Supplemental Digital Content is available in the text

The determination of herpes simplex virus (HSV) infection using a PCR assay is one of the most commonly requested tests for analysis of cerebrospinal fluid (CSF), although only a very low proportion of results are positive. A previously reported study showed that selecting only those CSF samples with >5 leukocytes/mm 3 or a protein level of >50 mg/dl was adequate for the diagnostic workup. The aim of the present study was to assess the reliability of alternative acceptance criteria based on elevated CSF white blood cell counts (>10 cells/mm 3 ). We analyzed all requests for HSV PCR received between January 2008 and December 2011. CSF samples were accepted for analysis if they had >10 cells/mm 3 or if the sample was from an immunocompromised patient or a child aged <2 years. In order to evaluate our selection criteria, we identified those CSF samples with a leukocyte count of 5 to 10 cells/mm 3 or protein levels of >50 mg/dl in order to test them for HSV type 1 and 2 (HSV-1 and HSV-2) DNA. During the study period, 466 CSF samples were submitted to the microbiology laboratory for HSV PCR. Of these, 268 (57.5%) were rejected, and 198 (42.5%) were tested according to our routine criteria. Of the tested samples, 11 (5.5%) were positive for HSV DNA (7 for HSV-1 and 4 for HSV-2). Of the 268 rejected specimens, 74 met the criteria of >5 cells/mm 3 and/or protein levels of >50 mg/dl. Of these, 70 (94.6%) were available for analysis. None of the samples yielded a positive HSV PCR result. Acceptance criteria based on CSF leukocyte counts, host immune status, and age can help to streamline the application of HSV PCR without reducing sensitivity. Detection of herpes simplex virus (HSV) DNA in cerebrospinal fluid (CSF) using PCR assay has been validated for the diagnosis of central nervous system (CNS) herpes infections. HSV PCR is currently recognized as the reference method (1, 2). This sensitive but expensive test is included in the routine evaluation of many patients with suspected CNS infection, although most of the tests performed yield negative results (3). Therefore, screening for suitable diagnostic samples is necessary.Acceptance criteria based on elevated CSF leukocyte counts (Ͼ5 cells/mm 3 ) and protein levels (Ͼ50 mg/dl) have been proposed as a way to save health care costs without reducing sensitivity (4, 5); however, other acceptance criteria have not been evaluated.Most patients with viral CNS infection have an abnormal CSF leukocyte count, which usually ranges from 10 cells/mm 3 to 200 cells/mm 3 (6). The CSF leukocyte counts reported in previous studies (1, 2, 5-13) are summarized in Table 1. In our institution, we accepted CSF specimens for HSV PCR testing if they had Ͼ10 cells/mm 3 . This cutoff was based both on our experience before the study and on data from the studies cited above. We also accepted specimens collected from immunocompromised patients and children younger than 2 years of age.Clinical laboratories attempt to decrease costs while maintaining the quality of the diagnostic approaches used. A...