- Email: [email protected]
Field-trial evaluation of an automatic temperature measurement device placed in the reticulo-rumen of pre-weaned male calves
Livestock Science 189 (2016) 78–81
Contents lists available at ScienceDirect
Livestock Science journal homepage: www.elsevier.com/locate/livsci
Short communication
Field-trial evaluation of an automatic temperature measurement device placed in the reticulo-rumen of pre-weaned male calves Birte Tietgen a, Hans.-Joachim. Laue b, Martina Hoedemaker c, Steffi Wiedemann a,n a
University of Kiel, Animal Health, Institute of Animal Breeding and Husbandry, 24098 Kiel, Germany University of Applied Science Kiel, Animal Feeding, 24783 Osterrönfeld, Germany c University of Veterinary Medicine, Clinic for Cattle, 30173 Hannover, Germany b
art ic l e i nf o
a b s t r a c t
Article history: Received 8 July 2015 Received in revised form 8 May 2016 Accepted 11 May 2016
The assessment of the rectal temperature plays an important role in the early detection of bovine respiratory disease (BRD) in young calves, but the measurement is invasive and labor-intensive. The aims of this retrospective field-trial evaluation were to assess the correlation between the automatic measurements of the reticulo-ruminal (ReRu) temperature and the rectal temperature and to evaluate the diagnostic accuracy of a cumulative sum (CUSUM) control chart to detect pre-weaned calves suffering from BRD. In 150 male fattening calves (16.6 73.3 d at arrival) the ReRu-temperature was obtained every 5 min automatically with a wireless device over a period of 8 weeks. Data was averaged over periods of 30 min and 4–6 h (day periods). The 30-min means were further evaluated using CUSUM control charts. All calves were inspected by trained persons at least twice a day and rectal temperatures were measured in calves showing visible signs related to BRD. A ReRu hyperthermia (ReRu temperature Z40 °C) was detected in 139 calves over 30 min and in 99 calves over day periods, respectively. During the evaluation period 30 animals were affected by BRD (rectal temperatures Z 40 °C and one additional clinical sign). The correlations between rectal temperatures and both the corresponding 30 min as well as the day period ReRu temperatures was r ¼0.75. The sensitivity (Se) and specificity (Sp) of ReRu hyperthermia for the 30 min means to detect BRD were 77% and 97%, respectively, whereas means of the day periods had a Se of 61% and a Sp of 97%. The CUSUM test revealed a Se of 71% and Sp of 98% to detect BRD. On average, by the CUSUM method calves with clinical signs of BRD were identified 3.5 d earlier. In conclusion, the automated ReRu temperature measurement possesses potential for the early detection of febrile responses in very young calves warranting further investigations. & 2016 Elsevier B.V. All rights reserved.
Keywords: Bolus Bovine respiratory disease Calves Fever Reticulo-rumen temperature
1. Introduction Bovine respiratory disease (BRD) has substantial impact on the welfare and productivity of young calves on rearing and fattening farms. The clinical signs of BRD vary greatly and range from slight increases in the respiratory rate to severe dyspnea (Cockcroft, 2015). Also, affected animals regularly display an initial early increase in body core temperature in response to infections with viral and bacterial pathogens causing BRD (Grissett et al., 2015). The typical fever peak a few hours or days after infection is commonly unnoticed by farm personnel. One reason lies in the common technique to detect an increased body core temperature which is to insert a thermometer into the rectum of the animals. This is labor-intensive as well as invasive and can compromise n Corresponding author. Present address: Rhine-Waal University of Applied Sciences, Livestock Science and Environmental Impacts, Marie-Curie-Strasse 1, 47533 Kleve, Germany. E-mail address: Steffi[email protected] (S. Wiedemann).
http://dx.doi.org/10.1016/j.livsci.2016.05.005 1871-1413/& 2016 Elsevier B.V. All rights reserved.
animal welfare. As a result, in herds in which BRD is enzootic more than 50% of the pre-weaned calves show ultrasonographic lung consolidations, but only in approximately 41% of these calves visual BRD signs were detected and treated by the producer (Buczinski et al., 2013). Previously, the application of remote reticuloruminal (ReRu) temperature measuring boluses has been shown to be a reliable and less invasive method for the detection of increased body core or rectal temperatures in dairy cows (Bewley et al., 2008a; Adams et al., 2013) and beef cattle (Rose-Dye et al., 2011; Timsit et al., 2011b). To our knowledge the measurement of ReRu temperature has not yet been described in pre-weaned animals. In general pre-weaned dairy calves are more frequently affected by BRD (prevalence on farms in the US of 12–16%) than calves after weaning (6–11%; Guterbock, 2014). Also, with an age of 14 days onwards male calves from different origins are often commingled and transferred to one barn for fattening which causes major stress and a greater risk of disease transmission (Svensson et al., 2003). Therefore, it was the objective of this study in pre-weaned calves to determine the relationship of the ReRu
B. Tietgen et al. / Livestock Science 189 (2016) 78–81
temperature with the rectal temperature and to determine the applicability of a cumulative sum (CUSUM) control chart to detect animals with clinical signs of BRD.
