Use of 311 data to estimate pathogen risk and host distribution: Is this the right way to go?

Background/Question/Methods

Urbanization is placing increasingly greater proportions of the global population at risk of experiencing a major trauma capable of transforming built and natural environments. The potential for disease outbreaks among at-risk populations often motivates strategies for adaptive management based on perceived balances between risks and benefits to human well-being. Yet understanding of trauma and post-trauma recovery is strikingly disconnected from understanding of risk perception including identification of potential conditions that originate or perpetuate uncertainty. Greater understanding of sources and drivers of perceived risk is critical because perceptions of risk can be the primary– and sometimes only– factors determining how post-trauma interventions shape the recovery of socioecological communities. Moreover, relating perceptions to physical rick factors can help prevent the “pathology of infectious disease control”, which involves a narrowing focus of institutional responses to a change in socioecological conditions that may escalate rather than reduce disease risk.

Here we examine whether the perception of rodents parallels the distribution and abundance of Rattus sp. in post-Katrina New Orleans to determine the extent to which risk perceptions are decoupled from physical risk factors. We did so by comparing the geographical distribution and number of 311-calls of rodent complaints to exhaustive removal trapping-based estimates of population size for R. rattus and R. norvegicus over 80 blocks across the city. We also examined whether 311 calls and population sizes of Rattus sp. covary with land cover, human population density, and the prevalence of abandoned properties. 

Results/Conclusions

In 2014, the city of New Orleans received 910 rodent complaints; the majority of calls (63%) occurred between June and October, which corresponds to peak rodent breeding season. Nearly half of the calls (45%) originated from areas of the city with an intermediate burden of abandoned properties. Though Rattus sp. population sizes were also related to abandonment, the distribution and number of rodent complaints was not consistently related with population size estimates. Large Rattus sp. populations were often detected at complaint hot spots, but low call numbers did not correspond to smaller population sizes, which may be a consequence of low human densities in areas with high levels of abandonment. These findings suggest that measures of risk perception, like 311 call complaints, may become decoupled from physical risk factors due to mitigating conditions, and therefore may not serve as an accurate proxy for estimating the distribution or prevalence of zoonotic pathogens and host population size.