2018 Survey of antimicrobial drug use and stewardship practices in adult cows on California dairies: post-Senate Bill 27

Questionnaire design

A survey instrument was designed to collect information about AMD use in adult cows on California dairies. The questionnaire was pre-tested using in-person interviews with extension and outreach specialists and veterinarians. The questionnaire consisted of 43 questions partitioned into three sections: herd information, health management, AMD use, practices and perspectives. Of the 43 questions, 40 were multiple choice questions with the option to specify any available additional information, three questions required the respondent to fill in the requested information. An optional section was included to allow respondents to provide contact information if they were interested in participating in a continuation study and to provide feedback about the questionnaire. The study was reviewed by the University of California, Davis Institutional Review Board and was granted exemption approval (IRB Number: 1537295-1).

Questionnaire administration and data collection

A list of all 1,282 licensed, grade A dairies in CA in 2017 provided by CDFA served as a sampling frame. A grade A dairy produces milk used in any products intended for consumption in fluid form (also called fluid grade milk or market milk). A confidential identification number was randomly assigned to all dairies on the list. Each dairy was uniquely identified by this number during analysis to maintain confidentiality. Dairies were mailed a survey packet containing a copy of the questionnaire, a postage-paid addressed business reply envelope, and an information cover letter. The cover letter introduced the dairy producers to the survey and provided an explanation that the term antimicrobial included antibiotics, drugs that are naturally produced by other microorganisms and that can kill or inhibit the growth of other microorganisms, and also synthetic chemicals such as sulfonamides. Producers were instructed that for the purpose of this survey questions will refer to all AMDs as antibiotics regardless of their origin. The estimated time for producers to complete the survey was 30 minutes. To improve questionnaire response rate, each dairy was mailed a reminder card two weeks after the initial questionnaire package. A second and third copy of the questionnaire, each followed by a reminder card, were similarly mailed a month and six months later, respectively, to dairies that had not previously responded. All mailings were completed between April and September 2018.

Questionnaire sections

Statistical analysis

Exploratory data analysis

Data on the counties where the dairies were located were categorized into three regions (Fig. 1)—Northern California (NCA), Northern San Joaquin Valley (NSJV), and Greater Southern California (GSCA) (Love et al., 2016). For the purposes of descriptive analysis, the herd size and RHA on respondent dairies were categorized based on statewide herd distribution. The statewide mean herd size and RHA was 1,304 milking cows and 23,500 lbs/cow, respectively (California Department of Food and Agriculture (CDFA), 2018). We categorized herd size as <1,305 milking cows (less than the state mean), 1,305 to 3,500 milking cows (larger herds), and >3,500 milking cows to distinguish the largest herds in the state. We categorized RHA as <10,660 kg/cow (<23,500 lbs/cow), less than the state mean; and ≥10,660 kg/cow. BTSCC was categorized as <100,000, 100,000–199,999, and ≥200,000 cells/mL. The <100,000 cells/mL BTSCC represent the high quality milk herd designation commonly associated with bonus payment to such dairies. The 100,000–199,999 cells/mL category represent the industrywide acceptable goal for BTSCC in dairy herds, while the category with ≥200,000 cells/mL represents the high BTSCC herds.

Proportions and their standard error (SE) were computed for categorical and ordinal variables. Mean and SE were computed for continuous variables. Confidence intervals for proportions were calculated using the normal distribution approximation method (Daniel, 2005). Descriptive statistics were performed using Stata 15 (Stata Corp, College Station, TX, USA). Incidence rate for disease treatment was calculated as: number of new cases divided by total cow time at risk, expressed per 100 milking cow months.

Descriptive analyses

A total of 149 (11.6%) survey responses from 19 of the 31 dairy producing counties in CA were received through the three mailings (Fig. 2). Following the distribution of dairy herds statewide (California Department of Food and Agriculture (CDFA), 2018), the greatest proportion of responses came from GSCA, followed by NSJV and NCA. Respondents’ herd characteristics are summarized in Table 1. The overall survey response rate among conventional and certified organic dairies was 11.2% (132/1176) and 15.1% (16/106), respectively. Owners of the dairies constituted the majority of respondents (79%). Additional results are summarized below by management practice, specifically for dry-off practices, vaccination, treatment, drug tracking, drug choices by disease, VCPR, and producers’ perspectives on antimicrobial stewardship.