2. Material and methods The field trial was performed according to the requirements of the federal animal welfare committee of the ministry of Schleswig-Holstein (reference number: V 242.7224.121–26). The study was conducted on a commercial rearing and fattening beef bull farm in northern Germany between September 2013 and November 2013. Pre-weaned male Holstein-Friesian calves (n ¼150) with an initial age of 16.6 d (SD 73.3 d) and an initial weight of 53.8 kg (SD 75.2 kg) were bought from different farms and assigned to one of six straw bedded pens on the day of arrival. Calves were fed with a total of 38.2 kg milk replacer per calf by three automatic calf feeders over a period of 8 wk. Concentrates, hay and water were offered for ad libitum intake at all times. During the trial period of 8 wk calves’ average daily gain was 778775 g/d. The ReRu temperature was measured with an automatic temperature measuring device in the form of a bolus (70 g; 6.6 cm 2.1 cm; Medria, Châteaubourg, France). On the day of arrival the bolus was swallowed by all animals after application with a bolus gun for small ruminants. Each bolus had an internal battery and transmitted the temperature every 5 min via radio link to the base station. Prior to the start of the experiment, the precision of all temperature measuring devices was tested in a water bath (gradual increase by 1 °C from 35.0 °C up to a maximum of 41.5 °C). Mean values for the five min bolus-temperature values were calculated for 30-min intervals and day periods (means of periods from 06.00 to 09.59 h ¼‘morning’, 10.00– 13.59 h¼ ‘midday’, 14.00–17.59 h¼‘afternoon’, 18.00 h to 23.59 h ¼‘evening’, 00.00–05.59 h¼ ‘night’). Throughout the experiment two trained persons alternately inspected all animals at least twice a day in addition to the barn personnel. If a calf displayed visible signs of BRD such as depression, ocular/nasal discharge, enforced breathing or coughing a detailed clinical inspection was performed. This included the assessment of the general condition, the evaluation of the respiratory tract (respiratory rate, lung auscultations) and the measurement of the rectal temperature as an approximation of the body core temperature. Only one digital thermometer was used throughout the trial which was inserted to a depth of approximately 8 cm into the rectum and had contact to the rectum wall throughout the measurement (HS Digital-Thermometer; Henry Schein, Wien, Austria). Calves with rectal temperatures Z 40 °C and, at least, one additional sign of BRD were diagnosed with BRD and treated for this condition. Follow up examinations of calves took place as long as calves displayed clinical signs related to BRD. In total, 219 clinical examinations were performed on 79 calves. Nine calves died during the trial mainly displaying clinical signs of BRD. The pathological examination of two calves which died without prior clinical signs of BRD indicated an infection with M. haemolytica and a rapid development of general septicemia. One calf did not wake up from anesthesia during the dehorning process. The health treatment of the animals is described in detail in the Supplementary material for this article. The software package SAS 9.2 (SAS inst. Inc., Cry, NC, USA) was applied for the retrospective statistical analysis. The means of the ReRu temperatures for the respective time periods were analyzed separately (30-min interval and day periods). To reduce the effect of water intake which could not be observed for technical reasons all ReRu temperature values below 37 °C were excluded from analysis. The rectal temperatures obtained by the two trained persons at the health inspection procedures were regarded as
79
comparative values. Descriptive analysis was performed to evaluate the number of calves displaying high ReRu temperatures (Z40 °C; ReRu hyperthermia) and markedly increased rectal temperatures (Z40 °C) with one further clinical sign of BRD. Pearson rank coefficients of correlations and the regression between rectal and associated ReRu temperatures (30–min intervals and day periods) were assessed. The cumulative sum (CUSUM) control chart was applied for the early detection of increased rectal temperatures (De Vries and Reneau, 2010). In this study the differences between the 30-min interval bolus temperature and the reference temperature (mean of all ReRu temperatures: 39.04 °C) were accumulated for each calf. For the CUSUM control chart four parameters were specified: the reference temperature, the standard deviation (0.46 °C) and two determinants of the process limits (h¼3, k ¼1.5). The k-value is the reference or allowable value, which is related to the extent of change to be detected. The h-value determines the control limit; an overstepping implies that the process is ‘out of control’. The k- and h-value were manually tested for best detection of ReRu hyperthermia and appropriate h-and k-values were chosen (Krieter et al., 2009). A one-sided test was performed. Thereby, only values exceeding the upper limit triggered an alarm which in this case was regarded as ReRu hyperthermia. The reliability of the ReRu temperature measurements and of the CUSUM values to assess spontaneously occurring BRD cases was further determined with regard to their sensitivity (Se), specificity (Sp), positive predictive value (PPV), and negative predictive value (NPV) (Petrie and Watson, 2013). To assess the discrimination power between BRD and non-BRD cases and to calculate the optimal cut-off value for ReRu-temperature to detect BRD, receiver operating characteristic (ROC) curves were created using PROC LOGISTIC in SAS (Zweig and Campbell, 1993). The calculated cut-off value for ReRu-temperature to detect a BRD case was at 40.0 °C.
3. Results and discussion To the authors’ knowledge, this is the first report on the use of wireless temperature measuring ReRu boluses for the early detection of clinical signs of BRD in pre-weaned calves. Remote rumen temperature boluses were previously successfully tested to continuously monitor the body core or rectal temperature in dairy cattle, beef bulls and heifers with more than 260 kg live weight (Bewley et al., 2008a; Small et al., 2008; Rose-Dye et al., 2011; Timsit et al., 2011a; Adams et al., 2013). In the pre-trial water bath analysis, the precision and accuracy of the boluses were good (difference of 0.2 °C between the boluses and of 0.15 between mean of all boluses and the digital thermometer). Most boluses could be administered without any complications in two week old calves. No loss via feces or per regurgitation occurred during the trial period, although further longterm studies should address the sustainability and food safety of the boluses. In a few calves, the ventral throat of the calves had to be massaged in the direction of the sternum, which enabled the bolus to pass the base of the tongue. Over the trial period, more than 90% of the expected 5-min values were correctly transferred. In the beginning of the trial some data were not available due to connection problems between some boluses and the base station in the barn which could only be solved by repeated moving of the base station. In total, rectal temperatures Z40.0 °C and one further clinical sign were observed in 30 animals (Table 1). Slightly elevated rectal temperatures ( 439.4 °C, but o 40.0 °C) and one further clinical sign were detected in 14 animals and enforced breathing or other signs of respiratory distress without an increase in rectal temperature in 17 animals. The incidence of clinical BRD signs was
80
B. Tietgen et al. / Livestock Science 189 (2016) 78–81
