The implementation and use of molecular methods is usually more expensive when compared to the classic microbiological diagnostic tests, since they usually require a relatively sophisticated laboratory infrastructure and equipments, and reagents and consumables are also typically more costly. In addition, the procedures may be more demanding and/or have increased complexity, in particular to prevent the occurrence of cross-contamination that may lead to false-positive results, which in turn require more specialized technical personnel. However, the simplification and increasing automation of various procedures have been contributing to a greater integration of molecular diagnostics methods in clinical laboratories. With the currently available technology, it is now possible to process and extract nucleic acids from hundreds of samples and proceed to analysis for defined genetic markers in less than a working day. Thus, the ability of molecular-based methods to analyze hundreds or thousands of samples in a short time, and a greater speed to reach the final results (also potentially contributing to limit the use of unnecessary empirical antimicrobial therapies), may generate substantial savings, making these technologies more competitive in terms of cost per analysis. An earlier detection of the infectious agents in diseased or carrier animals may also limit the spread of these pathogens to neighbor animals (and even to humans) and improve the overall quality of the veterinary service.

Another important aspect is related to the sensitivity of molecular methods. It is widely recognized that the analytical sensitivity of these methods is extremely high, particularly in those based on the amplification of nucleic acids, often being possible to detect the presence of a single copy of the target nucleic acid in the reaction mixture. In contrast, the performance of molecular methods may become unsatisfactory for the detection of nucleic acids directly from samples, yielding false-negative results. A problem still largely unsolved concerns the processes of extraction and purification of nucleic acids from biological samples (e.g., blood or tissues). The extraction and purification of microbial nucleic acids from some biological matrices prove to be complicated, due to the intrinsic characteristics of the agents (e.g., difficulty to lyse cell walls, such as in bacteria from the Mycobacterium tuberculosis complex), due to the presence of a much higher proportion of nucleic acids from host animals, or related to the presence of inhibitory compounds in the nucleic acid extracts that may affect the success of the amplification reactions [16]. Protocols for the extraction of nucleic acids, including several steps of mechanical, chemical, or enzymatic lysis, and several purification steps can be efficiently used in research, but are not practical or are too expensive for routine use. Thus, the availability of efficient, consistent, and reproducible nucleic acid extraction processes adapted for difficult biological matrices, yet affordable and simple to use, is still an unmet need that has hindered the widespread routine use of molecular methods in the veterinary microbiology laboratory. The consistency of the extraction process becomes even more critical if the aim involves the quantification of the pathogenic agents in the samples.

Another major challenge of molecular diagnostics lies in the lack of harmonization of the operational procedures used among different laboratories, which contrasts greatly with the situation of conventional diagnostic methods, whose use is very standardized. In veterinary medicine, molecular tests are largely designed in-house and implemented locally in different laboratories, as there are fewer commercial diagnostic kits available (also due to the smaller IVD market). Thus, to detect the same agent in the same type of sample, laboratories may use different methods for the extraction of nucleic acids and/or discrepant primer sequences and/or diverse amplification conditions. Consequently, the comparison of results among different laboratories is complicated and may even be contradictory, an illustrative example being the discrepant results for the PCR-based diagnosis of feline immunodeficiency virus (FIV) infection in cats by different laboratories using parallel samples [31]. The level of optimization and exigency in the implementation of molecular diagnostics methods varies greatly from laboratory to laboratory, from those that simply reproduce a particular system that has been described elsewhere in the literature, offering little quality control to those who effectively conduct a technical validation and quality control of the methods, assessing their specificity, sensitivity, and positive and negative predictive values, among other features.

Important in this context is the availability of reference materials and adequate gold standard diagnostic methods to validate and assess the performance of molecular tests. International agencies, such as the OIE, are making efforts to alert for the need for international standardization and validation of molecular tests for the use in veterinary microbiology laboratories and are making available relevant guidelines and procedures toward that aim. Nevertheless, since there is currently no major regulatory validation hurdles for the marketing of veterinary molecular diagnostic tests for infectious diseases, veterinarians and technicians must keep themselves informed and be smart buyers of these tests or services [32].

Perhaps one of the major challenges with regard to the molecular diagnostics technologies is precisely the interpretation of the clinical relevance of a positive or negative test. A negative result, which may indicate the absence of the agent, should always be interpreted considering the limitations of the specificity and sensitivity of the molecular test since, as mentioned above, false-negative results may occur, for example, due to the inefficiency of the nucleic acid extraction methods. With respect to the positive tests, the interpretation could be considered straightforward, for example, in cases where the presence of a particular pathogenic agent implies the occurrence of disease. However, a positive amplification result does not always mean that the agent is present in the sample in a viable state. The agent may be dead or nonviable, or there may be only remnants of nucleic acids (e.g., in animals that began therapy or when the agents have been successfully eliminated by the immune system) [33]. An interesting but yet unanswered question for different infectious diseases is how long the nucleic acid of causal agent may remain detectable in the samples after treatment and/or recovery of the animals. In other cases, nucleic acid detection may not mean that the animal is effectively ill, for example, in cases of latent infections, in asymptomatic carrier animals, in hosts in equilibrium with the agent, and when dealing with ubiquitous microorganisms. The vaccination of animals may also interfere with the performance or interpretation of molecular diagnostics assays, the detection of FIV in vaccinated cats being a good illustrative example of this limitation [31]. Moreover, when the disease is caused by opportunistic pathogens belonging to the normal microbial flora of animals, the interpretation of a positive molecular test also becomes more difficult, being necessary to differentiate between colonization and active infection. Veterinary doctors and technicians should be aware of the enhanced accuracy and speed of molecular tests for the diagnosis of infections but also to understand their limitations. Thus, clear algorithms for the use of molecular tests must be settled for each disease (and eventually for each animal species), establishing the appropriate specimen to collect, the suitable timing for collection, and the proper test to use, namely, regarding qualitative versus quantitative response, and whether it is fit for purpose (i.e., fulfills the needs of the situation). The clinician should be aware of the pathogenesis of each pathogen and collect the most adequate sample for its detection since the analysis of the wrong sample type is a common mistake leading to false-negative results [34]. A clear reporting of molecular data by diagnostic laboratories, an effective communication between the laboratory and the clinical settings, and clear interpretative criteria for the results have also to be established, requiring well-designed clinical studies. Molecular results should be analyzed considering the limitations of the techniques, the epidemiological context, and, preferably, where appropriate, in complement to the classical methods in a search for various relevant phenotypic and genotypic characteristics of the agents. Ultimately, the veterinarian should bring together all these results within the context of the set of clinical signs and symptoms exhibited by the suspect animal(s), to reach a final diagnosis.