Proportion positive for Epstein-Barr virus, cytomegalovirus, human herpesvirus 6, Toxoplasma, and human immunodeficiency virus types 1 and 2 in heterophile-negative patients with an absolute lymphocytosis or an instrument-generated atypical lymphocyte flag

Objectives.-To determine the proportion of patients with evidence of an acu te infection due to Epstein-Barr virus (EBV), cytomegalovirus (CMV), human herpesvirus 6 (HHV-6, Toxoplasma, or human immunodeficiency virus types 1 a nd 2 (HIV-1 and HIV-2) in heterophile-negative patients with an absolute ly mphocytosis or an instrument-generated atypical lymphocyte flag, and to dev elop a cost-effective testing algorithm for managing such heterophile-negat ive patients. Design.-We conducted a prospective investigation of 70 selected outpatients who tested negative for heterophile antibody in association with an absolu te lymphocytosis or instrument-generated atypical lymphocyte flag. The cont rol population consisted of 50 patients who were heterophile negative and h ad a normal absolute lymphocyte count and no instrument-generated atypical lymphocyte flag. Setting.-A large outpatient laboratory system. Intervention.-Viral serology for HHV-6 was performed by immunofluorescence, and all other serologies were performed by enzyme-linked immunoassay. All testing was for immunoglobulin (Ig) M antibodies, except in the case of HIV . Results.-The proportion of study patients positive for EBV was 40% (28/70); for CMV, 39% (27/70); for HHV-6, 25% (16/65); for Toxoplasma, 3% (2/70); a nd for HIV, 0% (0/70). All 50 control patients were negative for EBV IgM an tibodies. When patients with more than 1 positive viral test were excluded from analysis, positivity was 20% (9/45) for EBV, 22% (10/45) for CMV, 9% ( 4/45) for HHV-6, and 2% (1/45) for Toxoplasma. Utilizing hypothesis-generat ing logistic regression models, Downey type II atypical lymphocytes were si gnificantly associated with EBV positivity (P = .006), while Downey type II I lymphocytes were significantly associated with HHV-6 positivity (P = .016 ), and there was a trend for the association of Downey type I lymphocytes w ith CMV positivity (P = .097). Conclusions.-A positive viral serology was identified in 70% of study patie nts. Multiple positive serologies complicate establishing a definitive diag nosis. Potential cost savings may be associated with the use of an appropri ate testing algorithm.