Table 1 Number of animals with BRD rectal temperature Z40 °C þ one further sign of BRD) and reticulo-ruminal (ReRu) temperatures Z 40 °C over different time periods as well as number and duration of hyperthermia (hyp) periods per affected animal during the 8 trial weeks. Signs of BRD ReRu hyperthermia 5 min 30 min
No. of calves No. of hyp periods (Mean 7 SD)b Duration of hyp periods (h)b
30 1.39 7 0.7 337 17
Day periodsa
147 139 99 4247 489 64.57 76.1 7.17 8.0 357 41
327 38
307 34
a day period ¼06.00–09.59 h ¼ ‘morning’, 10.00 h to 13.59 h ¼‘midday’, 14.00 h to 17.59 h ¼ ‘afternoon’, 18.00 h to 23.59 h ¼ ‘evening’, 00.00 h to 05.59 h ¼ ‘night’ b only animals with respective hyperthermia periods were included in the calculation
highest during wk 3–4 (n ¼ 17) and wk 5–6 (n ¼29) of trial. Throughout the trial period diarrhea was noted in 14, thickened or infected navels in 22 and otitis in 2 animals, respectively. The correlations between rectal temperatures and both the corresponding 30 min as well as the day period ReRu temperatures was r ¼0.75 (r2 ¼0.56), respectively. These correlations were slightly higher than previous results in dairy cattle (r ¼0.65; Adams et al., 2013). In beef bulls a stronger association between rectal and ruminal temperatures were reported ranging from r ¼0.89 to r ¼0.91 (Rose-Dye et al., 2011; Timsit et al., 2011a). Timsit et al., (2011a) measured the correlation between ReRu temperature and rectal temperature during episodes of ruminal hyperthermia (4 39.9 °C lasting more than 6 h; r ¼0.89), which in our study still resulted in a lower associations (r ¼0.75). Rose-Dye et al., (2011) determined the correlation after a Mannheimia haemolytica challenge after which most animals displayed increased ruminal and rectal temperatures. A higher association was found between 30-min ReRu and rectal temperatures in the first three trial weeks (r ¼ 0.74) in comparison to the last three trial weeks (r ¼0.67). This could be explained by an increasing contribution of microbial fermentation to the ReRu temperature due to the developing fore stomach (Swanson and Harris, 1958). Also, a higher water intake due to increased concentrate intakes during later time periods could have had an effect on the ReRu temperature as was shown in adult dairy cattle (Bewley et al., 2008b). The mean number of animals displaying ReRu hyperthermia depended on the time periods for which data were averaged (Table 1). At least once, almost all calves transmitted increased ReRu temperatures (98% of calves; 2.6% of all 5-min values) or displayed 30-min interval temperatures above the threshold (93% of calves), whereas in 66% of all animals ReRu hyperthermia was determined over day periods. The mean duration of ReRu hyperthermia (32–55 h) in this study was shorter than the average duration of ruminal hyperthermia in 7–9 months old beef bulls (118 h 7112 h; Timsit et al., 2011b). For all animals with unnoticed visual signs of BRD an average number of 225 7237 ReRu hyperthermia 5-min values were transmitted, whereas 35.0 738.9 intervals of 30-min and 3.8 72.8 d periods indicated an outbreak of BRD. The duration of undetected ReRu hyperthermia periods was shorter in this study than previously reported in beef bulls (75% lasting less than 31 h vs. less than 42 h; Timsit et al., 2011a). In this regard, more than half of the day period values indicating ReRu hyperthermia were measured in the evening (25.7%) or at night (27.1). Higher body temperatures later in the day with concomitant fermentation processes in the rumen of the resting animal can lead to an increased ReRu temperature (Metz et al., 1987). Moreover, undetected short periods of ReRu hyperthermia might be explained with an effective immunological response of these
calves (Small et al., 2008). The animals were vaccinated during the peak of clinical cases of BRD. Although elevated numbers of ReRu hyperthermia periods were detected two days after vaccination, a clear separation of the effect of an immunological reaction to the vaccination from naturally occurring BRD was not possible. Besides, a rapid onset of the disease could reduce the chance of an early visual detection. Six out of eight calves that died of disease related causes during the trial period displayed ReRu hyperthermia periods up to 5 d before death, but three of them never showed obvious clinical symptoms. In the remaining two calves which displayed long lasting disease phases neither rectal nor ReRu hyperthermia were detected five days prior to death. Instead, in these animals periods of hypothermia were examined which warrants further investigations on the applicability of the bolus to detect a decline in body core temperature, e.g. due to diarrhea. Also, additional studies are needed to evaluate the economic relevance of undetected ReRu hyperthermia periods in pre-weaned calves under practical conditions which will help to refine the alarm threshold values. The evaluation of the ReRu temperature to predict spontaneously occurring BRD cases was based on the Se, Sp, PPV and NPV which generated good results particularly for the evaluation of the 30-min means and the CUSUM values (Table 2). It has to be taken into account that the significances of the obtained PPV and NPV are limited for other scenarios, e.g. for farms with a lower disease risk. Also, continuous instead of punctual measurements of the rectal temperatures would have probably resulted in higher values for these parameters because of the possible occurrence of short febrile events. The time difference between the excess of the upper limit of the CUSUM method and the time point at which the calf was noticed by health monitoring with clinical BRD was 79.7 h 770.1 h, which is longer than the previously reported 24 h (Timsit et al., 2011a). Up to 5 d prior to recognition of true fever by the personnel in all animals the 30-min ReRu values, day period values and the CUSUM indicated hyperthermia at least once (Fig. 1). The closer the time to the assessment of clinical BRD signs approached, the higher was the percentage of animals that were detected by ReRu temperature analysis. It has to be noted that one of the key objectives of this study was to assess the duration between first detection of increased ReRu hyperthermia and visual detection of BRD signs under field conditions. Therefore, rectal temperature was not assessed on a daily basis and possible phases of increased rectal temperatures with or without minor further signs of BRD might not have been detected. However, it can be assumed that under practical conditions without the regular visual inspection of two additional trained observers time periods until first detection of clinical signs might have been even longer which is supported by the previously reported low treatment rate of pre-weaned calves with consolidated lungs by the producer (41%; Buczinski et al., 2013). It still remains unknown whether an early detection might help to prevent contamination of pen mates by early isolation of one or few infected calves. Yet, an early detection and treatment of BRD drastically improves the chance of a complete recovery and reduces the need for antimicrobial treatments of an entire group of high-risk calves, e.g. upon arrival at a farm (Cusack Table 2 Sensitivity (Se), specificity (Sp), positive and negative predictive value (PPV and NPV) of reticulo-ruminal (ReRu) temperature measurements for the prediction of BRD (rectal temperature Z 40 °C þ one further sign of BRD; n¼ 197 measurements of rectal temperature). ReRu data
Se
Sp
PPV
NPV
AUC
Means of 30 min Means of day period CUSUM
77% 61% 71%
97% 97% 98%
73% 69% 86%
98% 97% 98%
0.969 0.958 0.855
B. Tietgen et al. / Livestock Science 189 (2016) 78–81
81
Landschaft’ for the first author is gratefully appreciated. We thank the staff of the fattening farm in Western-Pomerania, Northern Germany and V. Fengler and S. Langer for their assistance.
