On-farm testing and selective treatment of clinical mastitis

Mastitis is the most common reason for administering antibiotics to dairy cows in the UK. On-farm culture (OFC) programmes emerged first as a management tool as dairy farm size increased and following a shift in aetiology towards Gram‑negative (G−) organisms, which may not benefit from antimicrobial therapy.

These approaches may reduce the level of antimicrobial use on some farms, although the most effective route to improving cow welfare, economic benefit and reduction in antimicrobial use is still to concentrate on mastitis prevention as a first priority.

OFC has been used to identify clinical cases of mastitis caused by Gram-positive (G+) pathogens that would benefit from antibiotic treatment. Non-severe cases that are culture negative or caused by G− pathogens with high rates of spontaneous cure are generally managed without use of antimicrobials.

Methods used for OFC are simpler than those used in professional diagnostic laboratories and based on the goal of simply directing therapy; they are not expected to be as accurate (Lago and Godden, 2018). Methods are usually performed by farmers with some technical training, but it is important that they understand the limitations of these techniques.

Most methods used for OFC are laboratory shortcuts that are based on use of selective or chromogenic media. Aseptic milk samples are taken, pre-treatment, from non-severe cases of clinical mastitis by the farmer and plated up using the test kit of choice. After incubation for 12 to 24 hours, a decision support tree guides actions taken: treatment with antibiotics, NSAIDs, no treatment, marking for culling or segregation. Plates exhibiting no growth may be cultured for a further 24 hours.

The goals of OFC are to reduce the proportion of cases that receive antimicrobials, without significantly reducing cow welfare, clinical or bacteriological cure rates. This can prove cost-effective through the reduction in discarded milk, as well as reducing unnecessary antibiotic use and the risk of antibiotic residue failure.

Use of OFC can be particularly useful in aiding decision-making when mastitis detection aids, such as those used in automatic milking systems, have identified preclinical mastitis. OFC systems have also been used for screening herd additions and for the identification of quarters for antibiotic treatment in selective dry cow therapy, where individual cow somatic cell count data is not available.

Scientific evaluation of the concept

The concept of using selective treatment for clinical mastitis, which means treating different bacteria with different antibiotics or none at all, has been practised and researched for many years.

Some organic farms in the US, UK and elsewhere have ceased using antimicrobials in dairy farming completely, but the discussion of the implications of this approach is beyond the scope of this article. We are discussing the choice of whether a case receives antibiotics based on the results of an on‑farm test kit.

The most common approach is to give antimicrobials to G+ cases only, while mild or moderate cases (grade one or two) caused by G− bacteria or showing no growth are not receiving them. Beyond the antimicrobial choice, both types of mastitis are increasingly treated with anti-inflammatory drugs, for which good evidence exists beyond the reduction of pain − for example, lower somatic cell counts and reduced culling rate (McDougall et al, 2009) and improved reproductive performance (McDougall et al, 2016).

Do mild or moderate cases caused by G− organisms need antimicrobials? Schukken et al (2011) compared a five-day course of intramammary ceftiofur, a third-generation cephalosporin, with no antimicrobial treatment in mastitis cases caused by Escherichia coli and Klebsiella, and found significantly higher clinical and bacteriological cure rates, and lower subsequent culling rates in the treated cows.

In contrast, Pinzón-Sánchez et al (2011) developed a decision tree for mastitis treatments based on published literature, emphasising that little difference exists in bacteriological cure rates in G− or no growth cases due to antimicrobial use (Table 1).

Table 1. Bacteriological cure rates of different bacteria and treatment regimes for multiparous cows, taken from the literature by Pinzón-Sánchez et al (2011).
		0 days	5 days
Gram-positive	Staphylococcus aureus	0%	20%
	Environment Streptococcus	25%	65%
	CNS	55%	75%
Gram-negative	Escherichia coli	75%	85%
	Klebsiella	35%	45%
No growth		90%	90%

As a consequence they see OFC as cost effective on farms using routine extended treatment protocols.

Lago et al (2011) carried out a field study on eight US dairies, comparing treatment after OFC with blanket treatment, and found no significant differences in cure rates, somatic cell counts, recurrence rates, culling and so on.

However, an 11% reduction occurred in bacteriological cure rate in the OFC group, and the given samples size would have required a 14% difference to show significance, so may have been underpowered, as pointed out by Down et al (2017).

