Comparison of body temperature in donkeys using rectal digital, infrared, and mercury-in-glass thermometers during the hot-dry season in a tropical savannah.

The study aimed at comparing variations in body temperature values recorded using rectal digital, infrared, and mercury-in-glass thermometers in donkeys during the hot-dry season, prevailing under tropical savannah conditions. Thirty donkeys that served as subjects were divided into three groups of adults, yearlings, and foals. Values of the body temperature of each donkey were recorded bihourly, starting from 06:00 h till 18:00 h, by digital (5-cm depth of insertion), mercury-in-glass (3 cm depth), and infrared thermometers. The values obtained by each type of the thermometer were compared with those recorded using a 15-cm digital probe (Model HI935007, Hanna Instruments, range -50.0 to 150.0°C; accuracy ± 0.2°C) which served as the gold standard. Dry-bulb temperature (34.00 ± 0.50°C), temperature-humidity index (79.65 ± 0.15), and wet-bulb globe temperature (28.00 ± 0.50) index peaked at 14:00 h. The mean body temperatures for rectal probe, digital, mercury-in-glass, and infrared thermometers were 38.35 ± 0.11°C, 37.24 ± 0.04°C, 36.76 ± 0.06°C, and 36.92 ± 0.07°C, respectively. In comparison to the rectal probe, the mean bias for digital (-1.11 ± 0.05°C), mercury-in-glass (-1.59 ± 0.07°C), and infrared thermometers (-1.38 ± 0.07°C) was large. The Passing-Bablok regression plot demonstrated significant deviation from linearity (p < 0.01) when digital, infrared, and mercury-in-glass thermometers were compared to the rectal probe. The area under the curve (AUC) for digital (AUC: 0.7005 ± 0.01 [95%: 0.6853 - 0.7310], infrared (AUC: 0.6711 ± 0.01 [95%: 0.6322 - 0.7100], and mercury-in-glass (AUC: 0.6321 ± 0.01 [95%: 0.6001 - 0.7873] thermometers showed poor accuracy with low sensitivity. In conclusion, the use of digital, mercury-in-glass, and infrared thermometers in recording body temperature in donkeys during the hot-dry season underestimated the values. Their use in measuring body temperature may result in wrong diagnosis, and compromise the control of hyperthermia and diseases associated with thermoregulatory impairments in donkeys.