Appendix A. Supplementary material Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.livsci.2016.05.005.
Days before detection of BRD
Fig. 1. Percentage of detected calves five days before the observation of BRD (rectal temperature Z 40 °Cþ one further sign of BRD) using reticulo ruminal temperatures. Data was averaged over 30 min and day periods (ReRu hyp. ¼ Reticulo-ruminal hyperthermia, temperature Z 40.0 °C). The averaged 30 min values were further analyzed by the CUSUM control chart method.
et al., 2003). Nonetheless, increased ReRu temperatures should be validated by a detailed examination of the animal for clinical BRD signs (e.g. measurement of the rectal temperature and investigation of the respiratory tract) to ensure prudent use of antimicrobials. All animals with further clinical BRD symptoms should also be closely monitored because in 15 out of 29 animals (52%) that only displayed lethargy, coughing or other symptoms an increase in 30-min ReRu temperature values around the day of detection was observed (average 13.2 712.6 increased 30-min periods per animal). According to the information of the manufacturer the batteries of the boluses serve for two to three years. Although the battery life time was not tested in this study under field conditions this method has the potential to detect diseases which result in increased ReRu temperatures throughout the whole fattening period in cattle. In dairy herds the ReRu boluses could also be used for automatic evaluations during future life of heifers, e.g. observation of the estrus cycle or prediction of calving events (Cooper-Prado et al., 2011). Moreover, a continuous surveillance of many animals at one time is possible, which cannot be realized with regular time-consuming measurement of the rectal temperature in big cattle herds. In conclusion, automatic remote reticulo-ruminal temperature boluses are promising tools for the early detection of BRD in preweaned calves. They also have the potential to improve the responsible use of antimicrobials.
Conflict of interest None.
Acknowledgment The scholarship of the ‘Stiftung Schleswig-Holsteinische
References Adams, A.E., Olea-Popelka, F.J., Roman-Muniz, I.N., 2013. Using temperature-sensing reticular boluses to aid in the detection of production diseases in dairy cows. J. Dairy Sci. 96, 1549–1555. Bewley, J.M., Einstein, M.E., Grott, M.W., Schutz, M.M., 2008a. Comparison of reticular and rectal core body temperatures in lactating dairy cows. J. Dairy Sci. 91, 4661–4672. Bewley, J.M., Grott, M.W., Einstein, M.E., Schutz, M.M., 2008b. Impact of intake water temperatures on reticular temperatures of lactating dairy cows. J. Dairy Sci. 91, 3880–3887. Buczinski, S., Forté, G., Francoz, D., Bélanger, A.M., 2013. Comparison of thoracic auscultation, clinical score, and ultrasonography as indicators of bovine respiratory disease in preweaned dairy calves. J. Vet. Int. Med 28, 234–242. Cockcroft, P., 2015. Bovine medicine. Wiley. Cooper-Prado, M.J., Long, N.M., Wright, E.C., Goad, C.L., Wettemann, R.P., 2011. Relationship of ruminal temperature with parturition and estrus of beef cows. J. Anim. Sci. 89, 1020–1027. Cusack, P.M.V., McMeniman, N., Lean, I.J., 2003. The medicine and epidemiology of bovine respiratory disease in feedlots. Aust. Vet. J. 81, 480–487. De Vries, A., Reneau, J.K., 2010. Application of statistical process control charts to monitor changes in animal production systems. J. Anim. Sci. 88, E11–E24. Grissett, G.P., White, B.J., Larson, R.L., 2015. Structured literature review of responses of cattle to viral and bacterial pathogens causing bovine respiratory disease complex. J. Vet. Intern. Med. 29, 770–780. Guterbock, W.M., 2014. The impact of BRD: the current dairy experience. Animal health research reviews/Conference of Research Workers in Animal Diseases 15, 130–134. Krieter, J., Engler, J., Tölle, K.H., Timm, H.H., Hohls, E., 2009. Control charts applied to simulated sow herd datasets. Livest. Sci. 121, 281–287. Metz, J., Wiersma, F., Rossing, W., Van Den Berg, V., 1987. First experiences with a telemetry system for measurements of body temperature by dairy cows. 3rd Symp. Automation in Dairying, Wageningen, the Netherlands. IMAG., 185–197. Petrie, A., Watson, P., 2013. Statistics for veterinary and animal science. Wiley,. Rose-Dye, T.K., Burciaga-Robles, L.O., Krehbiel, C.R., Step, D.L., Fulton, R.W., Confer, A.W., Richards, C.J., 2011. Rumen temperature change monitored with remote rumen temperature boluses after challenges with bovine viral diarrhea virus and Mannheimia haemolytica. J. Anim. Sci. 89, 1193–1200. Small, J.A., Kennedy, A.D., Kahane, S.H., 2008. Core body temperature monitoring with passive transponder boluses in beef heifers. Can. J. Anim. Sci. 88, 225–235. Svensson, C., Lundborg, K., Emanuelson, U., Olsson, S.-O., 2003. Morbidity in Swedish dairy calves from birth to 90 days of age and individual calf-level risk factors for infectious diseases. Prev. Veter. Med. 58, 179–197. Swanson, E.W., Harris Jr., J.D., 1958. Development of rumination in the young calf. J. Dairy Sci. 41, 1768–1776. Timsit, E., Assie, S., Quiniou, R., Seegers, H., Bareille, N., 2011a. Early detection of bovine respiratory disease in young bulls using reticulo-rumen temperature boluses. Vet. J. 190, 136–142. Timsit, E., Bareille, N., Seegers, H., Lehebel, A., Assie, S., 2011b. Visually undetected fever episodes in newly received beef bulls at a fattening operation: occurence, duration and impact on performance. J. Anim. Sci. 89, 4272–4280. Zweig, M.H., Campbell, G., 1993. Receiver-operating characteristic (ROC) plots: a fundamental evaluation tool in clinical medicine. Clin. Chem. 39, 561–577.
