Adoption of farm biosecurity practices among smallholder poultry farmers in Kenya - An application of latent class analysis with a multinomial logistic regression

Abstract

Sub-Saharan Africa has a growing demand for poultry, but productivity in the sector has not increased to meet this demand. One major constraints in the sector is diseases. Many farmers currently use clinical control measures that involve treating birds with antibiotics upon detecting an infection. However, this approach has presented the misuse of antibiotics, leading to antimicrobial resistance, which could have catastrophic effects going by different projections. We evaluate the uptake of preventive approaches to disease management, otherwise known as biosecurity measures and the effect of the adopted practices on animal health outcome among poultry farmers in Nyanza region of Kenya. The study applies latent class analysis, which is a model-based clustering approach to categorize poultry farmers into low, moderate, and high biosecurity adoption classes. We find low adoption of biosecurity measures across all classes of smallholder poultry farmers in Nyanza. However, correlation analysis show that increased uptake of biosecurity measures is associated with positive poultry health outcomes. This is as demonstrated by lower mortality rates among farmers characterized by higher adoption of biosecurity measures. Lastly, we implement a multinomial logistic regression to assess determinants of class membership and our analysis shows that information access is the greatest driver of biosecurity adoption. Farmers who had access to information on biosecurity measures were 25 % more likely to belong to the class of farmers adopting more biosecurity practices - high adoption class- and 21 % less likely to be in the moderate adopters class. As such, the study recommends enhanced information dissemination to improve the uptake of biosecurity measures.