Stratified analyses

The percentage of respondents that blanket treated cows with IMM AMD and/or teat sealant at dry-off were comparable across herd sizes; however, respondents whose average number of milking cows were 3,500 or more exclusively blanket treated cows at dry-off (Appendix 6). Conversely, the percentages varied by the region where the dairy is located, approximately half of respondents’ blanket treated in NCA, while approximately 93% and 96% blanket treated in NSJV and GSCA, respectively. Approximately 50%, 81%, and 100% of respondents’ with herd sizes <1,305, 1,305–3,500, and >3,500, respectively, tracked AMD treatments given to cows using computer software only or in combination with all or any of markings on the animal, paper records kept in barn or office, and white/chalk board or other temporary marking. Similar patterns were observed across region, approximately 30%, 54%, and 83% in NCA, NSJV, and GSCA, respectively (Appendix 6). The incidence rates of adult cows treatments per month for mastitis, metritis, or lameness increased as the average number of milking cows in the herd increased. On average, 2.9 ± 0.6, 2.9 ± 0.3, and 3.3 ± 1.4 mastitis cases per 100 adult cows were treated monthly in dairies with herd sizes <1,305, 1,305–3,500, and >3,500 adult cows, respectively (Appendix 6). Across region, 0.9 ± 0.3, 2.9 ± 0.6, and 3.5 ± 0.4 mastitis cases per 100 adult cows were treated monthly in dairies situated in NCA, NSJV, and GSCA, respectively. A greater percentage of producers who included a veterinarian in the decision of which oral AMD to purchase had written or computerized animal health protocols for their dairy compared to producers who did not include a veterinarian in the decision, 87% versus 66% (Appendix 6).

Multiple factor analysis and hierarchical clustering

The first two principal component dimensions of the MFA explained approximately 14.02% of variability in the survey responses, 9.59% for the first dimension and 4.43% for the second. Seven components with correlation coefficient ≥ 0.4 were identified and retained (Table 7). On the first dimension, four components accounted for 62.7% of the total variance in the data. These components were disease management practices (mastitis, metritis, lameness, and pneumonia management practices), herd demography, AMD usage information, and cow dry-off protocols. On the second dimension, three components (AMD use stewardship practices, familiarity with FDA “medically important antimicrobial” term, and producer perceptions of AMD on dairies) accounted for 43.1% of the total variance in the data. The cluster analysis of conventional dairy participants partitioned study dairies into two possible clusters (cluster C1 and cluster C2) based on the height of the dendrogram; while the analysis of organic dairy participant grouped into a single cluster.

Conventional dairy clusters

The profiles of the conventional dairy clusters are described in Table 8, while their comparison to the organic dairy cluster is described in Appendix 7. Herd size varied between the clusters with the largest herds in cluster C1. Cluster C1 included dairies from the three regions and all breed types; while C2 included dairies in NSJV and GSCA region, mainly Holstein and herds with more than one breed.

Most of the cluster C1 dairies (95.1%) implemented blanket treatment of dry cows with IMM AMD and teat sealant, compared to dairies in cluster C2 (80.0%). Notably, dairies in cluster C1 used multiple AMD classes (cephalosporins, penicillins, and aminoglycosides) for dry cow treatment, dairies in cluster C2 used mainly cephalosporins. Approximately two thirds of the dairies in cluster C1 tracked both AMD treatments given to cows (69.7%) and withdrawal periods for AMD treated cows (66.4%) using a computer, in addition to other methods. The reverse was true for dairies in cluster C2, approximately two fifths (60.0%) do not track AMD treatments and two thirds (77.8%) do not track AMD withdrawal periods using a computer.

For mastitis treatment, dairies in cluster C1 relied mostly on clinical signs and laboratory testing, and tended to treat cows while culture results, if any, were pending. Dairies in cluster C2 relied mostly on clinical signs. Most C1 dairies (98.4%) treat mastitis with AMD, compared to 75.0% in C2. For metritis treatment, dairies in cluster C2 totally relied on oral/injectable AMD, while cluster C1 relied on both oral/injectable AMD and intrauterine treatment. Similarly, dairies in cluster C2 managed lameness in adult cattle using mostly hoof treatment and less of oral/injectable AMD compared to dairies in C1.

