Assessment of farmers' compliance in implementing recommended cow comfort changes and their effects on lying time, stall and cow cleanliness within smallholder dairy farms in Kenya.

Our study aimed to evaluate farmers' compliance in implementing recommendations of farm-specific cow comfort changes, and the effects of these changes on lying time, stall cleanliness and cow cleanliness using a randomized controlled trial carried out on 100 smallholder dairy farms in Kenya, with 62 and 11 farms remaining in the intervention and control groups, respectively. On the first farm visit, data loggers were attached on lactating cows to determine lying time and questionnaires utilized to collect baseline data. Three days later, stall design and management recommendations were given to the intervention group of farmers orally and in written form. After an average of 39 ± 7 days, data loggers were re-attached, compliance was assessed, and a post-intervention questionnaire was administered to the intervention group on the third visit. Three days subsequent to the first and third visits, data loggers were removed from all cows. Data were analysed in Stata 14.2® using proportion tests and Kruskal-Wallis rank tests to compare cleanliness scores and lying time, respectively. Interaction effects between treatment groups and visits were assessed using multivariable mixed linear and logistic regression models. While 46 of the 62 intervention farmers (74%) made at least one recommended change to cow comfort, 63% of the 324 overall recommendations were implemented. The odds of a recommendation being implemented were significantly higher when:1) major recommendations were given relative to minor recommendations (OR = 6.28); 2) recommendations were related to floor characteristics (floor softness and flatness) in comparison to recommendations related to stall design (OR = 3.14). The odds of compliance were lower on: 1) farms where the farm-hands received the recommendations compared to farms that had the female principal farmer receive the recommendations (OR = 0.01); 2) farms that had recommended changes related to roof, alley and sharps fixes relative to stall design fixes (OR = 0.13). Post-intervention, stall, udder and upper hind-leg cleanliness scores improved significantly (p < 0.0001, p = 0.021 and p = 0.017, respectively) in the intervention farms but not in the control farms. There was no significant difference in lying times between intervention and control farms, with 0.6 and 0.2 h/day increases being recorded in the intervention and control groups, from the 10.9 and 10.4 h/day at baseline, respectively. Giving farm-specific cow comfort recommendations to smallholder dairy farmers in Kenya, and providing them with a participatory role in the formulation and implementation of improvement recommendations ensured good acceptance and a high degree of implementation, and led to a subsequent improvement in cow comfort and cleanliness.