Epidemiological and partial budget analysis for treatment of subclinical Staphylococcus aureus intramammary infections considering microbiological and cytological scenarios.

An innovative method was investigated to aid in the elimination of Staphylococcus aureus (S. aureus) intramammary infections (IMI) from dairy herds. A stochastic model explore the economic benefit of three-day or eight-day treatment of subclinical IMI in all S. aureus infected cows or in only those with a somatic cell count (SCC) exceeding 200,000 cells/ml. An epidemiological model was developed to run parallel to the economic model that would predict the S. aureus IMI likely to persist, develop new infections and clinical mastitis. In the economic model a first algorithm was used to consider the low prevalence (LP) scenario and made use of S. aureus prevalence information provided by retrospective analysis of microbiological and cytological results in South Africa (2008-2012). The data used considered Staphylococcus aureus prevalence from [1.495; 1.595] to [6.72; 6.95] for SCC≤200,000 and SCC>200,000 cells/ml respectively. A second algorithm considered the high prevalence (HP) scenario to evaluate a simulated situation with a 5[U1] [R12] to 25% prevalence. Scenarios of low or high transmission ratio (TR) were included in the model according to the hygiene management on the farm. Probabilities and costs were calculated over 255days. The economic models predicted average cost indices for low S. aureus IMI and low TR to vary from -3179 ZAR (South African Rands) when subclinical cases with SCC higher than 200,000 cell/ml were treated for eight days, to -3663 ZAR when all subclinical S. aureus IMI were treated for three days. With a HP and high TR of S. aureus the average cost indices changed from -18,042 ZAR when none to -5433 ZAR per 255days when all S. aureus IMI were treated for eight days. The epidemiological model in this study predicted substantial benefit of treatment mainly in high TR scenarios. New IMI decreased up to77% in the three-day and up to 91% in the eight-day treatment scenarios. In the HP scenarios, persistent IMI were reduced by 94%. The number of clinical cases predicted with no treatment for subclinical infections was higher than the total number of clinical and subclinical cases in scenarios where cows were treated three or eight days. Initial prudent treatment of subclinical IMI resulted in less overall treatments and less new, persistent and clinical cases. Combined results of economic and epidemiological models indicated that the option that cost the least did not always have the best epidemiological outcome. Models may assist in optimising and balancing decisions relating to financial and IMI.