Process mining to identify average time patterns of changes in FAMACHA© scores in ewes and guide monitoring intervals

Frequent blanket treatment with anthelmintics (AH) leads to selection of resistant parasites. Targeted Selective Treatment (TST) methodologies were developed to support more discriminating decisions on AH treatment of individual animals. Criteria like FAMACHA© scores (F-scores, a measure of anemia) detect changes in animal health to trigger treatment, but require effort and so the appropriate frequency of monitoring must be considered. Process mining can be used to define optimal intervals for TST application by analyzing flows between states over time. The aim of this study was therefore to evaluate the average time taken for changes in the F-scores of adult ewes using process mining. The study was performed by analyzing a database containing F-score evaluations from 2006 to 2022, using UpFlux© software. From 2017-2022 ewes were categorized based on their physiological state. Ewes in late gestation and lactation were classified as likely to be more susceptible to parasites, while non-pregnant and early pregnant ewes were considered as less susceptible. When considering data for the entire flock from 2017 to 2022, an average of 97 ewes were evaluated weekly, and they spent a mean of 78.5 days at F-score 1 (F1) and 25.0 days at F2. Based on F-score fluctuations, ewes progressed from F1 to F2 in an average of 59.1 days and from F2 to F3 in an average of 26.3 days. Ewes in the more susceptible physiological states maintained F1 and F2 scores for an average of 34.1 and 20.0 days, respectively. Their transition from score F2 (non-anemic) to F3 (anemic) occurred in an average of 19.7 days. Ewes in less susceptible physiological states maintained F1 for 68.2 days and F2 for 24.5 days, while F2 changed to F3 after an average of 29.6 days. Based on the results, evaluations for TST should be conducted more frequently (every 15 days) in ewes during late pregnancy and lactation, while a 30-day interval is sufficient for other physiological states. This method could be used to evaluate optimal intervals for F-scoring in other settings, to guide efficient animal health monitoring and intervention strategies.