Bovine respiratory disease complex continues to be an important health issue in the beef cattle industry.1–5 In a national survey, > 33% of cow-calf producers agreed or strongly agreed that BRDC is economically important to their operation.6 In the same study,6 respiratory tract disease was reported as the cause of death for 8.2% of the calves that died before weaning. In 2 multiyear studies,7,8 the cumulative incidence of preweaning BRDC within herds ranged from 3.3% to 23.5% and 8.5% to 65.4%. Calves with preweaning BRDC had mean body weights that were 16.5 and 7.7 kg (36.3 and 16.9 lb) less than the mean body weights of their cohorts at weaning.7,9 Morbidity associated with preweaning BRDC can negatively affect calf sale weight and can result in deaths among preweaned calves, thereby reducing the weights and number of calves that are sold by cow-calf operators and impacting herd profitability.
Risk factors for BRDC for both feedlot cattle and preweaned dairy calves have been well documented.10–14 However, management of preweaned beef calves is much different from that of preweaned dairy calves; therefore, it cannot be assumed that risk factors for preweaning BRDC in the beef cattle industry would be similar to those in the dairy industry. To our knowledge, there are no reports of large-scale studies to investigate management factors associated with preweaning BRDC in cow-calf herds.
The USDA National Animal Health Monitoring System Beef 2007–2008 study generated data on management factors and disease occurrences for US cow-calf herds.15 The purpose of the study reported here was to use those data to provide a unique assessment of potential associations between cow-calf herd management practices and the rate of BRDC in preweaned beef calves. Identification of cow-calf management factors that are associated with BRDC-related morbidity in preweaned calves would allow further focus on potential management changes or preventive medicine approaches that are most likely to offer the largest positive impact on this important disease syndrome.
Materials and Methods
Survey design—Data for this study were collected in phases I and II.15 From a sampling frame maintained by the National Agricultural Statistics Service, which included all known cow-calf operations in 24 selected states, a stratified random sample of operations was selected. The survey was a single-stage design with sampling across strata (on the basis of state and herd size) without replacement. Herd size on January 1, 2007, within state was the stratification variable. States included in this survey were Alabama, Arkansas, California, Colorado, Florida, Georgia, Idaho, Iowa, Kansas, Kentucky, Louisiana, Mississippi, Missouri, Montana, Nebraska, New Mexico, North Dakota, Oklahoma, Oregon, South Dakota, Tennessee, Texas, Virginia, and Wyoming. States were chosen on the basis of the stated goal that the survey would represent at least 70% of the beef cows and cow-calf producers in the United States.
Survey data—A total of 4,001 producers were selected to participate in phase I (General Beef Management Report) of the survey. The phase I questionnaire contained 110 questions and was administered face-to-face by trained enumerators from the National Agricultural Statistics Service between October 22 and November 30, 2007. Individuals that completed phase I were invited to participate in phase II. Phase II of the study (Veterinary Services Initial Visit) involved a questionnaire that contained 51 questions, which was administered by veterinarians representing the USDA APHIS in January 2008.
Of those 4,001 producers selected during phase I, 1,033 (25.8%) provided complete information and were eligible for participation in phase II. Of those eligible producers, 567 (54.9%) consented to participate in phase II of the study. Reasons for nonparticipation of eligible producers in phase II included refusal to participate (35.3%), inability to be contacted (7.8%), and having no beef cows on the operation (2.0%). The initial dataset used for the present study contained answers given by the 567 participants who completed both survey phases I and II. Twenty-four states were represented in which 28.6 million (87.8%) of the beef cows and 603,000 (79.6%) of the cow-calf operations in the United States were located.16
Data validation for phase I was performed by each state's National Agricultural Statistics Service personnel, and data validation for phase II was performed by the USDA Center for Epidemiology and Animal Health staff.15 Additional data validation was completed by the National Animal Health Monitoring System staff after the data sets from each state were combined. All respondents’ data were statistically weighted to reflect the population from which the individuals were selected. The weights represented the inverse of the probability of being selected and included adjustments for nonresponse.17
Statistical analysis—All data analyses were performed with a statistical software programa that provides a platform for assessing survey data with weighted responses. The outcome variable of interest for each herd was the rate of respiratory tract disease in calves during the interval between birth and weaning (ie, BRDC rate). Herd-level rates were defined on the basis of reported counts of calves treated with an antimicrobial for respiratory tract disease and total calf-days at risk from birth to weaning. Calf-days at risk for each herd were initially determined as the reported number of calves born alive multiplied by the reported mean number of days from birth to weaning. Calf-days at risk were then adjusted on the basis of any calf deaths and respiratory tract-associated morbidity that developed prior to weaning. The reported count of calf deaths for 3 periods (birth to 24 hours after birth, 24 hours to 3 weeks after birth, and 3 weeks after birth to weaning) was contained in the data set. Morbidity was not reported for each of these periods but only as a total count for illnesses that developed prior to weaning. All calf-days for calves that died within 24 hours after birth were subtracted from the initial calf-day count at risk estimate. For calf deaths during other periods and for all preweaning morbidities, calves contributed only half (midpoint) of their calf-days at risk for the respective period.
