Contact Tracing Simulation: Breaking Disease Transmission Networks

simulator intermediate ~10 min
Loading simulation...
~35% of network infected with 8 traces/day, 70% compliance

In a network of 80 people with 5 average contacts, tracing 8 contacts per day with 70% compliance limits the outbreak to about 35% of the network. Without tracing, the same network sees 70%+ infection. Speed and compliance are the two critical levers.

Formula

Tracing effectiveness: E = trace_speed × compliance / (new_cases × avg_degree)
R_effective with tracing: R_t ≈ R₀ × (1 - E)
Containment condition: trace_speed × compliance > R₀ × avg_degree

Racing the Pathogen

Contact tracing is fundamentally a race between public health workers and the pathogen. When an infected person is identified, tracers must find, notify, and isolate their contacts before those contacts become infectious and pass the disease onward. The critical variable is the ratio of tracing speed to the disease's serial interval. If tracers can identify and isolate 80% of contacts within one serial interval, containment is achievable even for moderately transmissible diseases.

Network Topology Matters

Diseases do not spread through homogeneous populations — they follow the architecture of human social networks. Most people have a moderate number of contacts, but a few hubs connect to many others. These high-degree nodes are both the greatest risk (superspreaders) and the greatest opportunity (tracing them protects the most people). The simulation builds a random network with tunable average degree, showing how network connectivity shapes both transmission and tracing effectiveness.

The Compliance Bottleneck

Identifying contacts is only half the battle. Traced individuals must actually comply with isolation instructions to prevent onward transmission. Studies during COVID-19 found that isolation compliance ranged from 40% to 80% depending on financial support, housing conditions, and trust in public health authorities. Even perfect tracing fails if compliance is too low — the simulation lets you directly observe this effect by adjusting the compliance slider.

Visualizing Transmission Chains

This simulation renders the social network as an animated graph. Infected nodes glow red, traced nodes turn cyan, and recovered nodes fade to gray. Watch as infection waves ripple outward from patient zero, then see tracing efforts chase the wavefront. The chains-broken counter shows how many potential transmission paths were severed by timely isolation, quantifying the invisible work of contact tracers.

FAQ

How does contact tracing reduce disease spread?

Contact tracing identifies people exposed to an infected individual, then isolates them before they can transmit the pathogen onward. By removing potential transmission links from the network faster than the disease can traverse them, tracing breaks chains of infection. Effectiveness depends on speed (finding contacts before they become infectious), completeness (identifying most contacts), and compliance (contacts actually isolating).

What determines whether contact tracing can contain an outbreak?

Three factors: the disease's serial interval (time between successive cases), the fraction of contacts that can be traced, and the speed of tracing. If tracing can identify and isolate contacts before they reach peak infectiousness, containment is possible. For COVID-19, pre-symptomatic transmission made this extremely difficult. For SARS (2003), where transmission occurred mainly after symptom onset, tracing was decisive.

Why is network structure important for contact tracing?

In heterogeneous networks where some people have far more contacts than average, superspreaders can overwhelm tracing capacity. A single superspreading event might generate 30+ contacts that all need tracing simultaneously. Network structure also determines whether traced individuals are central (breaking many potential paths) or peripheral (having little impact on overall transmission).

How does digital contact tracing compare to manual tracing?

Digital tracing (smartphone apps) can identify contacts in hours rather than days and captures contacts the index case may not remember (e.g., strangers on public transport). However, it requires high adoption rates (>60%), misses contacts without smartphones, and raises privacy concerns. The most effective approach combines digital and manual tracing, using technology for speed and human interviewers for completeness.

Sources

Embed

<iframe src="https://homo-deus.com/lab/epidemiological-modeling/contact-tracing/embed" width="100%" height="400" frameborder="0"></iframe>
View source on GitHub