Contents lists available at ScienceDirect
Livestock Science journal homepage: www.elsevier.com/locate/livsci
Short communication
Field-trial evaluation of an automatic temperature measurement device placed in the reticulo-rumen of pre-weaned male calves Birte Tietgen a, Hans.-Joachim. Laue b, Martina Hoedemaker c, Steffi Wiedemann a,n a
University of Kiel, Animal Health, Institute of Animal Breeding and Husbandry, 24098 Kiel, Germany University of Applied Science Kiel, Animal Feeding, 24783 Osterrönfeld, Germany c University of Veterinary Medicine, Clinic for Cattle, 30173 Hannover, Germany b
art ic l e i nf o
a b s t r a c t
Article history: Received 8 July 2015 Received in revised form 8 May 2016 Accepted 11 May 2016
The assessment of the rectal temperature plays an important role in the early detection of bovine respiratory disease (BRD) in young calves, but the measurement is invasive and labor-intensive. The aims of this retrospective field-trial evaluation were to assess the correlation between the automatic measurements of the reticulo-ruminal (ReRu) temperature and the rectal temperature and to evaluate the diagnostic accuracy of a cumulative sum (CUSUM) control chart to detect pre-weaned calves suffering from BRD. In 150 male fattening calves (16.6 73.3 d at arrival) the ReRu-temperature was obtained every 5 min automatically with a wireless device over a period of 8 weeks. Data was averaged over periods of 30 min and 4–6 h (day periods). The 30-min means were further evaluated using CUSUM control charts. All calves were inspected by trained persons at least twice a day and rectal temperatures were measured in calves showing visible signs related to BRD. A ReRu hyperthermia (ReRu temperature Z40 °C) was detected in 139 calves over 30 min and in 99 calves over day periods, respectively. During the evaluation period 30 animals were affected by BRD (rectal temperatures Z 40 °C and one additional clinical sign). The correlations between rectal temperatures and both the corresponding 30 min as well as the day period ReRu temperatures was r ¼0.75. The sensitivity (Se) and specificity (Sp) of ReRu hyperthermia for the 30 min means to detect BRD were 77% and 97%, respectively, whereas means of the day periods had a Se of 61% and a Sp of 97%. The CUSUM test revealed a Se of 71% and Sp of 98% to detect BRD. On average, by the CUSUM method calves with clinical signs of BRD were identified 3.5 d earlier. In conclusion, the automated ReRu temperature measurement possesses potential for the early detection of febrile responses in very young calves warranting further investigations. & 2016 Elsevier B.V. All rights reserved.
Keywords: Bolus Bovine respiratory disease Calves Fever Reticulo-rumen temperature
1. Introduction Bovine respiratory disease (BRD) has substantial impact on the welfare and productivity of young calves on rearing and fattening farms. The clinical signs of BRD vary greatly and range from slight increases in the respiratory rate to severe dyspnea (Cockcroft, 2015). Also, affected animals regularly display an initial early increase in body core temperature in response to infections with viral and bacterial pathogens causing BRD (Grissett et al., 2015). The typical fever peak a few hours or days after infection is commonly unnoticed by farm personnel. One reason lies in the common technique to detect an increased body core temperature which is to insert a thermometer into the rectum of the animals. This is labor-intensive as well as invasive and can compromise n Corresponding author. Present address: Rhine-Waal University of Applied Sciences, Livestock Science and Environmental Impacts, Marie-Curie-Strasse 1, 47533 Kleve, Germany. E-mail address: Steffi[email protected] (S. Wiedemann).
http://dx.doi.org/10.1016/j.livsci.2016.05.005 1871-1413/& 2016 Elsevier B.V. All rights reserved.
animal welfare. As a result, in herds in which BRD is enzootic more than 50% of the pre-weaned calves show ultrasonographic lung consolidations, but only in approximately 41% of these calves visual BRD signs were detected and treated by the producer (Buczinski et al., 2013). Previously, the application of remote reticuloruminal (ReRu) temperature measuring boluses has been shown to be a reliable and less invasive method for the detection of increased body core or rectal temperatures in dairy cows (Bewley et al., 2008a; Adams et al., 2013) and beef cattle (Rose-Dye et al., 2011; Timsit et al., 2011b). To our knowledge the measurement of ReRu temperature has not yet been described in pre-weaned animals. In general pre-weaned dairy calves are more frequently affected by BRD (prevalence on farms in the US of 12–16%) than calves after weaning (6–11%; Guterbock, 2014). Also, with an age of 14 days onwards male calves from different origins are often commingled and transferred to one barn for fattening which causes major stress and a greater risk of disease transmission (Svensson et al., 2003). Therefore, it was the objective of this study in pre-weaned calves to determine the relationship of the ReRu
B. Tietgen et al. / Livestock Science 189 (2016) 78–81
temperature with the rectal temperature and to determine the applicability of a cumulative sum (CUSUM) control chart to detect animals with clinical signs of BRD.