A total of 804 survey questions were available as independent variables for the study. Many variables (eg, type of animal identification used) were not retained for analysis given the lack of plausibility for potential associations with the outcome of interest. Descriptive analyses, including histograms and scatterplots, were completed for each variable. Categorical variables for which 1 level represented > 95% of the herds were excluded from further analysis. Contingency tables were used to assess the similarities of variables among herds. For variables that had > 95% of the respondents answering similarly to multiple related questions, a single variable was created to represent responses to these related questions. For example, animals administered a BVDV vaccine and animals administered an IBR vaccine were included in 1 variable representing vaccination against both BVDV and IBR (as opposed to vaccination against infection with 1 pathogen only or no vaccination). After these initial data evaluation steps, which included assessments of plausibility, descriptive statistics, and correlations among variables and a combination of some variables into a single variable, 86 independent variables were further evaluated for inclusion in a multivariable model.
Associations between BRDC rate and various independent variables were evaluated in weighted analyses involving negative binomial regression models with the Taylor series linearized variance estimation and single sampling unit centered at the grand mean.17,18 Each independent variable was initially assessed in univariable models. Independent variables were retained for further analysis when the P value for an unconditional association was ≤ 0.30. Continuous independent variables were assessed for the assumption of linearity by categorization into quantiles. When estimated effects of quantiles indicated nonlinearity, the variable was entered into the model via the categorical coding. To check for potential multicollinearity, associations among variables that were retained following the univariable analysis (P ≤ 0.30) were assessed by means of a χ2 test and Spearman correlation for binary and nominal or ordinal variables, respectively. In addition, collinearity was assessed by the addition of 1 covariate at a time to the bivariable analyses of other possible collinear variables.
Variables that remained following the univariable analyses were entered into a multivariable model, and manual backward selection was used to select independent variables that were significantly (P ≤ 0.05) associated with the outcome. In addition, herd size was assessed both as a potentially important predictive variable and potential confounder of other management factors. Each variable that was not retained during the initial backward elimination process was later reoffered to the model to reassess significance and check for confounding. Any variable reoffered to the final model that was significant or resulted in a coefficient change ≥ 20% for any other variable was retained in the model. This process continued until no variables reoffered to the model were eligible for retention. After the final model was determined, comparisons among levels within variables were made via adjusted Wald tests.
Results
Sample—As reported by the 567 producers who completed phases I and II of the survey, 443 operations had calves born alive during the study period and yielded data on BRDC in preweaned calves; unweighted descriptive statistics for these herds were determined. Among the 443 operations, the median number of beef cows on the operation was 128 (range, 2 to 5,847) on January 1, 2008; the mean ± SD number of beef cows on the operation was 269.7 ± 467.2. The median and mean percentage of the herds composed of animals > 10 years of age was 10.6% (range, 0% to 82.6%) and 13.5 ± 13.6%, respectively. The median and mean percentage of cows < 5 years of age was 43.1% (range, 0% to 100%) and 42.9 ± 21.6, respectively. For these herds, the median number of calves born alive in 2007 was 125 (range, 1 to 6,400). For the calves born alive, the mean percentage of calves affected by respiratory tract disease was 3.0 ± 7.1% and the mean interval from birth to weaning was 207.2 ± 35.5 days (median, 210 days [range, 90 to 300 days]). The mean number of calf deaths in the interval from birth to 24 hours after birth was 3.1 ± 6.5 (median, 1.0 [range, 0 to 70]). The mean number of calf deaths in the interval from 24 hours to 3 weeks after birth was 3.3 ± 7.9 (median, 1 [range, 0 to 70]) and in the interval from 3 weeks after birth to weaning was 4.0 ± 10.4 (median, 2 [range, 0 to 125]). Mean BRDC rate was 1.5 ± 3.7 cases/10,000 calf-days (median, 0.18 cases/10,000 calf-days [range, 0 to 75.0 cases/10,000 calf-days]). The distribution of within-herd BRDC rates within the study population was determined (Figure 1).
