grid-3-1"> <div class="extension-tile gridset-a"><div>Tönshoff B<br>et al<br>Findings</div></div> <div class="extension-tile gridset-b"><div>Parallel<br>Studies</div></div> <div class="extension-tile gridset-c animated bounceInDown"><div>What Causes<br>SARS-CoV-2<br>Spread?</div></div> </div> --- exclude: true layout: false class: full-screen hide-count <div class="grid-3-2"> <div class="extension-tile gridset-a"><div>Case from 1981</div></div> <div class="extension-tile gridset-b"><div>Prevalence</div></div> <div class="extension-tile gridset-c"><div>Incidence</div></div> <div class="extension-tile gridset-d animated fadeInRight"><div>Inference from<br>Exposures &<br>Outcomes</div></div> <div class="extension-tile gridset-e animated bounceInUp"><div>Relative Risk<br>&<br>Odds Ratios</div></div> <div class="extension-tile gridset-f animated slideInLeft"><div>Data Types<br>&<br>Distributions</div></div> </div> --- exclude: true class: full-screen hide-count <div class="grid-3-2"> <div class="extension-tile gridset-a animated flipInY"><div>Case from 1981</div></div> <div class="extension-tile gridset-b animated flipInX"><div>Prevalence</div></div> <div class="extension-tile gridset-c animated flipInY"><div>Incidence</div></div> <div class="extension-tile gridset-d animated flipInX"><div>Inference from<br>Exposures &<br>Outcomes</div></div> <div class="extension-tile gridset-e animated flipInY"><div>Relative Risk<br>&<br>Odds Ratios</div></div> <div class="extension-tile gridset-f animated flipInX"><div>Data Types<br>&<br>Distributions</div></div> </div> --- layout: false class: full-screen hide-count <div class="grid-3-1"> <div class="extension-tile gridset-a"><div>Tönshoff B<br>et al<br>Findings</div></div> <div class="extension-tile gridset-grey"><div>Parallel<br>Studies</div></div> <div class="extension-tile gridset-grey"><div>What Causes<br>SARS-CoV-2<br>Spread?</div></div> </div> --- ## Population Studied .pad-left[ - Southwest Germany April-May 2020 - Freiburg, Heidelberg, Tübingen, and Ulm - enrollment April 22 to May 15, 2020 - inclusion: child 1-10 years, live with parent in state of Baden-Württemberg - exclusion: severe congenital disease, immunodeficiency, <u>prior known COVID-19</u> - <u>schools closed except "exceptional child care" (580 / 2443 = 24%)</u> ] --- ## Cross-Sectional Study Design .pad-left[ - 2482 adult-child pairs - primary analysis: mixed effects logistic regression (Model 1): `$${positive} \sim Bernoulli(p)$$` `$$logit(p) = \beta_{parent} \cdot parent + \alpha + \alpha_{family} + \epsilon$$` - secondary analyses (Models 2 & 3): `$${positive} \sim \beta_{parent} \cdot parent + \beta_{young} \cdot young + \alpha + \alpha_{family}$$` `$${positive} \sim \beta_{parent} \cdot parent + \beta_{young} \cdot young + \beta_{siblings} \cdot siblings \\ + \beta_{sex} \cdot sex + \beta_{daycare} \cdot daycare + \beta_{contact} \cdot contact + \alpha + \alpha_{family}$$` ] --- ## Seroprevalence in Adults & Kids .pad-left[ - 1.9% of adults seropositive - 0.9% of children seropositive - pre-school age (1-5 years): 0.8% - school age (6-10 years): 1.0% - attending childcare: 0.5% - not attending childcare: 1.0% ] --- ## Comparing Adults & Kids .pad-left[ - "Model 1 yielded an estimated prevalence for children aged 1 to 10 years of 0.6% (95% CI, 0.3-1.0; Table 5). This estimated seroprevalence was 3-fold lower than in parents (1.8% [95% CI, 1.2-2.4])." - "The lower SARS-CoV-2 seroprevalence in young children compared with their corresponding parent is an important observation, because <span style='background-color:lightyellow'>it indicates that children are very unlikely to have boosted the COVID-19 outbreak</span> in southwest Germany during the period of investigation." - "<span style='background-color:lightyellow'>This contrasts with other respiratory tract infections, such as influenza or pneumococci, in which children can play a prominent role for the dissemination of the disease</span>. The higher frequency and proximity of social contacts of children are considered to be a major driver of virus transmission."<sup>1,2,3,4</sup> ] .footnote[ 1 - Viboud C et al <u>ICS</u> 2004; 2 - Pedro PA et al <u>Vaccine</u> 2005; 3 - Caucehmez S <u>Nature</u> 2008;<br>4 - Mossong J et al <u>PLoS Medicine</u> 2008 ] --- layout: false class: full-screen hide-count <div class="grid-3-1"> <div class="extension-tile gridset-grey"><div>Tönshoff B<br>et al<br>Findings</div></div> <div class="extension-tile gridset-b animated flipInY"><div>Parallel<br>Studies</div></div> <div class="extension-tile gridset-grey"><div>What Causes<br>SARS-CoV-2<br>Spread?