Most dairies in cluster C1 (97.5%) strongly agreed that administration of appropriate AMD, dose, route and duration are important AMD stewardship practices compared to less in C2 (80.0%). However, responding to questions relating to AMD resistance, more dairies in cluster C2 (40.0%) strongly agreed with the statement “Any use of antibiotics may result in infections more difficult to treat in the future”, whereas, fewer dairies in cluster C1 (11.6%) agreed with same statement.

Organic dairy cluster

All the survey’s organic dairies formed a single cluster characterized by multiple breeds with Holstein and Jersey making up 37.5% and 12.5%, respectively and the remaining majority being other breeds. The dairies in the cluster were predominantly in NCA (81.2%) with three respondent in GSCA (18.8%), had a median herd size of 325 milking cows, and a median annual rolling herd average of 8,235 kg per cow. Two organic dairy (12.5%) reported use of blanket treatment of dry-cows depending on the season with an external teat sealant, one organic dairy (6.3%) reported use of blanket treatment of dry-cows with organic intramammary infusion (Phyto-Mast, Bovin-ity Health LLC, Narvon, PA). One organic dairy (6.3%) reported use of selective dry-cow treatment with an organic intramammary infusion (Phyto-Mast), and one other dairy reported use of selective dry-cow treatment with external teat sealant depending on season and mastitis history. Approximately two thirds (60.0%) of the organic dairies reported use of abnormal milk and sample testing at a milk quality laboratory as basis for mastitis treatment decisions, the remaining relied on abnormal milk findings only. Of the organic dairy respondents, 18.8% indicated treatment of mastitis with AMD, while 50% reported treatment of metritis with intrauterine bolus, and 100% reported treatment of lameness with bolus or injection of AMD and hoof treatment. Few organic dairies in the cluster used computers to track AMD treatments (41.7%) and withdrawal periods (37.5%). Three fifths (60.0%) of the survey’s organic dairies indicated use of OTC AMD, or OTC and prescription AMD prior to January 1st 2018. The remaining organic dairy responses (40.0%) indicated no OTC AMD, prescription but no OTC AMD, neither OTC nor prescription, or no response.

Most of the organic respondents (76.9%) identified the following AMD stewardship practices as very important: administration of the appropriate dose, route and duration of AMD; good record keeping of treatments and treatment dates; and observing withdrawal periods and drug residue avoidance. However, on a statement that any use of AMD may result in infections more difficult to treat in future, 38.5% of the survey’s organic dairies agreed or strongly agreed, 38.5% disagreed or strongly disagreed, while 23.0% were neutral.

Dry cow treatment practices (dry-off protocols)