Analysis results—Of the 86 variables assessed in the initial analyses, 21 were found to have an unconditional association (P ≤ 0.30) with the outcome variable (Table 1). Crude IRRs for these variables provided unadjusted estimates for the magnitude and direction of effects for factors associated with calf respiratory tract disease rates. Results of these analyses indicated that several variables of interest were not associated (P > 0.30) with BRDC rates. Herd size, divided in quartiles, and the proportion of cows < 5 years of age in the herd were not associated with BRDC rates. Certain disease management variables were also not associated with BRDC rates, including how often the calving pen was used to house sick cows; vaccination of 22-day-old to weaning-age calves with a 4-way vaccine against IBR and BVDV, parainfluenza-3 virus, and BRSV infections; consultation with a veterinarian regarding disease prevention; number of defined breeding seasons; and the separation of cow-calf pairs from pregnant cows. General management practices, including mean number of days from birth to dehorning, use of creep feed, use of artificial insemination, and body condition scoring, also were not associated with the outcome variable.
Results of univariable negative binomial regression analyses indicating management practices that were associated (P ≤ 0.30) with the rate of respiratory tract disease (BRDC) among preweaned beef calves in 443 US cow-calf operations for which data were obtained in a 2-phase survey.
Variable | No. of operations | IRR | 95% CI of the IRR |
---|---|---|---|
Vaccination of 22-day-old to weaned calves against Mannheimia (Pasteurella) sp infection | |||
No | 293 | ||
Yes | 149 | 0.55 | 0.30–1.01 |
Vaccination of cows with 4-way vaccine against IBR and BVD, parainfluenza-3 virus, and BRSV infections | |||
No | 260 | ||
Yes | 182 | 1.61 | 0.68–3.80 |
Provision of antimicrobials in feed given to calves to prevent respiratory tract disease | |||
No | 408 | ||
Yes | 35 | 2.65 | 0.89–7.92 |
No of times calves vaccinated for respiratory disease from birth to weaning* | |||
0 | 67 | ||
1 | 111 | 2.82 | 1.04–7.69 |
2 | 175 | 2.79 | 1.09–7.18 |
3 | 52 | 0.99 | 0.36–2.76 |
Importation of preweaned calves with dams | |||
No | 418 | ||
Yes | 25 | 2.99 | 1.00–8.93 |
Importation of bred heifers | |||
No | 402 | ||
Yes | 41 | 0.32 | 0.13–0.77 |
Importation of weaned steers | |||
No | 412 | ||
Yes | 31 | 2.83 | 0.82–9.74 |
No. of visits to the operation made by visitors in an average month† | |||
0 | 33 | ||
1 or 2 | 79 | 2.02 | 0.37–11.05 |
3–5 | 92 | 0.49 | 0.10–2.43 |
6–30 | 114 | 0.39 | 0.08–1.82 |
> 30 | 116 | 1.03 | 0.22–4.83 |
Best description of genetic composition of calves | |||
Single breed | 122 | ||
Composite | 32 | 2.32 | 1.08–4.99 |
2-breed cross | 196 | 3.09 | 1.37–6.97 |
3-breed cross | 93 | 4.23 | 1.70–10.53 |
No. of live calves with horns born this year ≥ 1 | |||
No | 234 | ||
Yes | 209 | 1.92 | 0.84–4.36 |
Mean weaning weight of replacements (kg) | |||
< 228 | 130 | ||
229–250 | 37 | 0.76 | 0.27–2.15 |
> 250–273 | 126 | 2.11 | 0.71–6.33 |
> 273 | 43 | 1.00 | 0.40–2.51 |
Mean time from birth to weaning (d) | |||
< 180 | 132 | ||
181–210 | 162 | 1.51 | 0.64–3.56 |
211–230 | 40 | 2.23 | 0.41–12.02 |
231–315 | 109 | 0.53 | 0.23–1.18 |
Do the same people buy the operation's calves each year? | |||
No | 209 | ||
Yes | 232 | 1.78 | 0.83–3.84 |
Classification of cow-calf operation as a primary, supplemental, or other source of income | |||
Primary | 209 | ||
Supplemental | 209 | 0.68 | 0.26–1.80 |
Other | 25 | 0.31 | 0.08–1.19 |
