A comprehensive analysis of inspection and audit reports in animal-derived food production by local competent authorities: A study from Emilia-Romagna, Italy (2020–2024)

Food safety is essential to public health and is ensured through official controls that verify compliance across the food supply chain. This study examines the frequency and nature of non-compliances identified through inspections and audits in Emilia-Romagna, Italy, from 2020 to 2024. A survey explored challenges faced by veterinarians in official controls. Over 540,000 inspections and 40,000 audits were performed in approximate to 3000 production activities handling products of animal origin. Inspection non-compliance rates remained low, from 1.7 % in 2020 to around 2.3-2.4 % in later years. Few inspections led to enforcement actions: 89 legal referrals, 151 administrative penalties, and 55 product seizures. Audits showed higher non-compliance rates (4.6 %), particularly in poultry slaughterhouses (8.5 %) and meat processing facilities (7.2 %), though rarely resulting in sanctions. The study underscores methodological and functional differences between inspections and audits, the former being more effective in identifying operational issues, and the latter in detecting systemic deficiencies. While audits revealed more non-compliances, this quantitative difference lacks consistent support from broader literature and deeper analysis, suggesting both tools have complementary strengths shaped by context and methodology. The Complaint Index highlighted imbalances in inspection frequency, pointing to sectors where resources could be better allocated. However, limitations such as fragmented data systems hinder comprehensive risk-based planning. Official veterinarians' survey responses indicated a need for improved training, better checklists, and time optimization. The findings advocate for a transition from broad risk categories to targeted, risk-based inspections, supported by digital tools to enhance the efficiency and effectiveness of official controls.