Disease Spread Models to Estimate Highly Uncertain Emerging Diseases Losses for Animal Agriculture Insurance Policies: An Application to the U.S. Farm‐Raised Catfish Industry

Abstract

Emerging diseases (ED) can have devastating effects on agriculture. Consequently, agricultural insurance for ED can develop if basic insurability criteria are met, including the capability to estimate the severity of ED outbreaks with associated uncertainty. The U.S. farm-raised channel catfish (Ictalurus punctatus) industry was used to evaluate the feasibility of using a disease spread simulation modeling framework to estimate the potential losses from new ED for agricultural insurance purposes. Two stochastic models were used to simulate the spread of ED between and within channel catfish ponds in Mississippi (MS) under high, medium, and low disease impact scenarios. The mean (95% prediction interval (PI)) proportion of ponds infected within disease-impacted farms was 7.6% (3.8%, 22.8%), 24.5% (3.8%, 72.0%), and 45.6% (4.0%, 92.3%), and the mean (95% PI) proportion of fish mortalities in ponds affected by the disease was 9.8% (1.4%, 26.7%), 49.2% (4.7%, 60.7%), and 88.3% (85.9%, 90.5%) for the low, medium, and high impact scenarios, respectively. The farm-level mortality losses from an ED were up to 40.3% of the total farm inventory and can be used for insurance premium rate development. Disease spread modeling provides a systematic way to organize the current knowledge on the ED perils and, ultimately, use this information to help develop actuarially sound agricultural insurance policies and premiums. However, the estimates obtained will include a large amount of uncertainty driven by the stochastic nature of disease outbreaks, by the uncertainty in the frequency of future ED occurrences, and by the often sparse data available from past outbreaks.

Forensic Examination of a Boar Died During Transport Aimed at Reimbursement of Insured Animal

“…Conventional empirical methods are of limited use to estimate the impact of new risk data, are usually scarce and/or may not be relevant to the new disease. Hence, the feasibility of alternative methods to quantify the impact of risk should be further explored (Coble et al, 2006;Zagmutt et al, 2013). Th e price of risk (insurance) is a premium, which exists in all types of insurance.…”

Modeling Finite‐Time Failure Probabilities in Risk Analysis Applications

“…A risk analytic approach to modeling such losses at a single‐farm level was recently proposed in Ref. for the purpose of assessing the possibility of launching agricultural insurance coverage to protect catfish producers. The authors apply stochastic disease spread modeling (DSM), implemented via Monte Carlo (MC) simulation, in order to estimate the total loss cost, (see Equation in Ref.…”

“…The authors apply stochastic disease spread modeling (DSM), implemented via Monte Carlo (MC) simulation, in order to estimate the total loss cost, (see Equation in Ref. ), due to fish mortality resulting from the spread of an ED in fish ponds belonging to a farm.…”

“…Following Ref. , it is again assumed that disease spread between farms can be triggered by direct and indirect contacts, modeled by a Poisson process. The numerical results in Ref.…”

“…The numerical results in Ref. reveal that the total loss caused by fish mortalities for a single farm, , is a random variable that follows a certain distribution that is estimated using MC simulation. Thus, in a multi‐farm system, where an ED spreads from one farm to another, the loss from each individual farm, , can be modeled by a random variable , .…”