Percentage of operator's total hours worked (on and off operation) devoted to the cow-calf operation | |||
< 25 | 113 | ||
26–50 | 115 | 0.79 | 0.32–1.94 |
51–99 | 78 | 2.43 | 0.69–8.55 |
100 | 137 | 0.70 | 0.29–1.69 |
Use of estrus synchronization | |||
No | 349 | ||
Yes | 94 | 1.51 | 1.08–2.13 |
Use of ultrasonography to detect pregnancy | |||
No | 395 | ||
Yes | 48 | 1.35 | 0.82–2.32 |
Use of pelvic measurement as a reproductive management tool | |||
No | 381 | ||
Yes | 62 | 3.46 | 1.11–10.76 |
Duration of labor before intervention is initiated (h) | |||
≤ 2 | 261 | ||
> 2 | 141 | 0.57 | 0.25–1.31 |
No. of cows treated for any disease except respiratory tract disease < 1 | |||
No | 243 | ||
Yes | 196 | 1.58 | 0.69–3.65 |
No. of cows treated for respiratory tract disease ≥ 1 | |||
No | 373 | ||
Yes | 66 | 1.98 | 0.61–6.42 |
Vaccination against ≥ 1 of the following infections: IBR and infections with BVDV, parainfluenza-3 virus, BRSV, Mannheimia haemolytica, Pasteurella multocida, and Histophilus somni.
Visitors included employees, veterinarians, neighbors, nutritionists, commercial haulers, and others.
The final multivariable model included 6 independent variables that were significantly associated with herd-level BRDC rates (Table 2). To assess the form of the model that best fit the data (Poisson or negative binomial), the dispersion parameter point estimate (α) and 95% CIs were evaluated. Because the 95% CI for the dispersion parameter (2.53 to 4.54) did not include 0, the negative binomial model was deemed the most appropriate.19 Although the reported number of visits by outsiders in an average month (each individual person being counted as a separate visit) was significantly associated with the outcome, there were no differences between the referent category (0 visits by outsiders/mo) and effects for a herd that received 1 or 2 visits/mo (P … 0.256), 3 to 5 visits/mo (P … 0.316), 6 to 30 visits/mo (P … 0.147), or > 30 visits (P … 0.668). However, the direction of the effects varied: operations for which 1 or 2 visits/mo were reported had higher BRDC rates than did the operations for which 3 to 5 visits/mo (P … 0.008) or 6 to 30 visits/mo (P … 0.001) were reported. Similarly, operations for which 3 to 5 visits (P … 0.020) or 6 to 30 visits (P … 0.001) were reported had lower BRDC rates than did the operations for which > 30 visits/mo were reported. There was no difference (P … 0.273) in BRDC rate in herds that received 1 or 2 visits/mo and herds that received > 30 visits/mo. Compared with single-breed herds, herds with calves from 2-breed crosses (P … 0.005) and 3 or more breed crosses (P ≤ 0.001) had higher BRDC rates. Compared with herds with single-breed calves, herds with composite calves had higher BRDC rates, although the difference was not significant (P … 0.051). Additionally, cow-calf operations that respondents classified as supplemental sources of income had lower BRDC rates (P … 0.015), compared with operations that respondents classified as primary sources of income. The BRDC rate for herds that were classified as sources of income other than primary or supplemental (ie, kept for pleasure) did not differ from the rates for herds kept as primary (P … 0.087) or supplemental (P … 0.522) sources of income.
Results of a multivariable negative binomial regression model indicating management practices that were associated (P ≤ 0.05) with the rate of respiratory tract disease (BRDC) among preweaned beef calves in 443 US cow-calf operations for which data were obtained in a 2-phase survey.