2. Material and methods The field trial was performed according to the requirements of the federal animal welfare committee of the ministry of Schleswig-Holstein (reference number: V 242.7224.121–26). The study was conducted on a commercial rearing and fattening beef bull farm in northern Germany between September 2013 and November 2013. Pre-weaned male Holstein-Friesian calves (n ¼150) with an initial age of 16.6 d (SD 73.3 d) and an initial weight of 53.8 kg (SD 75.2 kg) were bought from different farms and assigned to one of six straw bedded pens on the day of arrival. Calves were fed with a total of 38.2 kg milk replacer per calf by three automatic calf feeders over a period of 8 wk. Concentrates, hay and water were offered for ad libitum intake at all times. During the trial period of 8 wk calves’ average daily gain was 778775 g/d. The ReRu temperature was measured with an automatic temperature measuring device in the form of a bolus (70 g; 6.6 cm 2.1 cm; Medria, Châteaubourg, France). On the day of arrival the bolus was swallowed by all animals after application with a bolus gun for small ruminants. Each bolus had an internal battery and transmitted the temperature every 5 min via radio link to the base station. Prior to the start of the experiment, the precision of all temperature measuring devices was tested in a water bath (gradual increase by 1 °C from 35.0 °C up to a maximum of 41.5 °C). Mean values for the five min bolus-temperature values were calculated for 30-min intervals and day periods (means of periods from 06.00 to 09.59 h ¼‘morning’, 10.00– 13.59 h¼ ‘midday’, 14.00–17.59 h¼‘afternoon’, 18.00 h to 23.59 h ¼‘evening’, 00.00–05.59 h¼ ‘night’). Throughout the experiment two trained persons alternately inspected all animals at least twice a day in addition to the barn personnel. If a calf displayed visible signs of BRD such as depression, ocular/nasal discharge, enforced breathing or coughing a detailed clinical inspection was performed. This included the assessment of the general condition, the evaluation of the respiratory tract (respiratory rate, lung auscultations) and the measurement of the rectal temperature as an approximation of the body core temperature. Only one digital thermometer was used throughout the trial which was inserted to a depth of approximately 8 cm into the rectum and had contact to the rectum wall throughout the measurement (HS Digital-Thermometer; Henry Schein, Wien, Austria). Calves with rectal temperatures Z 40 °C and, at least, one additional sign of BRD were diagnosed with BRD and treated for this condition. Follow up examinations of calves took place as long as calves displayed clinical signs related to BRD. In total, 219 clinical examinations were performed on 79 calves. Nine calves died during the trial mainly displaying clinical signs of BRD. The pathological examination of two calves which died without prior clinical signs of BRD indicated an infection with M. haemolytica and a rapid development of general septicemia. One calf did not wake up from anesthesia during the dehorning process. The health treatment of the animals is described in detail in the Supplementary material for this article. The software package SAS 9.2 (SAS inst. Inc., Cry, NC, USA) was applied for the retrospective statistical analysis. The means of the ReRu temperatures for the respective time periods were analyzed separately (30-min interval and day periods). To reduce the effect of water intake which could not be observed for technical reasons all ReRu temperature values below 37 °C were excluded from analysis. The rectal temperatures obtained by the two trained persons at the health inspection procedures were regarded as
79
comparative values. Descriptive analysis was performed to evaluate the number of calves displaying high ReRu temperatures (Z40 °C; ReRu hyperthermia) and markedly increased rectal temperatures (Z40 °C) with one further clinical sign of BRD. Pearson rank coefficients of correlations and the regression between rectal and associated ReRu temperatures (30–min intervals and day periods) were assessed. The cumulative sum (CUSUM) control chart was applied for the early detection of increased rectal temperatures (De Vries and Reneau, 2010). In this study the differences between the 30-min interval bolus temperature and the reference temperature (mean of all ReRu temperatures: 39.04 °C) were accumulated for each calf. For the CUSUM control chart four parameters were specified: the reference temperature, the standard deviation (0.46 °C) and two determinants of the process limits (h¼3, k ¼1.5). The k-value is the reference or allowable value, which is related to the extent of change to be detected. The h-value determines the control limit; an overstepping implies that the process is ‘out of control’. The k- and h-value were manually tested for best detection of ReRu hyperthermia and appropriate h-and k-values were chosen (Krieter et al., 2009). A one-sided test was performed. Thereby, only values exceeding the upper limit triggered an alarm which in this case was regarded as ReRu hyperthermia. The reliability of the ReRu temperature measurements and of the CUSUM values to assess spontaneously occurring BRD cases was further determined with regard to their sensitivity (Se), specificity (Sp), positive predictive value (PPV), and negative predictive value (NPV) (Petrie and Watson, 2013). To assess the discrimination power between BRD and non-BRD cases and to calculate the optimal cut-off value for ReRu-temperature to detect BRD, receiver operating characteristic (ROC) curves were created using PROC LOGISTIC in SAS (Zweig and Campbell, 1993). The calculated cut-off value for ReRu-temperature to detect a BRD case was at 40.0 °C.
3. Results and discussion To the authors’ knowledge, this is the first report on the use of wireless temperature measuring ReRu boluses for the early detection of clinical signs of BRD in pre-weaned calves. Remote rumen temperature boluses were previously successfully tested to continuously monitor the body core or rectal temperature in dairy cattle, beef bulls and heifers with more than 260 kg live weight (Bewley et al., 2008a; Small et al., 2008; Rose-Dye et al., 2011; Timsit et al., 2011a; Adams et al., 2013). In the pre-trial water bath analysis, the precision and accuracy of the boluses were good (difference of 0.2 °C between the boluses and of 0.15 between mean of all boluses and the digital thermometer). Most boluses could be administered without any complications in two week old calves. No loss via feces or per regurgitation occurred during the trial period, although further longterm studies should address the sustainability and food safety of the boluses. In a few calves, the ventral throat of the calves had to be massaged in the direction of the sternum, which enabled the bolus to pass the base of the tongue. Over the trial period, more than 90% of the expected 5-min values were correctly transferred. In the beginning of the trial some data were not available due to connection problems between some boluses and the base station in the barn which could only be solved by repeated moving of the base station. In total, rectal temperatures Z40.0 °C and one further clinical sign were observed in 30 animals (Table 1). Slightly elevated rectal temperatures ( 439.4 °C, but o 40.0 °C) and one further clinical sign were detected in 14 animals and enforced breathing or other signs of respiratory distress without an increase in rectal temperature in 17 animals. The incidence of clinical BRD signs was
80
B. Tietgen et al. / Livestock Science 189 (2016) 78–81