Our findings indicated that the majority of surveyed dairies (93.7%) treated all end-of-lactation cows presenting for dry-off with IMM AMD and/or teat sealants. About half (49.6%) of the respondents used IMM AMD, 44.1% used both IMM AMD and teat sealant, while 6.3% used only teat sealant at dry-off. The USDA-National Animal Health Monitoring System 2014 Dairy Study, (National Animal Health Monitoring System (NAHMS), 2016a), showed that 86.1% of dairy operations in the Western US (California, Colorado, Idaho, Texas, and Washington), and 79.7% in the Eastern US (Indiana, Iowa, Kentucky, Michigan, Minnesota, Missouri, New York, Ohio, Pennsylvania, Vermont, Virginia, and Wisconsin) treated all their cows at dry-off with IMM AMD, while 90.8% of the operations used some antimicrobials on at least some of the cows at dry-off. In contrast, 7.2% and 9.4% of the operations in the Western and Eastern US, respectively, did not use IMM AMD at dry-off. According to our survey, 14.7% of respondent dairies carried out selective dry cow treatment at dry-off; this estimate was close to the 13.9% and 20.3% reported by NAHMS for Western and Eastern US, respectively. The majority of the study dairy operations (61.6%) used the cephalosporins (Spectramast^®, Tomorrow^®) class of AMD in dry cow treatment, 33.6% used AMD containing penicillins (Orbenin^®, Boviclox^®, Albadry^®), while 14.4% used penicillin-aminoglycoside combinations (Quartermaster^®). Dry cow treatment statistics based on the NAHMS 2014 dairy study showed comparable estimate for the use of cephalosporins (63.3%) on Western US dairy operations (29% used Cefa-Dri^®/Tomorrow (cephapirin: first-generation cephalosporins), while 34.3% used Spectramast^® (ceftiofur: third-generation cephalosporins). The contrast was seen with use of AMD containing penicillins on dry-off treatment which was reportedly greater in our study 33.6% compared to the 26.8% (Boviclox^®, Dry-clox^®, Orbenin^®, Albadry^® plus) in NAHMS 2014 Dairy study; and with Quartermaster^® which was less in our study (14.4% compared to 24.0% in NAHMS 2014 Dairy study) (National Animal Health Monitoring System (NAHMS), 2016a). Several studies have described the benefits of dry cow therapy to udder health (Nickerson & Oliver, 2014; Owens et al., 2001; Vilar et al., 2018). The practice prevents mastitis in the early dry period, allows milk-producing tissue to redevelop in cured quarters, and reduces clinical mastitis at freshening. Blanket dry cow therapy has been used to treat cases of chronic mastitis caused by Staphylococcus aureus and Streptococcus agalactiae and to reduce new IMM infections early in the dry period (Bonsaglia et al., 2017). However, blanket treatment of all dry cows exposes cows without a history of mastitis to unnecessary antimicrobial use (Bonsaglia et al., 2017; Higgins et al., 2017). Recent studies have demonstrated the benefits of selective dry cow therapy as an effective alternative to blanket treatment with AMD (Bonsaglia et al., 2017; Cameron et al., 2015, 2014; Scherpenzeel et al., 2014). Some countries such as Finland, the Netherlands, and Denmark have banned blanket dry cow treatment (Bennedsgaard, Klaas & Vaarst, 2010; Scherpenzeel et al., 2016; Vilar et al., 2018). Selective dry cow treatment has been shown to result in: decreased use of antimicrobials for IMM infections during the dry period by as much as 21%, resulting in reduced incidence of AMR, and thus translating to a significant reduction in production cost (Cameron et al., 2014; Østerås, Edge & Martin, 1999). However, determination of an effective tool to identify cows that will benefit from selective dry cow treatment could be challenging.

Dairy cow health management and vaccination practices

Most dairy producers in the survey vaccinated lactating cows to prevent mastitis due to coliform bacteria (85.6%), respiratory disease (87.0%), and abortion and infertility (73.0%). Furthermore, 95.3% of the study dairies administered at least one type of vaccine to lactating cows for disease prevention. This percentage is higher than the 73.8% of dairy operations that administered any vaccine to cows for disease prevention reported by National Animal Health Monitoring System (NAHMS), (2016b), including, 18.1% of U.S. operations that vaccinated lactating cows against coliform mastitis (range, 2.4–50.8% depending on herd size) (National Animal Health Monitoring System (NAHMS), 2016a). Generally, NAHMS reported that more dairy operations in the western U.S. (86.3%) vaccinated their lactating cows for any disease compared to dairies in eastern U.S (72.6%) (National Animal Health Monitoring System (NAHMS), 2016b). Several other studies have highlighted the benefits of vaccination in preventing or reducing the severity of several infections in vaccinated cows (Ferguson, Galligan & Cortese, 1997; González et al., 1989; Gorden et al., 2018; Pereira et al., 2013; Prado et al., 2011) and their offspring (Dubrovsky et al., 2019a, 2019b; Maier et al., 2019). Scientific efforts should be expanded towards the development of alternative non-AMD disease treatment and prevention approaches geared towards reducing the incidence of AMR in livestock production.