Variable | P value | Level | β | SE (β) | IRR | 95% CI of the IRR |
---|---|---|---|---|---|---|
Intercept | −8.89 | 0.59 | ||||
Provision of antimicrobials in feed given to calves to prevent respiratory tract disease | 0.008 | No | Referent | |||
Yes | 1.24 | 0.46 | 3.46 | 1.39–8.60 | ||
Importation of bred heifers | 0.013 | No | Referent | |||
Yes | −0.93 | 0.37 | 0.40 | 0.19–0.82 | ||
Importation of weaned steers | 0.022 | No | Referent | |||
Yes | 0.96 | 0.42 | 2.62 | 1.15–5.97 | ||
No. of visits to the operation made by visitors in an average month† | 0.001 | 0 | Referent | |||
1–2 | 0.72 | 0.63 | 2.06 | 0.59–7.13 | ||
3–5 | −0.56 | 0.55 | 0.57 | 0.19–1.70 | ||
6–30 | −0.78 | 0.53 | 0.46 | 0.16–1.31 | ||
> 30 | 0.23 | 0.52 | 1.26 | 0.45–3.51 | ||
Best description of genetic composition of calves | 0.001 | Single breed | Referent | |||
Composite | 0.82 | 0.42 | 2.27 | 1.00–5.16 | ||
2-breed cross | 0.86 | 0.31 | 2.36 | 1.30–4.29 | ||
3-breed cross | 1.39 | 0.37 | 4.00 | 1.93–8.31 | ||
Classification of cow-calf operation as a primary, supplemental, or other (ie, hobby) source of income | 0.029 | Primary | Referent | |||
Supplemental | −0.74 | 0.31 | 0.48 | 0.26–0.87 | ||
Other | −1.17 | 0.68 | 0.31 | 0.08–1.18 |
Dispersion parameter (α) estimate, 3.39 (95% CI, 2.53 to 4.54).
See Table 1 for remainder of key.
Discussion
Using survey data that were representative of the US cow-calf industry, we were able to generate useful information on management practices associated with BRDC rates in preweaned beef calves. In the final model, 6 variables were found to be associated with preweaning BRDC rates. These effects may be indicative of levels of pathogen exposure within the herd, genetic or breed factors associated with the disease syndrome, or simply general management characteristics of certain cow-calf operations that either affect BRDC rates or indicate preparedness for or a response to diagnoses of BRDC among preweaned calves. Considering that we cannot assess the temporal relationships among these associations relative to onset of BRDC because of the cross-sectional design of the present study, it is not appropriate to assume causal relationships. Further evaluation of the mechanisms for the detected associations and potential disease mitigation opportunities is required; however, the data obtained in the present study were derived from a unique assessment of potential risk factors for preweaning BRDC rates in US cow-calf herds and may be useful for profiling or evaluating herds that are likely to have BRDC rates that are higher than the mean national rate.
The mean cumulative risk of respiratory tract disease in preweaned calves (3.0%) in the present study was higher than that found in 1 previous US study9 but slightly lower than values in other studies.20,21 One major reason for the mean percentage in the present study being higher than the 0.5% reported by Wittum et al9 may be that those researchers used a shorter period of time at risk (45 days) in their analysis. The interval from birth to weaning in the study by Dewell et al20 was approximately 200 days, which is comparable to the interval reported by respondents in the present study; however, the risk estimate (4.6%) in that other study20 was based on a much smaller number of calves located in a single herd. In combination, the results of these studies9,20,21 indicate that although risk of preweaning BRDC is relatively low, it still may vary among different subsets of the national cow-calf population. An incidence rate for BRDC was not reported for either of the previous studies9,20; thus, unique data for US cow-calf herds are shown in the present study (Figure 1).
In the present study, the multivariable model revealed that 6 management practices were significantly associated with the reported preweaning BRDC rate. Operations for which respondents reported that antimicrobials were fed to preweaned calves to prevent respiratory tract disease had higher BRDC rates than the operations for which this practice was not reported. However, it is possible that this association is reflective of an effect, rather than a potential cause. Because of the cross-sectional design of the present study, the BRDC history for each study herd was not known. The feeding of antimicrobials to preweaned calves may have been initiated in response to a recent or current BRDC outbreak or may have been a preventative measure implemented in the year during which the survey was conducted because of the development of BRDC-related illnesses during the preceding year at those identified operations. Little research investigating the benefit of feeding antimicrobials to prevent development of BRDC in preweaned calves has been reported, to our knowledge, but the practice is thought to be beneficial for postweaned calves in some instances.1 Although the term prevention is used as per the wording of the survey question, the antimicrobials may have been used for disease control because very few feed-grade antimicrobials are labeled for BRDC prevention and producers may consider prevention and control as equivalent. Thus, care must be exercised when evaluating the association between feeding antimicrobials and BRDC rates because of the possibility that feeding antimicrobials was in response to present or previous BRDC outbreaks and the potential for this observed association to be attributable to other unmeasured factors. However, it is possible that feeding antimicrobials to preweaned calves for the prevention of respiratory tract disease is a true risk factor for higher BRDC rates at the herd level.
