Inter-observer repeatability of indicators of consciousness in broiler chicken after waterbath stunning
Abstract
One of the main challenges in monitoring the state of consciousness in broiler chicken after waterbath stunning is the selection of the animal-based indicators (ABI) ensuring consistency of controls. To be relevant, ABI should meet three requirements, validity, feasibility and repeatability. The validity and feasibility of ABIs have been assessed by EFSA (2013). However, it is pending to assess the repeatability so that can be proposed a refined and validated list. Thus, the main goal of the study was to assess the inter-observer repeatability of the most valid and feasible ABIs for the state of consciousness after water bath stunning in broilers both before bleeding (tonic seizure, breathing, spontaneous blinking and vocalisation) and during bleeding (wing flapping, breathing, spontaneous swallowing and head shaking) and the correlation among them.
This study compared the assessment of 3 observers in 5,241 broilers from 19 batches of 6 different slaughterhouses in two EU countries. Data were analysed at individual broiler level and the combination of crude percentage of agreement (PoA) and Fleiss’ kappa (k) and its interpretation according to Fleiss (2003) was used to assess the inter-observer repeatability of the outcomes of some ABIs for the state of consciousness.
Before bleeding, the most repeatable ABI was vocalisation (PoA = 100%) followed by spontaneous blinking (PoA = 99.8%; k = poor), breathing (PoA = 98.9%; k = fair to good) and tonic seizure (PoA = 91.7%; k = fair to good).
However, both vocalisation and spontaneous blinking were artificially highly repeatable as hardly ever were observed. On the other hand, absence of tonic seizure was the less repeatable and was not correlated to other ABI before bleeding probably because tonic seizure occurred in some birds while the bird was still in the water bath. Therefore, it seems difficult to rely on the absence of tonic seizure to assess consciousness. Thus, we recommend focusing on presence of breathing as indicator of consciousness. However, presence of spontaneous blinking and vocalisation, although hardly ever observed, should not be neglected as indicators of consciousness and ineffective stunning.
During bleeding, the most repeatable ABI was spontaneous swallowing (PoA = 98.84; k = poor), followed by wing flapping (PoA = 98.2%; k = fair to good), head shaking (PoA = 96.4%; k = fair to good), and breathing (PoA = 88.2%; k = fair to good). However, spontaneous swallowing is artificially repeatable as was the least observed indicator. Therefore, we recommend focus on presence of breathing, head shaking and wing flapping assessment although less repeatable. Sometimes birds showed simultaneously more than one outcome of consciousness being breathing and head shaking and breathing and wing flapping the most observed combinations.
This work will serve at proposing a refined list of ABIs so that they can be used to assess the consciousness of broiler chickens in commercial slaughterhouses.