Dairy cow health protocols and antimicrobial treatment practices

Overall, 92.3% of the study dairy producers relied on a veterinarian only or veterinarian and other sources for information regarding AMD to use for treatment of adult cows. Half (49.3%) of the dairies included a veterinarian in their decision on which oral and injectable AMD are purchased and stocked, while 37.9% included a veterinarian in their decision on which AMD is used to treat sick cows. These proportions are consistent with the National Animal Health Monitoring System (NAHMS) (2018) survey which confirmed that 96.2% of U.S. dairy producers consulted a veterinarian or a drug label created by a veterinarian in their treatment decisions, 55.1% consulted with a veterinarian or the drug label created by the veterinarian for their decision on what drug to use while 44.0% relied on previous experience with the drug. From a 2002 survey of antibiotic usage on 113 dairy herds from 13 counties in Pennsylvania (Sawant, Sordillo & Jayarao, 2005), only 32% of dairy producers sought veterinarian advice before administering antibiotics to cows on their dairies, a proportion similar to what was observed in our survey. The above data suggests the need for increased veterinarian involvement with producers in their treatment decisions, in line with state regulation (FAC 14400–14408 (formerly SB-27)) which require that all MIADs, other than those intended to be fed to livestock (which require a VFD), may only be purchased and administered with a prescription from a California licensed veterinarian within a valid VCPR.

In our study, 75.0% of dairies confirmed having written/computerized animal health protocols for cows, mainly developed with veterinarian inputs for both treatment and preventative therapeutic uses. In a 2007 study aimed at determining South Carolina (SC) dairy farmers’ knowledge on the importance of prudent antibiotic use on dairy cows, Friedman et al. (2007) reported that only 32% of participating farms had written protocols for diagnosing and treating common medical conditions in the dairy herd. Sawant, Sordillo & Jayarao (2005) reported only 21% of dairy producers had written plans for treating sick animals. In the SC study, most farmers expressed that though their protocols may have been ‘unwritten’, they followed standard operating procedures in dairy herd management. In a survey of dairy herds in Washington State and Pennsylvania, Raymond, Wohrle & Call (2006) and Sawant, Sordillo & Jayarao (2005) concluded that the lack of written protocols for disease treatment on dairies could lead to inappropriate use of antibiotics which can promote the selection of resistant bacteria. Although, a high percentage (75.0%) of farmers in the current CA study confirmed having written/computerized animal health protocols compared to the Washington and Pennsylvania’s studies, research is needed to investigate this association further.

Antimicrobial drug selection, dosing and tracking practices

Drug inventory and treatment records help dairy producers track drug use and treated cows. Approximately half of the surveyed dairies reported not keeping a drug inventory log on their dairies, information which otherwise may help the producers monitor drug use on the dairy over time, track availability and stocking of drugs to avoid shortages or waste due to their expiration, and prevent accidental administration of expired drugs. A survey of dairies in Pennsylvania found only 50% of the dairies kept written or computerized records of antibiotic treatment that could be verified (Sawant, Sordillo & Jayarao, 2005). Other important benefits of a drug inventory to producers include the opportunity to monitor drug label changes over time and monitor costs incurred for AMDs over time and by drug class. Similarly, maintaining adequate treatment records benefits food safety, supports improved management of drugs used that could help prevent drug residues in milk and meat, helps ensure an effective herd health plan, and reduces farmer liability or violations during regulatory follow-up.

At least 36.8% of our survey respondents relied on bacterial culture and antibiotic sensitivity results in choosing a second antibiotic drug to treat a sick animal if the first was not successful. In addition to consulting with a veterinarian, laboratory testing may be a better approach after a treatment failure to confirm the bacterial etiology or its antimicrobial sensitivity if the diagnosis was correct.

Most (80.3%) of the dairies in our survey confirmed tracking/recording milk, meat, or milk and meat withdrawal intervals. According to NAHMS, more than 90.0% of U.S. dairy operations administered drugs that required a milk withdrawal period or a milk and meat withdrawal period (National Animal Health Monitoring System (NAHMS), 2016a). The National Milk Producers Federation (NMPF) recommends that producers keep written or electronic records on all cattle treated with drugs that require a withdrawal period for milk or meat, records should be maintained for at least two years or as otherwise required by Federal or State law, and should be accessible to everyone who works with the animals (National Milk Producers Federation (NMPF), 2019). Approximately 35% of dairies in our survey tracked milk or meat withdrawal intervals but not both; such a practice is not uncommon, as certain drugs in specific formulations (such as Ceftiofur) and that are used in adult cow treatment may only have meat and no milk withdrawal periods. Similarly, approximately 20% (28/142) of surveyed dairies did not track milk or meat withdrawal periods; 6 (20%) of these dairies were organic while the remaining 80% were conventional dairies. Not tracking milk or meat withdrawal period in this group could be related to not using AMDs on these dairies or may represent cases of drugs used that do not have a milk withdrawal period.