Importation of bred heifers was associated with lower rates of BRDC, but importation of weaned steers had a positive association with BRDC rates in the present study. Producers that market bred heifers may be more likely to practice complete health programs because of the high value of breeding stock and germ plasm.22 These complete health programs may include vaccination or other preventive health measures and management of genetic traits to affect risk factors such as dystocia and udder and teat conformation, which have been previously associated with development of BRDC in preweaned calves.21,23 It is plausible that herds that include bred heifers for market could have lower BRDC rates as a consequence of risk factor management, which subsequently results in lower BRDC rates among herds that import this class of animal. It is also plausible that herd managers have different health requirements for purchasing heifers than they do for purchasing steers. Exposure to BRDC pathogens through contact with other cattle populations, especially populations derived from multiple source herds, is believed to be a risk factor for preweaning BRDC.23 It is possible that steer purchasing occurs through marketing channels where exposure of animals to multiple pathogens is likely.24 The higher BRDC rates among herds in which steers were imported could be because the preweaned calves were exposed to BRDC pathogens not already present on the operation or because they were exposed to increased BRDC pathogen concentrations. Another possible explanation is the increased risk associated with importation of stocker cattle.4 The temporal relationships between importation of bred heifers or steers and their exposure to preweaned calves cannot be assessed in the present study; however, the data have suggested that importation of particular classes of cattle could potentially impact the BRDC rates among preweaned calves in cow-calf operations.
Biosecurity has been cited as an important management practice for control of preweaning BRDC, given that it may reduce BRDC pathogen introduction and transmission to calves.25 Pathogen transmission can occur not only via animal-to-animal contact, but also from environmental or fomite contamination; therefore, controlling the entry of outsiders to a cow-calf facility may be a logical biosecurity practice.25 Although the data set used in the present study did not allow assessment of the respondents’ biosecurity practices, it is possible that respondents who allowed visitors are cognizant of biosecurity as a disease control method and have effective biosecurity programs in place. The variable representing the number of outsider visits to a herd in an average month was significant in the final model, yet there were no significant differences in pairwise comparisons of herds that had 0 visits/mo and other herd-visit categories. In comparisons among herds that received at least 1 visit/mo, we found that herds that received 1 or 2 visits/mo or > 30 visits/mo had higher BRDC rates, compared with herds that received 3 to 5 visits/mo or 6 to 30 visits/mo. It is possible that these effects indicate that personnel at operations with few visitors are less concerned with biosecurity than are personnel at operations with moderate numbers of visitors. These results may also indicate that large numbers of visits may overwhelm some biosecurity programs at cow-calf operations, making them less effective than similar programs implemented for operations with moderate numbers of visits. In the questionnaire used for the present study, respondents were asked to count each individual person as a separate visit. It is possible that most visits did not involve contact with either outsider animals or herd animals, being instead visits unrelated to nonlivestock exposures. The overall variability in effects among different levels of the variable representing the reported number of herd visits in an average month and the lack of specificity in defining types of visitors (eg, employees, veterinarians, and cattle buyers), combined with the cross-sectional nature of the data, make it difficult to determine potential mechanisms for the observed associations. Furthermore, when evaluating the association between visitor load and BRDC rates from the study data set, caution is warranted because of the possibility that the number of visits in an average month is a proxy for other unmeasured factors.
In the present study, the multivariable model revealed that breed management may play a role in preweaning BRDC rates in US cow-calf operations. Operations that had calves that were 2- or 3-breed crosses or composites had higher BRDC rates, compared with rates for operations that had single-breed calves. This finding is in contrast with results of a study7 in which BRDC incidence was greater among single-breed preweaned calves, compared with incidences among 2-way cross, 3-way cross, or composite-herd calves. However, that study7 was completed on a large US federal research operation with similar management practices for both single- and multiple-breed groups. It is possible the single-breed operations in the present study were primarily purebred operations with enhanced disease prevention or that more intensive treatment regimens were practiced given the perceived added value of purebred calves. Additionally, in the present study, a definition for crossbred or composite was not provided; therefore, there may have been a difference in the breed composition between this study population and the study by Snowder et al.7 Because of the national scope of the present study, it was unlikely that the genetic traits of the single-breed herds were similar. It is more likely that other management practices that can affect BRDC rates, and differ among single- or multiple-breed herds, were present.
