Network Effects of Risk Behavior Change Following Prophylactic Interventions

Rajmohan Rajaraman, Zhifeng Sun, Ravi Sundaram, Anil VullikantiPLoS 2013 · DOI

Abstract

We formulated a network-based model to understand how risk behavior change in conjunction with failure of prophylactic interventions can lead to unintended outcomes where "less (intervention) is more (effective)." Our model captures the distinction between one- and two-sided risk behavior change. In one-sided situations (e.g. influenza/H1N1) it is sufficient for either individual in an interaction to exhibit risk behavior change whereas in two-sided situations (e.g. AIDS/HIV) it is necessary for both individuals in the interaction to exhibit risk behavior change, for a potential transmission of the disease. A central discovery is that this phenomenon occurs at differing levels of intervention coverage depending upon the "sidedness" of the interaction. We find that for one-sided interactions, sufficiently high vaccination coverage is necessary for mitigating the effects of risk behavior; for two-sided interactions, it is essential to combine prophylactic treatments with programs aimed at reducing risky behavior. Furthermore, again dependent on the "sidedness," targeting highly connected nodes can be strictly worse than uniformly random interventions at the same level of coverage.

Introduction

Public health interventions, such as vaccines and antivirals, are rarely 100% effective. A significant complication in their deployment is "risk behavior change," also known as moral hazard. Individuals who believe they are protected by an intervention may engage in riskier social or sexual behaviors, potentially increasing the overall transmissibility of a disease. While previous research has identified this "perverse" outcome, it often relied on simplistic differential equation models that treat populations as homogeneous masses.

This paper addresses this gap by applying network-based modeling to understand how the specific structure of human interactions and the "sidedness" of behavior change influence disease spread. The goal is to determine when interventions help and when they inadvertently exacerbate an epidemic.

Methodology: Modeling Sidedness in Networks

The authors utilize a discrete-time Susceptible-Infected-Recovered (SIR) model mapped onto various network topologies. In this model, each node exists in one of three states: unvaccinated, successfully vaccinated, or "failed" vaccinee. Failed vaccinees remain susceptible but exhibit an increased "boosted" transmissivity ($p_m$) due to behavior change.

A key innovation of this study is the concept of sidedness:

  • One-sided (e.g., Influenza): Transmission is boosted if either the infector or the infectee has a failed vaccine and acts riskily (e.g., attending a crowded event while symptomatic).
  • Two-sided (e.g., HIV): Transmission is only boosted if both parties engage in risky behavior (e.g., unprotected intercourse).
Illustration of the network model comparing one-sided and two-sided risk behavior change.

The study evaluates this model across five distinct network types to ensure robustness: human contact networks, social communication (Enron email), peer-to-peer (Gnutella), preferential attachment, and Erdős-Rényi random networks.

Key Insights and Results

The "Less is More" Phenomenon

The research identifies a non-monotonic relationship between vaccination coverage and epidemic severity. In many scenarios, the expected size of an outbreak increases as the vaccination rate rises, sometimes doubling the epidemic's severity before eventually declining. This "perversity" occurs at different thresholds based on sidedness:

  • In one-sided models, the epidemic worsens at low levels of vaccination.
  • In two-sided models, the danger peaks at high levels of vaccination.

The Targeted Strategy Trap

In standard epidemiological models, vaccinating "hubs" (the most connected individuals) is considered the most effective strategy. However, this paper reveals that if hubs act riskier after vaccination, they become "super-spreaders" of the boosted transmission rate. In these cases, targeted vaccination can be strictly worse than random vaccination. The epidemic's severity becomes a tug-of-war between the vaccine's success in removing nodes and the vaccine's failure triggering higher transmission rates on remaining edges.

Policy Implications and Conclusion

The findings suggest that public health strategies must be tailored to the nature of the disease and the associated behavior change:

  • For one-sided diseases like the flu, achieving very high coverage is the only reliable way to mitigate the effects of risk behavior.
  • For two-sided diseases like HIV, medical interventions must be strictly coupled with behavioral education programs, as the risk of an outbreak can actually increase as high coverage is approached.

While the model assumes static networks and binary risk levels, the observation of the "less is more" effect across all tested networks suggests it is a fundamental property of network-based disease dynamics. Future work will explore co-evolutionary models where behavior and disease spread interact dynamically.

Summary updated

Abstract

We formulated a network-based model to understand how risk behavior change in conjunction with failure of prophylactic interventions can lead to unintended outcomes where "less (intervention) is more (effective)." Our model captures the distinction between one- and two-sided risk behavior change. In one-sided situations (e.g. influenza/H1N1) it is sufficient for either individual in an interaction to exhibit risk behavior change whereas in two-sided situations (e.g. AIDS/HIV) it is necessary for both individuals in the interaction to exhibit risk behavior change, for a potential transmission of the disease. A central discovery is that this phenomenon occurs at differing levels of intervention coverage depending upon the "sidedness" of the interaction. We find that for one-sided interactions, sufficiently high vaccination coverage is necessary for mitigating the effects of risk behavior; for two-sided interactions, it is essential to combine prophylactic treatments with programs aimed at reducing risky behavior. Furthermore, again dependent on the "sidedness," targeting highly connected nodes can be strictly worse than uniformly random interventions at the same level of coverage.
BibTeX
@inproceedings{rajaraman2013network,
  author = {Rajmohan Rajaraman and Zhifeng Sun and Ravi Sundaram and Anil Vullikanti},
  title = {{Network Effects of Risk Behavior Change Following Prophylactic Interventions}},
  booktitle = {PLoS},
  year = {2013},
  doi = {10.1371/journal.pone.0064653},
}
Figure