Basis and antimicrobial choices for disease treatment

According to the survey, the number one reason for AMD use on CA dairies was for mastitis treatment, followed by metritis, lameness, pneumonia, and postoperative care; similar to the report by Sawant, Sordillo & Jayarao (2005) and Raymond, Wohrle & Call (2006). Approximately, three mastitis cases per 100 milking cow months were reported based on clinical signs and laboratory diagnosis. Based on NAHMS 2014 survey, 99.7% of dairy producers in 2013 reported having at least one case of mastitis, and one-fourth of all cows on their dairies had clinical mastitis (National Animal Health Monitoring System (NAHMS), 2016a). Two thirds of dairies in our survey individually treat mastitis cows with IMM AMD infusion, while the remaining used IMM infusion and systemic AMD, which is similar to the estimate reported in the NAHMS 2014 dairy study where 89.4 % of dairy operations used IMM AMD to treat mastitis. Cephalosporins remained the primary antimicrobials for treating mastitis, followed by penicillins, lincosamides, and tetracyclines as also evident from the NAHMS 2014 survey.

According to NAHMS 2014 dairy study (National Animal Health Monitoring System (NAHMS), 2018), 58.8%, 61.0%, and 54.1% of dairy operations treated some cows with antimicrobials for reproductive disease, lameness, and respiratory disease, respectively. In our survey, approximately 2 per 100 adult cows were treated for each of metritis and lameness per month, and 0.3 per 100 adult cows were treated for pneumonia. Tetracyclines were the primary antimicrobials used for intrauterine treatment of metritis and cephalosporins for systemic treatment. Lameness was managed in survey dairies with hoof treatment and systemic antibiotics, primarily with tetracyclines and cephalosporins.

Veterinarian-client-patient relationship (VCPR) and disease diagnosis practices

The basic elements of a VCPR are that the veterinarian agreed to assume the responsibility for the animal’s health and the producer agreed to comply with the veterinarian’s instructions; the veterinarian has sufficient knowledge of the animals to make a general diagnosis and assumes responsibility for follow-up care, including adverse reactions and/or treatment failures; and the veterinarian maintains all animal records (16 CCR 2032.1). Most dairies (92%) in our survey confirmed having a valid VCPR for their dairies. Further analysis of the survey questions revealed that the 8.0% (12/143) of producers who said they do not have a VCPR for their dairies may have either misunderstood the question, were not familiar with the term, or mistakenly selected the incorrect answer, as their responses to other questions relating to cow health on their farms showed direct involvement of a veterinarian in cow health care decisions. Specifically, of the 12 producers who indicated not having a VCPR, 10 included veterinarians as one of the sources they relied on for information about AMD used to treat cows. The remaining two dairies had either submitted non-routine samples for infectious disease diagnosis or had used other on-farm diagnostic techniques to guide treatment decisions with AMD for cows; both diagnostic activities were most likely associated with a veterinarian. Hence, continued education of dairy producers is necessary to update their understanding of the requirements of the current law and what a valid VCPR entails.