Operations were classified by respondents as a primary or supplemental source of income in the present study, and those classified as a supplemental source of income had lower BRDC rates. Operators who have herds that are primary income sources derive most of their income from the sale of calves.26 Because of the potential added importance of calf sales in these herds, it is possible that recognition of BRDC-related illnesses was heightened and personnel of primary income source operations were more adept at detecting BRDC-affected calves, resulting in more preweaned calves being treated for BRDC. Among the operations included in the present study, being a primary income source was also associated with greater herd size (data not shown; a previous report15 from phase 1 of the National Animal Health Monitoring System Beef 2007–2008 survey indicated distributions of herd size and income categories). Increased herd size may be associated with greater within-herd exposure to BRDC pathogens, which is considered to be a risk factor for preweaning BRDC.23 Additionally, a lack of quantity or quality of skilled employees has been associated with increased preweaning BRDC rates,23 and labor deficiencies could be more common in operations that are primary income sources. Because primary income source herds may have different pathogen exposure levels or disease recognition and control strategies and are likely managed much differently, compared with herds considered as supplemental income sources or other (hobby) classifications, income category may be a proxy for relationships among several risk factors that affect BRDC rates.
Although the design and implementation of the present study should have provided data that effectively represented a large proportion of US cow-calf herds, the external and internal validity of the study may be affected by study limitations. Because the study design was complex, a weighted analysis of the sample population was used to provide representation of the target population.19,27 We recognize that when nonresponse weighting is used, the assumptions required, if incorrect, may result in biased estimates of effects. Potential bias due to weighting was assessed in the present study by comparing population estimates (total number of cattle and number of operations within strata) with totals published by the National Agricultural Statistics Service. If the effects are unbiased, a weighted analysis allows for better external validity, and through the adjustment of response bias, more accurately reflects the cow-calf operator population. The survey participants volunteered to participate in the study; therefore, a self-selection bias may have been present. In evaluating the final model, we compared it with an unweighted model (data not shown) and found relatively similar estimates of effects. Internal validity may have been affected by the accuracy of the statistical models. Unfortunately, limitations in available survey analysis software for weighted analyses prevent extensive evaluations of model fit.19 However, it was clearly demonstrated that a negative binomial distribution was more appropriate for the study data than was a Poisson distribution. For these national study data, there was no uniform case definition for BRDC; instead, the number of calves treated with antimicrobials for respiratory tract disease was reported. Because the study population included many cow-calf operations throughout the United States, the case definition and disease recognition skills among participants likely differed. The potential misclassification of BRDC in the present study may or may not have differed among herds with respect to risk factors of interest. In addition, because of the retrospective nature of the study (operators were asked to recall events that occurred during the preceding year), recall bias may have affected both the reported number of BRDC events and the management practices that were present. Most of the management practices retained in the final model were not likely affected by recall bias, but several management practices in analyses of unconditional associations may have been. Similarly, because the days at risk were calculated considering deaths that occurred within specific time frames, some timing misspecification may have occurred. Combined with potential inaccuracies in the reported mean weaning days, these data may have led to differential or nondifferential misspecification of the rate exposure (calf-days at risk).
In the present study, analysis of the data revealed that several management practices of US cow-calf operations were associated with herd-level preweaning BRDC rates. Because of the cross-sectional nature of the collected data, it cannot be determined whether any of the associated management practices preceded or followed BRDC events. However, enhancing our understanding of these risk factors may help in identifying future disease mitigation opportunities and provides useful profiles of the US cow-calf herd attributes that are associated with increased rates of BRDC in preweaned calves.
ABBREVIATIONS
BRDC | Bovine respiratory disease complex |
BRSV | Bovine respiratory syncytial virus |
BVDV | Bovine viral diarrhea virus |
CI | Confidence interval |
IBR | Infectious bovine rhinotracheitis |
IRR | Incidence rate ratio |
svy, Stata, StataCorp LP, College Station, Tex.
References
1. Duff GC, Galyean ML. Board invited review: recent advances in management of highly stressed, newly received feedlot cattle. J Anim Sci 2007; 85: 823–840.
2. Snowder GD. Bovine respiratory disease in feedlot cattle: environmental, genetic, and economic factors. J Anim Sci 2006; 84: 1999–2008.
