Five Most Important Mistakes in Diagnostic Test Interpretation for Infectious Diseases
J. Sykes
Department of Medicine & Epidemiology, University of California-Davis, Davis, CA, USA
Over the past 15 years, there has been a focus on development of nucleic acid-detection assays to rapidly detect pathogens, such as polymerase chain reaction (PCR)-based assays and fluorescent in situ hybridization (FISH). Refinement of molecular diagnostic assays has continued to occur in veterinary diagnostic laboratories, with improved standardization and quality assurance. There have also been advances in creation of nucleic acid-based assays that can be used in the clinic. However, as with other assays, it remains important that practitioners understand the limitations of these assays and how to properly interpret test results.
The following pitfalls of diagnostic testing for infectious diseases should be considered:
1. Before selecting and interpreting a diagnostic test, veterinarians should always determine whether a diagnostic test used is an organism-detection test versus an antibody-detection test. The detection of an organism implies the presence of the pathogen itself, whereas the pathogen may no longer be present when antibodies are detectable.
2. Detection of an organism, either by antigen testing, nucleic acid testing, or culture, does not imply that it is the cause of the disease (etiologic predictive value). The presence of an organism may be an incidental finding, especially when subclinical infection or colonization is widespread. In addition, for some infectious diseases, positive test results may occur as a result of recent vaccination.
3. A negative organism-detection test or antibody detection test result should not be construed to ‘rule out’ an infectious disease. Organisms may be present at undetectable levels, the sample size may be too small, or for PCR assays, assay design may limit the detection of certain strains of a pathogen. It is also possible that organisms are not being shed from the specimen collection site. Antibody tests are commonly negative early in the course of acute infectious diseases, or in very immunosuppressed animals.
4. Positive test results in very low prevalence areas must be interpreted carefully because they often represent false positive test results (low positive predictive value). This includes overinterpretation of unsuspected positive test results that appear when an infectious disease ‘panel’ is requested, including pathogens that the veterinarian was not initially suspecting.
5. Veterinarians should consider whether multiple different types of diagnostic tests may be required for a specific infectious disease, which should be interpreted in light of the time course of illness and the clinical findings. In particular, the results of antibody-detection tests may complement the results of organism-detection tests.