*2.1. Dataset*

With the permission of the OVs and the director of the control post, surveillance reports from 2010 to 2015 filed by the director of the control post (CP) IT CE 07/PS Bitritto (Bari, Italy) in compliance with Council Regulation (EC) No. 1255/1997 (Article 5(h)), and double-checked by OVs from the Local Health Authority (Bari, Italy) were used for this manuscript.

The CP (authorization CE 07/PS; 41◦227 N, 16◦50 9 E) measured approximately 8000 square metres during the data collection period and could contain up to 1200 bovines, 1850 sheep/goats, 500 pigs and 150 horses. It had a lorry wash, three hay barns, five animal houses and many different pens, equipped with watering and feeding points, bedding and resting facilities. The animal houses had adequate natural and artificial light to ensure proper inspection of the animals by the OVs.

At the control post, three OVs working for the Local Health Authority (Bari, Italy), including one of the authors (DT) checked the animals during loading, and immediately before they left the CP, to assess their fitness for transport following the criteria set out in Annex 1 of Council Regulation (EC) No. 1/2005. Depending on the day of arrival, the animals were checked by one of the three OVs. However, those three OVs had each received official EU training to assess fitness for transport and used the same check list to compile the surveillance reports. In compliance with the (EC) No. 1/2005, the OVs judged animals as unfit to continue the journey when they were either seriously injured or presented clinical signs of a pathological process, including severe lameness (animals unable to move independently without pain or to walk unassisted), prolapse, severe open wound, and respiratory and gastro-enteric disease (animals with massive nasal discharge, severe dyspnoea and pleurodynia, animals with severe diarrhoea or acute abdominal pain, respectively). Such animals were either humanely euthanized, slaughtered in the local slaughterhouse or treated depending on the severity of the case after the assessment performed by the OVs.

The following parameters were recorded in the surveillance reports: date and time of arrival, species, number of animals transported per truck (NATT), country of provenance, TRACES code, truck registration number, transport company, number of dead on arrival (DOA), of those which died at the control post (DCP), of animals judged unfit to continue travel (UFT), day and time of departure, and country of final destination. Since these reports were official documents, there were no missing data.

Based on the surveillance reports, the dates were expressed as month and season. Using the category which was reported in each TRACES, the category of transported animals was added to the dataset. The categories were chosen in compliance with Council Regulation (EC) No. 1/2005, for bovines and sheep/goats the following categories were used: small calves (100 kg), medium-sized calves (200 kg), heavy calves (325 kg), heavy cattle (550 kg), very heavy cattle (>700 kg), lambs, and sheep/goats; whereas for pigs, the following categories were applied: light fattening (110–120 kg), heavy fattening (130–150 kg), and breeding (>200 kg).

The space allowance per animal was also added to the dataset. It was calculated by dividing the available floor space in each truck by the number of animals transported in the truck and expressed as m2/animal. It was not possible to calculate the space allowance in kg/m<sup>2</sup> because the surveillance reports did not report the total weight of the truck load.

The transport companies were aggregated by country of origin (i.e., nationality) and the total number of dead animals was calculated as the sum of DOA and DCP.