Table 1 Number of animals with BRD rectal temperature Z40 °C þ one further sign of BRD) and reticulo-ruminal (ReRu) temperatures Z 40 °C over different time periods as well as number and duration of hyperthermia (hyp) periods per affected animal during the 8 trial weeks. Signs of BRD ReRu hyperthermia 5 min 30 min
No. of calves No. of hyp periods (Mean 7 SD)b Duration of hyp periods (h)b
30 1.39 7 0.7 337 17
Day periodsa
147 139 99 4247 489 64.57 76.1 7.17 8.0 357 41
327 38
307 34
a day period ¼06.00–09.59 h ¼ ‘morning’, 10.00 h to 13.59 h ¼‘midday’, 14.00 h to 17.59 h ¼ ‘afternoon’, 18.00 h to 23.59 h ¼ ‘evening’, 00.00 h to 05.59 h ¼ ‘night’ b only animals with respective hyperthermia periods were included in the calculation
highest during wk 3–4 (n ¼ 17) and wk 5–6 (n ¼29) of trial. Throughout the trial period diarrhea was noted in 14, thickened or infected navels in 22 and otitis in 2 animals, respectively. The correlations between rectal temperatures and both the corresponding 30 min as well as the day period ReRu temperatures was r ¼0.75 (r2 ¼0.56), respectively. These correlations were slightly higher than previous results in dairy cattle (r ¼0.65; Adams et al., 2013). In beef bulls a stronger association between rectal and ruminal temperatures were reported ranging from r ¼0.89 to r ¼0.91 (Rose-Dye et al., 2011; Timsit et al., 2011a). Timsit et al., (2011a) measured the correlation between ReRu temperature and rectal temperature during episodes of ruminal hyperthermia (4 39.9 °C lasting more than 6 h; r ¼0.89), which in our study still resulted in a lower associations (r ¼0.75). Rose-Dye et al., (2011) determined the correlation after a Mannheimia haemolytica challenge after which most animals displayed increased ruminal and rectal temperatures. A higher association was found between 30-min ReRu and rectal temperatures in the first three trial weeks (r ¼ 0.74) in comparison to the last three trial weeks (r ¼0.67). This could be explained by an increasing contribution of microbial fermentation to the ReRu temperature due to the developing fore stomach (Swanson and Harris, 1958). Also, a higher water intake due to increased concentrate intakes during later time periods could have had an effect on the ReRu temperature as was shown in adult dairy cattle (Bewley et al., 2008b). The mean number of animals displaying ReRu hyperthermia depended on the time periods for which data were averaged (Table 1). At least once, almost all calves transmitted increased ReRu temperatures (98% of calves; 2.6% of all 5-min values) or displayed 30-min interval temperatures above the threshold (93% of calves), whereas in 66% of all animals ReRu hyperthermia was determined over day periods. The mean duration of ReRu hyperthermia (32–55 h) in this study was shorter than the average duration of ruminal hyperthermia in 7–9 months old beef bulls (118 h 7112 h; Timsit et al., 2011b). For all animals with unnoticed visual signs of BRD an average number of 225 7237 ReRu hyperthermia 5-min values were transmitted, whereas 35.0 738.9 intervals of 30-min and 3.8 72.8 d periods indicated an outbreak of BRD. The duration of undetected ReRu hyperthermia periods was shorter in this study than previously reported in beef bulls (75% lasting less than 31 h vs. less than 42 h; Timsit et al., 2011a). In this regard, more than half of the day period values indicating ReRu hyperthermia were measured in the evening (25.7%) or at night (27.1). Higher body temperatures later in the day with concomitant fermentation processes in the rumen of the resting animal can lead to an increased ReRu temperature (Metz et al., 1987). Moreover, undetected short periods of ReRu hyperthermia might be explained with an effective immunological response of these
calves (Small et al., 2008). The animals were vaccinated during the peak of clinical cases of BRD. Although elevated numbers of ReRu hyperthermia periods were detected two days after vaccination, a clear separation of the effect of an immunological reaction to the vaccination from naturally occurring BRD was not possible. Besides, a rapid onset of the disease could reduce the chance of an early visual detection. Six out of eight calves that died of disease related causes during the trial period displayed ReRu hyperthermia periods up to 5 d before death, but three of them never showed obvious clinical symptoms. In the remaining two calves which displayed long lasting disease phases neither rectal nor ReRu hyperthermia were detected five days prior to death. Instead, in these animals periods of hypothermia were examined which warrants further investigations on the applicability of the bolus to detect a decline in body core temperature, e.g. due to diarrhea. Also, additional studies are needed to evaluate the economic relevance of undetected ReRu hyperthermia periods in pre-weaned calves under practical conditions which will help to refine the alarm threshold values. The evaluation of the ReRu temperature to predict spontaneously occurring BRD cases was based on the Se, Sp, PPV and NPV which generated good results particularly for the evaluation of the 30-min means and the CUSUM values (Table 2). It has to be taken into account that the significances of the obtained PPV and NPV are limited for other scenarios, e.g. for farms with a lower disease risk. Also, continuous instead of punctual measurements of the rectal temperatures would have probably resulted in higher values for these parameters because of the possible occurrence of short febrile events. The time difference between the excess of the upper limit of the CUSUM method and the time point at which the calf was noticed by health monitoring with clinical BRD was 79.7 h 770.1 h, which is longer than the previously reported 24 h (Timsit et al., 2011a). Up to 5 d prior to recognition of true fever by the personnel in all animals the 30-min ReRu values, day period values and the CUSUM indicated hyperthermia at least once (Fig. 1). The closer the time to the assessment of clinical BRD signs approached, the higher was the percentage of animals that were detected by ReRu temperature analysis. It has to be noted that one of the key objectives of this study was to assess the duration between first detection of increased ReRu hyperthermia and visual detection of BRD signs under field conditions. Therefore, rectal temperature was not assessed on a daily basis and possible phases of increased rectal temperatures with or without minor further signs of BRD might not have been detected. However, it can be assumed that under practical conditions without the regular visual inspection of two additional trained observers time periods until first detection of clinical signs might have been even longer which is supported by the previously reported low treatment rate of pre-weaned calves with consolidated lungs by the producer (41%; Buczinski et al., 2013). It still remains unknown whether an early detection might help to prevent contamination of pen mates by early isolation of one or few infected calves. Yet, an early detection and treatment of BRD drastically improves the chance of a complete recovery and reduces the need for antimicrobial treatments of an entire group of high-risk calves, e.g. upon arrival at a farm (Cusack Table 2 Sensitivity (Se), specificity (Sp), positive and negative predictive value (PPV and NPV) of reticulo-ruminal (ReRu) temperature measurements for the prediction of BRD (rectal temperature Z 40 °C þ one further sign of BRD; n¼ 197 measurements of rectal temperature). ReRu data
Se
Sp
PPV
NPV
AUC
Means of 30 min Means of day period CUSUM
77% 61% 71%
97% 97% 98%
73% 69% 86%
98% 97% 98%
0.969 0.958 0.855
B. Tietgen et al. / Livestock Science 189 (2016) 78–81
81
Landschaft’ for the first author is gratefully appreciated. We thank the staff of the fattening farm in Western-Pomerania, Northern Germany and V. Fengler and S. Langer for their assistance.
