Diagnostic Stewardship—Leveraging the Laboratory to Improve Antimicrobial Use

Antimicrobial stewardship programs have emerged as a means to address inappropriate antimicrobial use, manage costs, decrease drug resistance, and prevent medication-related adverse events. The traditional stewardship model relies on pharmacists, infectious disease physicians, or both, providing feedback to clinicians. Culture-based and non–culture-based diagnostic tests help establish the presence or absence of infection. Although routine, the process of ordering and interpreting diagnostic tests is complex and frequently results in diagnostic error.1 The decision to order a test should be guided by careful clinical evaluation, recognition of a clinical syndrome, and estimation of the pretest likelihood of the condition for which the test is obtained. Tests are ordered, specimens collected and processed, and results reported. Clinicians then interpret these results and decide whether to initiate or continue treatment.1 However, clinicians often order common tests for patients without symptoms specific for the disease process (ie, those with a very low pretest likelihood of infection), eg, Clostridium difficile stool testing among patients without diarrhea, or urine cultures among patients without symptoms referable to the urinary tract. When positive test results are obtained in these and other scenarios, unnecessary therapy is often prescribed, even though the results represent false-positive findings or colonization rather than true infection.1,2 The problem with ordering tests in the setting of low pretest likelihood of disease is magnified by the availability of increasingly sensitive molecular tests, many of which are combined into “syndromic” testing panels.3 Some panels detect more than 2 dozen targets simultaneously, and future next-generation sequencing tests will detect the presence of any microbial genetic material.3 The pretest likelihoods of infection attributable to each target in these assays vary substantially, further complicating the interpretation of positive test results and potentially contributing to overtreatment.3