Decision analysis model for paratuberculosis control in commercial dairy herds

Abstract

A previous economic test-and-cull decision analysis model has been strengthened and updated with current epidemiologic information. Created using Excel and PrecisionTree software, the model incorporates costs and benefits of herd management changes, diagnostic testing, and different management actions based on test results to control paratuberculosis in commercial dairy herds. This novel "JD-Tree" model includes a herd management decision node (four options), a test/no test decision node (two options), a diagnostic test choice decision node (five options), test result chance nodes (four levels of possible results), and test action decision nodes (three options; cull, manage, no action). The model culminates in a chance node for true infection status. Outcomes are measured as a net cost-benefit value to the producer. The model demonstrates that improving herd management practices to control infection spread (hygiene) is often more cost-effective than testing; not all herds should test as part of a paratuberculosis control program. For many herds, low-cost tests are more useful than more sensitive, higher cost tests. The model also indicates that test-positive cows in early stages of infection may be retained in the herd to generate farm income, provided they are managed properly to limit infection transmission. JD-Tree is a useful instructional tool, helping veterinarians understand the complex interactions affecting the economics of paratuberculosis control and to define the accuracy and cost specifications of better diagnostic tests.