Dairy farmer practices and perspectives

Animal welfare audit programs are designed to improve welfare standards in livestock operations by ensuring animals are humanely raised. Dairy quality assurance programs, on the other hand, are designed to improve product quality through assessments and monitoring, thereby promoting consumer health, environmental health, and dairy cattle health and welfare. Most dairies in our survey participated in welfare (97.5%) and quality assurance programs (52.9%) in the previous year, highlighting the dedication of CA producers to maintaining high standards of animal welfare and product quality. Based on the NAHMS 2014 dairy study, approximately 45.9% of all dairy operations nationwide participated in a quality assurance program, with higher participation by operations in Western US (66.8%) compared to in the Eastern US (43.9%) (National Animal Health Monitoring System (NAHMS), 2018). The majority (98.6%) of producers in our survey were aware that all uses of medically important AMD in livestock require a VFD or prescription. Such a high percent shows the successful impact of SB-27 legislation on responding CA dairies and producers, the majority of whom had used OTC AMD on their dairies prior to January 1, 2018 and have made changes, such as, decreasing usage of AMD which were previously available OTC, increased usage of alternatives to AMD, and introduced management changes leading to better herd health as reported by a third of the surveyed dairies.

Multiple factor analysis and hierarchical clustering

The MFA identified seven components that accounted for 99.7% of the variation in the data. The seven components identified the important questions that could be considered for a risk assessment tool for AMD use and AMR on CA dairies in future studies, similar to the approaches adopted for bovine respiratory disease (Love et al., 2016; Maier et al., 2019) and Johne’s disease (Berghaus et al., 2005) risk assessment.

Cluster C1 consisted of large-scale conventional dairies (median herd size: 1,265 cows; median RHA: 11,340 kg/cow) situated mainly in the GSCA and NSJV region. Dairies in this cluster were characterized by their AMD usage in dry cow management (blanket treatment of all dry cows) at the end of each lactation, and in the management of disease conditions such as mastitis, metritis, and lameness in adult cows. The cluster can be linked with good knowledge of AMD stewardship practices as evident by the high frequency of AMD treatment and withdrawal tracking, most often using a computer, and overall agreement to statements regarding drug use stewardship practices and AMR in their dairy operations.

Cluster C2 consisted of mid-sized conventional dairies (median herd size: 715 cows; median RHA: 10,092 kg/cow) located in NSJV and GSCA regions. Producers in this cluster could be linked with some knowledge of AMD use stewardship practices as characterized by their high level of agreement (100%) to statements regarding AMD use stewardship practices. They tend to show more agreement regarding AMR issues raised in the study compared to dairies in clusters C1. Producers in this cluster could benefit from outreach and education relating to use of computer in record keeping, AMD use stewardship practices and their impact on AMR in dairy operations.

The organic cluster consisted of small-sized dairies mainly located in NCA and reported little or no usage of AMD in their dairy operations. Organic dairies do not use AMD amongst other prohibited chemicals (such as pesticides, hormones with exception of oxytocin) unless sick animals require antibiotics. Emergency AMD treatments to maintain the welfare of the sick animals are allowed on organic dairies; however, such animals have to be segregated from the organic herd and sold to a non-organic market in addition to strict documentation and reporting requirements (Code of Federal Regulations (CFR), 2011). The survey’s organic dairies had mixed responses on AMD stewardship practices with approximately two-fifths agreeing that any use of AMD may result in infections more difficult to treat in the future.

The current study data were collected based on voluntary participation, in accordance with the SB 27 from CA licensed grade A dairy producers. The mean dairy herd size in our survey was 1,550 cows per herd, which was greater than a 2017 estimate of average herd size of 1,304 cows per herd (California Department of Food and Agriculture (CDFA), 2018); which may suggest over representation of large herds in our survey. However, response rates across regions in our survey were similar to the regional distribution of herds published in the California Department of Food and Agriculture (CDFA) (2018) report. Therefore, potential bias due to regional difference was minimal as all 3 regions were proportionately represented in our survey. In addition, survey responses may reflect respondent recall and/or reporting bias and may not be representative of the actual practices on their dairies. Furthermore, AMD usage and treatment practices in respondent dairies represented the period of time shortly after the implementation of SB 27 on January 2018 and so may not represent changes that have taken place since the survey was administered. Hence, a follow-up survey is necessary to capture the long-term impact of possible changes in stewardship practices among CA dairy producers. Additional studies are needed to measure factors associated with herd demography, drug use treatment practices, and changes in management and stewardship practices since SB-27. Further studies correlating AMR profile of dairies with levels of AMD usage on the dairies will be beneficial to the understanding of AMD use and the potential for AMR development on CA dairies.