</div></div> </div> --- ## Other Pediatric Seroprevalence Studies .pad-left[ - Seroprevalence results vary according to location & phase of global pandemic<sup>1-3</sup> - United States: - St. Louis<sup>4</sup>: 1.7% kids, 3.1% adults - Mississippi<sup>5</sup>: 14.5% <6mos, 14.3% 6-11mos, 9.9% 1-8yrs, 9.0% 9-17yrs - "cumulative infection rates... among persons aged <18 years were similar to those among persons aged 18–49 years, the age group with the highest seroprevalence during the period" ] .footnote[ 1 - Ismail et al _Lancet Infect Dis_ 2021; 2 - Wang et al _Pediatr Investig_ 2020; 3 - Roarty et al _Lancet Infect Dis_ 2020;<br>4 - Smith et al _mSphere_ 2021; 5 - Hobbs et al _MMWR_ 2021 ] --- layout: false class: full-screen hide-count <div class="grid-3-1"> <div class="extension-tile gridset-grey animated"><div>Tönshoff B<br>et al<br>Findings</div></div> <div class="extension-tile gridset-grey animated"><div>Parallel<br>Studies</div></div> <div class="extension-tile gridset-c animated flipInX"><div>What Causes<br>SARS-CoV-2<br>Spread?</div></div> </div> --- ## Concerns About COVID-19 in Kids .pad-left[ - **Pediatric COVID-19 Disease**: - mercifully, severe primary disease and MIS-C are rare - **Safety of Kid-Proximate Workers**: - the in-person school debate<sup>1</sup> - **Contribution of Pediatric Population to Larger COVID-19 Spread**: - impact of systematic under-reporting of pediatric cases? - how strong is your prior that kids spread viruses? ] .footnote[ 1 - Gold JAW et al _MMWR_ 2021 ] --- ## Models for SARS-CoV-2 Spread .pull-left[ - Implicit model for SARS-CoV-2 `\(R\)`: `$$R_{t} \sim {age}_{t}$$` - Does age <u>mediate</u> SARS-CoV-2 spread? - What is the evidence for differences in individual viral load/shedding? - What is the evidence for an impact of age distribution on transmission events? ] .pull-right[
] --- ## Models for SARS-CoV-2 Spread .pull-left[ - Implicit model for SARS-CoV-2 `\(R\)`: `$$R_{t} \sim {age}_{t}$$` - Does age <u>mediate</u> SARS-CoV-2 spread? - What is the evidence for differences in individual viral load/shedding? - What is the evidence for an impact of age distribution on transmission events? ] .pull-right[
] --- ## Models for SARS-CoV-2 Spread .pull-left[ - Implicit model for SARS-CoV-2 `\(R\)`: `$$R_{t} \sim {age}_{t}$$` - Does age <u>mediate</u> SARS-CoV-2 spread? - What is the evidence for differences in individual viral load/shedding? - What is the evidence for an impact of age distribution on transmission events? ] .pull-right[
] --- ##Models for SARS-CoV-2 Spread .pull-left[ - Alternative model for SARS-CoV-2 `\(R\)`: `$$R_{t} \sim {community\ mobility}_{t}$$` `$${community\ mobility}_{t} \sim {school\ policy}_{t}$$` - Differences observed across age groups reflect differences in community mobility, which are impacted by school policy. ] .pull-right[
] --- ##Models for SARS-CoV-2 Spread .pull-left[ - Alternative model for SARS-CoV-2 `\(R\)`: `$$R_{t} \sim {community\ mobility}_{t}$$` `$${community\ mobility}_{t} \sim {school\ policy}_{t}$$` - Differences observed across age groups reflect differences in community mobility, which are impacted by school policy. ] .pull-right[
] --- ##Models for SARS-CoV-2 Spread .pull-left[ - Alternative model for SARS-CoV-2 `\(R\)`: `$$R_{t} \sim {community\ mobility}_{t}$$` `$${community\ mobility}_{t} \sim {school\ policy}_{t}$$` - Differences observed across age groups reflect differences in community mobility, which are impacted by school policy. ] .pull-right[
] --- ## Measurement Challenge ⇾ Model Parameters .pad-left[ - Monod M et al. Age groups that sustain resurging COVID-19 epidemics in the United States. _Science_. 2021 Feb 2; PMID: 33531384 - "at least 65 of 100 COVID-19 infections originate from individuals aged 20-49" - Foursquare social mobility data: not available for 0-19 years - COVID-19 infections underreported in pediatric population - "school opening is associated with an estimated 25.7% [14.5%-40.5%] increase of COVID-19 infections and a 5.9% [3.4%-9.3%] increase in COVID-19 attributable deaths" - **significantly greater if pediatric case numbers substantially underreported** ] --- exclude: true layout: false class: full-screen hide-count <div class="grid-3-1"> <div class="extension-tile gridset-a animated flipInX"><div>Tönshoff B<br>et al<br>Findings</div></div> <div class="extension-tile gridset-b animated flipInY"><div>Parallel<br>Studies</div></div> <div class="extension-tile gridset-c animated flipInX"><div>What Causes<br>SARS-CoV-2<br>Spread?</div></div> </div> --- class: middle center hide-count background-image: url(img/sarscov2_cdc_blur.png) background-size: cover # <span style='color:white'>Questions?</span> <!-- .callout-url-bottom[ .fade-in[<span style="font-size:0.5em; color:white">slides ↓ </span>] <span style="link-color:white">[bjklab.org](http://www.bjklab.org)</span> ] -->