3. Fulton R. Bovine respiratory disease research (1983–2009). Anim Health Res Rev 2009; 10: 131–139.
4. Smith RA. Therapeutic management of bovine respiratory disease complex. In: Bovine respiratory disease: a sourcebook for the veterinary professional. Trenton, NJ: Veterinary Learning Systems, 1996.
5. Edwards A. Respiratory diseases of feedlot cattle in central USA. Bovine Pract 1996; 30: 5–7.
6. USDA APHIS Veterinary Services. NAHMS beef 2007–08 part IV: reference of beef-cow calf management practices in the United States. 2009; 10147.
7. Snowder GD, Van Vleck LD, Cundiff LV, et al. Influence of breed, heterozygosity, and disease incidence on estimates of variance components of respiratory disease in preweaned beef calves. J Anim Sci 2005; 83: 1247–1261.
8. Muggli-Cockett NE. Genetic analysis of bovine respiratory disease in beef calves during the first year of life. J Anim Sci 1992; 70: 2013–2019.
9. Wittum TE, Woollen NE, Perino LJ, et al. Relationships among treatment for respiratory tract disease, pulmonary lesions evident at slaughter, and rate of weight gain in feedlot cattle. J Am Vet Med Assoc 1996; 209: 814–818.
10. Van Donkersgoed J. Epidemiological study of enzootic pneumonia in dairy calves in Saskatchewan. Can J Vet Res 1993; 57: 247–254.
11. Svensson C, Lundborg K, Emanuelson U, et al. Morbidity in Swedish dairy calves from birth to 90 days of age and individual calf-level risk factors for infectious diseases. Prev Vet Med 2003; 58: 179–197.
12. Svensson C, Hultgren J, Oltenacu PA. Morbidity in 3–7-month-old dairy calves in south-western Sweden, and risk factors for diarrhoea and respiratory disease. Prev Vet Med 2006; 74: 162–179.
13. Lago A, McGuirk SM, Bennett TB, et al. Calf respiratory disease and pen microenvironments in naturally ventilated calf barns in winter. J Dairy Sci 2006; 89: 4014–4025.
14. Sanderson MW, Dargatz DA, Wagner B. Risk factors for initial respiratory disease in United States' feedlots based on producer-collected daily morbidity counts. Can Vet J 2008; 49: 373–378.
15. USDA APHIS Veterinary Services. NAHMS beef 2007–08 part I: reference of beef-cow calf management practices in the United States, 2007–2008. 2008; 10147.
16. USDA APHIS Veterinary Services. NAHMS beef 2007–08 part II: reference of beef-cow calf management practices in the United States, 2007–2008. 2009; 10147.
17. Dargatz DA, Hill GW. Analysis of survey data. Prev Vet Med 1996; 28: 225–237.
18. Stata Corp. Stata survey statistics reference manual, release 10. College Station, Tex: Stata Press, 2007.
19. Heeringa S, West B, Berglund P. Chapter 3.6: foundations and techniques for design-based estimation and inference. In: Applied survey data analysis. Boca Raton, Fla: Chapman & Hall/CRC, 2010; 68–74.
20. Dewell R, Hungerford LL, Keen JE, et al. Association of neonatal serum immunoglobulin G1 concentration with health and performance in beef calves. J Am Vet Med Assoc 2006; 228: 914–921.
21. Chenoweth PJ, Sanderson MW. Health and production management in beef cattle breeding herds. In: Radostits O, ed. Herd health: food animal production medicine. 3rd ed. Philadelphia: WB Saunders Co, 2001; 509–580.
22. Wittum TE, Salman MD, King ME, et al. Individual animal and maternal risk factors for morbidity and mortality of neonatal beef calves in Colorado. Prev Vet Med 1994; 19: 1–13.
23. Stokka GL. Prevention of respiratory disease in cow/calf operations. Vet Clin North Am Food Anim Pract 2010; 26: 229–241.
24. Edwards TA. Control methods for bovine respiratory disease in feedlot cattle. Vet Clin North Am Food Anim Pract 2010; 26: 273–284.
25. Callan RJ, Garry FB. Biosecurity and bovine respiratory disease. Vet Clin North Am Food Anim Pract 2002; 18: 57–77.
26. Wikse SE, Kinsel ML, Field RW, et al. An epidemiologic approach to solving beef herd production and disease problems. Vet Med 1994; 87: 495–506.
27. Levy PS, Lemeshow S. Sampling of populations. 4th ed. Hoboken, NJ: John Wiley & Sons, 2008.