Science - USA (2021-07-09)

SCIENCE sciencemag.org

GRAPHIC: K. FRANKLIN/

SCIENCE

; (ILLUSTRATIONS OF PEOPLE) PROPUBLICA’S WEEPEOPLE FONT

illness ( 15 ). Therefore, structural conditions

that affect an individual’s network and expo-

sure risk are likely far more predictive than

individual choices in determining whether

the infection will be a terminal event or lead

to multiple downstream infections. Thus,

comprehensively addressing the needs of

a few with disproportionate risks can avert

more downstream infections than eliminat-

ing a small risk among many.

How can public health strategies address

individual heterogeneity and differential in-

fection risk? Early in the pandemic, there was

an assumption of relative homogeneity in the

risks of infection and the potential impact of

interventions across the population. This was

included in modeling to inform public health

approaches. Compartmental models, which

divide populations into distinct sections and

assume that individuals in these groups have

the same characteristics, are mostly used to

model COVID-19 cases and the impact of in-

terventions. However, they infrequently in-

tegrate the effects of differential population

mixing, socioeconomic factors, and networks

across compartment effects. It is now clear

that individual heterogeneity has large-scale

effects on disparities seen in the risk of in-

fection and disease burden, which is con-

firmed in network-based disease modeling ( 1 ,

6 , 7 , 11 ). Public health policies implemented

based on the assumption of equal risk of ac-

quisition and transmission across all socio-

economic groups, ages, and occupations left

certain communities exposed to a higher risk

of infection, resulting in differential burdens

of disease ( 1 – 3 , 7 ). Leveraging network het-

erogeneity in infectious disease models may

better demonstrate these differential risks

observed in real-life epidemiological analyses

and the benefits of prioritizing intensive and

targeted interventions to those with differen-

tial risks, given the potential for larger num-

bers of averted downstream infections.

The intersection between direct and indi-

rect onward transmission risks reinforces the

need for effective and pragmatic strategies to

break chains of transmission, especially in

people at high risk of infection. Policy inter-

ventions should consider the overall number

of contacts that a person has and, subse-

quently, downstream infections averted based

on differential impacts in different communi-

ties ( 13 ). For example, people living in mul-

tigenerational households, serving in high-

exposure occupations, and residing in densely

populated communities could be prioritized

for temporary housing support, assurance

of employee benefits such as paid leave, and

vaccine outreach services. Considering this

differential impact, targeted interventions

through network-adaptive resource-based

interventions could be leveraged according

to individual and network-level needs. Such

an adaptive approach could inform model-

ing and prioritize specific resource-based

intervention strategies, including testing

aligned with lived realities, housing support

if insufficient space to isolate, and paid leave

from work to support quarantine and iso-

lation, combined with outreach testing and

infection prevention and control support in

workplaces. In addition, network-adaptive

vaccination strategies prioritize those with

large networks based on contact heteroge-

neity. For example, looking at vaccination

rates in England by deprivation, vaccination

coverage is clearly lower in more deprived ar-

eas, where the risk of infection and disease

burden is higher. Although this may be due

to multiple reasons, including lack of access

to care and inability to take time off work,

there is a need for areas of high and enduring

transmission risk to accelerate vaccination

to match the increased risks of infection and

onward transmission.

The focus of COVID-19 response strategies

has often been on behavior change as a pri-

mary means of decreasing contact networks

and thus transmission chains. However,

contact patterns are driven, in large part, by

socioeconomic inequities and structural rac-

ism and are nonmodifiable at the individual

level in the absence of specific support. Thus,

nonadaptive public health interventions

fail to address individual heterogeneities

and have left socioeconomically marginal-

ized communities at risk of infection, death,

and economic hardship. There is a risk that

lower vaccine uptake among these commu-

nities could perpetuate existing inequalities.

Therefore, it is vital that community-led vac-

cine delivery strategies are strengthened.

The disparities that have defined

COVID-19 epidemiology could have been

readily predictable given historical data on

pandemics. The next respiratory pandemic

will also be defined by similar disparities.

Using network-driven strategies to inform

rapidly emergent epidemic responses repre-

sents an evidence-based and equitable path

forward where the aim is to invest more

to prevent infections in a  person with dis-

proportionate risks because disease burden

and downstream infection risks vary sub-

stantially ( 6 , 7 , 13 ). j

REFERENCES AND NOTES

1. T. D. Hollingsworth, J. Public Health Policy 30 , 328
   (2009).


2. M. Cevik, J. L. Marcus, C. Buckee, T. C. Smith, Clin. Infect.
   Dis. 10.1093/cid/ciaa1442 (2020).


3. M. E. Sundaram et al., medRxiv
   10.1101/2020.11.09.20223792 (2020).


4. Public Health England (PHE) Transmission Group,
   “Factors contributing to risk of SARS-CoV2 transmis-
   sion in various settings” (PHE, 2020); https://bit.
   ly/3xIbuMM.


5. A. Goyal, D. B. Reeves, E. F. Cardozo-Ojeda, J. T. Schiffer,
   B. T. Mayer, eLife 10 , e63537 (2021).


6. S. Mishra, S. D. Baral, Lancet Infect. Dis. 20 , 155 (2020).


7. S. Mishra, J. C. Kwong, A. K. Chan, S. D. Baral, CMAJ 192 ,
   E684 (2020).


8. J. E. Lemieux et al., Science 371 , eabe3261 (2021).


9. Y.-H. Chen et al., medRxiv 10.1101/2021.01.21.21250266
   (2021).


10. N. I. Lone et al., Lancet Reg. Health Eur. 1 , 100005
    (2021).


11. K. Sun et al., Science 371 , eabe2424 (2021).


12. X. Qiu et al., Clin. Microbiol. Infect. 27 , 511 (2021).


13. B. G. Link, J. Phelan, J. Health Soc. Behav. 1995 , 80
    (1995).


14. COVID-19 Advisory for Ontario, Update on COVID-19
    projections: Ontario COVID-19 science advisory table
    and modelling consensus table (2020); https://bit.
    ly/3wRzHjY.


15. M. Murti et al., medRxiv 10.1101/2020.11.25.20239038
    (2020).
    10.1126/science.abg0842

Case A Case B

Network

social distancing

Contact

frequency

Cumulative

contacts

Healthy Infected Social distancing

Downstream infection risks vary according to network patterns

Case A depicts a person with a small network, who can work from home and self-isolate if needed. Case B

represents a person who works in a public-facing job or in an unsafe workplace and lives in a multigenerational

or large household. Overall risk of exposure and onward transmission risk differ substantially between these

two individuals, representing a disproportionately high transmission chain in case B. Intervention strategies

should focus on breaking chains of downstream transmission.

9 JULY 2021 • VOL 373 ISSUE 6551 163

0709Perspectives.indd 163 7/1/21 6:34 PM