Other field studies found no differences in any of the parameters tested. Vasquez et al (2017) compared blanket treatment of clinical mastitis with five days of intramammary ceftiofur with culture-based treatment (G+ cases treated only with one day of intramammary cefapirin) and found no difference in days to clinical cure, subsequent yields, and somatic cell counts and culling. Antimicrobial usage was reduced by 32% in the culture-based cows. These outcomes contradict the earlier work by Schukken et al (2011).

McDougall et al (2018) looked purely at re-treatment rates, and found no difference between blanket and selective treatments.

Fuenzalida and Ruegg (2019) compared intramammary ceftiofur for zero days (control), two days and eight days for the treatment of G− mastitis, and found differences in bacteriological cure for Klebsiella, but not for E coli.

In summary, Ruegg (2018) estimated that only 20% to 33% of antimicrobial treatments for mastitis in the US are beneficial and recommended a targeted, ideally culture-based approach.

It is beyond the scope of this article to compare the different test kits available, but peer-reviewed evaluations have been carried out for MastDecide (Leimbach and Krömker, 2018), Vetorapid (Viora et al, 2014) and Accumast (Ferreira et al, 2018), among others, with sensitivity and specificity figures given.

Sipka et al (2021) pointed out the importance of training, observing higher agreement rates of test kits with standard laboratory practices in trained staff.

Schmenger et al (2020) evaluated a targeted mastitis treatment protocol in a field study on five German farms. This protocol not only withholds antimicrobials for cases likely to cure spontaneously, but also to cases considered as “not worthy”, due to somatic cell count and clinical mastitis history. They found a significant reduction in antimicrobial use of more than 50% in the targeted treated groups with no difference in bacteriological cure rates, in spite of reported compliance issues on the farms.

Blueprint for success

For success:

• The farmer needs to be trained to take a milk sample aseptically, inoculate and incubate the media correctly, and understand the health and safety risks. On occasion these systems can be operated by the laboratory or veterinary practice, using milk samples provided by the farmer.
• Veterinary involvement is vital to oversee correct technique and decision-making, as well as initially ensuring appropriate mastitis epidemiology.
• The OFC kit needs to perform to a sufficiently high sensitivity and specificity (Royster et al, 2014; Ferreira et al, 2018; Malcata et al, 2019; Viora et al, 2014; Leimbach and Kromker, 2018; Mansion-de Vries et al, 2014). If the sensitivity is too low for detecting G+ organisms, an important opportunity for treatment will be lost. If the specificity is too low then too many mastitis cases will receive unnecessary antibiotic treatment. The most important predictor of success is a high negative predictive value  for G+ organisms, which could be attained through a combination of high sensitivity for – and a low prevalence of – G+ pathogens.
• Farm staff must be motivated, and able to follow protocols and make correct treatment decisions.
• Quality control measures to monitor both diagnostic accuracy and contamination. It is recommended that pre-treatment samples are frozen so as to be available for confirmatory culture.
• Ongoing monitoring by the veterinary surgeon should include results of confirmatory bacteriology, cure rates, antibiotic use, recurrence rates, culling rates and proportions of G+/G-/no growth/contamination.

What can go wrong?

The following points are what can go wrong and how to prevent this:

• OFC must be carried out correctly to avoid zoonotic risks to human health.
• Incubator function needs to be monitored to ensure the correct temperature.
• Poor sampling or plating up technique will result in media contamination and it is important that training incorporates recognition of contaminated cultures.
• Test reading errors can occur due to under-diagnosis; particularly missing slow-growing or pinprick colonies, or where the incubation temperature is too low. Overdiagnosis can result from the inability to recognise contamination (three or more separate isolates), single colonies (particularly not in association with the inoculate) or misidentification of milk globules.
• A high prevalence of G+ organisms can result in most cases being treated, which may remove any cost benefit. Also, the delay may result in a reduction in cure rates, although this has not been demonstrated experimentally. It has been estimated that these systems are only cost-effective when the prevalence of G+ mastitis pathogens falls below 50%, but this assumes a reduction in cure rate of 5% (Down et al, 2017).
• Where poor compliance or decision-making cannot be rectified by further training, it is suggested that OFC is discontinued.