Appendix A. Supplementary material Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.livsci.2016.05.005.
Days before detection of BRD
Fig. 1. Percentage of detected calves five days before the observation of BRD (rectal temperature Z 40 °Cþ one further sign of BRD) using reticulo ruminal temperatures. Data was averaged over 30 min and day periods (ReRu hyp. ¼ Reticulo-ruminal hyperthermia, temperature Z 40.0 °C). The averaged 30 min values were further analyzed by the CUSUM control chart method.
et al., 2003). Nonetheless, increased ReRu temperatures should be validated by a detailed examination of the animal for clinical BRD signs (e.g. measurement of the rectal temperature and investigation of the respiratory tract) to ensure prudent use of antimicrobials. All animals with further clinical BRD symptoms should also be closely monitored because in 15 out of 29 animals (52%) that only displayed lethargy, coughing or other symptoms an increase in 30-min ReRu temperature values around the day of detection was observed (average 13.2 712.6 increased 30-min periods per animal). According to the information of the manufacturer the batteries of the boluses serve for two to three years. Although the battery life time was not tested in this study under field conditions this method has the potential to detect diseases which result in increased ReRu temperatures throughout the whole fattening period in cattle. In dairy herds the ReRu boluses could also be used for automatic evaluations during future life of heifers, e.g. observation of the estrus cycle or prediction of calving events (Cooper-Prado et al., 2011). Moreover, a continuous surveillance of many animals at one time is possible, which cannot be realized with regular time-consuming measurement of the rectal temperature in big cattle herds. In conclusion, automatic remote reticulo-ruminal temperature boluses are promising tools for the early detection of BRD in preweaned calves. They also have the potential to improve the responsible use of antimicrobials.
Conflict of interest None.
Acknowledgment The scholarship of the ‘Stiftung Schleswig-Holsteinische
References Adams, A.E., Olea-Popelka, F.J., Roman-Muniz, I.N., 2013. Using temperature-sensing reticular boluses to aid in the detection of production diseases in dairy cows. J. Dairy Sci. 96, 1549–1555. Bewley, J.M., Einstein, M.E., Grott, M.W., Schutz, M.M., 2008a. Comparison of reticular and rectal core body temperatures in lactating dairy cows. J. Dairy Sci. 91, 4661–4672. Bewley, J.M., Grott, M.W., Einstein, M.E., Schutz, M.M., 2008b. Impact of intake water temperatures on reticular temperatures of lactating dairy cows. J. Dairy Sci. 91, 3880–3887. Buczinski, S., Forté, G., Francoz, D., Bélanger, A.M., 2013. Comparison of thoracic auscultation, clinical score, and ultrasonography as indicators of bovine respiratory disease in preweaned dairy calves. J. Vet. Int. Med 28, 234–242. Cockcroft, P., 2015. Bovine medicine. Wiley. Cooper-Prado, M.J., Long, N.M., Wright, E.C., Goad, C.L., Wettemann, R.P., 2011. Relationship of ruminal temperature with parturition and estrus of beef cows. J. Anim. Sci. 89, 1020–1027. Cusack, P.M.V., McMeniman, N., Lean, I.J., 2003. The medicine and epidemiology of bovine respiratory disease in feedlots. Aust. Vet. J. 81, 480–487. De Vries, A., Reneau, J.K., 2010. Application of statistical process control charts to monitor changes in animal production systems. J. Anim. Sci. 88, E11–E24. Grissett, G.P., White, B.J., Larson, R.L., 2015. Structured literature review of responses of cattle to viral and bacterial pathogens causing bovine respiratory disease complex. J. Vet. Intern. Med. 29, 770–780. Guterbock, W.M., 2014. The impact of BRD: the current dairy experience. Animal health research reviews/Conference of Research Workers in Animal Diseases 15, 130–134. Krieter, J., Engler, J., Tölle, K.H., Timm, H.H., Hohls, E., 2009. Control charts applied to simulated sow herd datasets. Livest. Sci. 121, 281–287. Metz, J., Wiersma, F., Rossing, W., Van Den Berg, V., 1987. First experiences with a telemetry system for measurements of body temperature by dairy cows. 3rd Symp. Automation in Dairying, Wageningen, the Netherlands. IMAG., 185–197. Petrie, A., Watson, P., 2013. Statistics for veterinary and animal science. Wiley,. Rose-Dye, T.K., Burciaga-Robles, L.O., Krehbiel, C.R., Step, D.L., Fulton, R.W., Confer, A.W., Richards, C.J., 2011. Rumen temperature change monitored with remote rumen temperature boluses after challenges with bovine viral diarrhea virus and Mannheimia haemolytica. J. Anim. Sci. 89, 1193–1200. Small, J.A., Kennedy, A.D., Kahane, S.H., 2008. Core body temperature monitoring with passive transponder boluses in beef heifers. Can. J. Anim. Sci. 88, 225–235. Svensson, C., Lundborg, K., Emanuelson, U., Olsson, S.-O., 2003. Morbidity in Swedish dairy calves from birth to 90 days of age and individual calf-level risk factors for infectious diseases. Prev. Veter. Med. 58, 179–197. Swanson, E.W., Harris Jr., J.D., 1958. Development of rumination in the young calf. J. Dairy Sci. 41, 1768–1776. Timsit, E., Assie, S., Quiniou, R., Seegers, H., Bareille, N., 2011a. Early detection of bovine respiratory disease in young bulls using reticulo-rumen temperature boluses. Vet. J. 190, 136–142. Timsit, E., Bareille, N., Seegers, H., Lehebel, A., Assie, S., 2011b. Visually undetected fever episodes in newly received beef bulls at a fattening operation: occurence, duration and impact on performance. J. Anim. Sci. 89, 4272–4280. Zweig, M.H., Campbell, G., 1993. Receiver-operating characteristic (ROC) plots: a fundamental evaluation tool in clinical medicine. Clin. Chem. 39, 561–577.