作者: Stephen P Blatt , Catherine R Lucey , Clifford A Butzin , Craig W Hendrix , Daniel R Lucey
DOI: 10.1001/JAMA.1993.03500050100034
关键词:
摘要: Objective. —To determine whether the total lymphocyte count (TLC) accurately predicts a low absolute CD4 + T-cell and percentage in persons infected with human immunodeficiency virus (HIV). Design. —Retrospective analysis of data collected US Air Force HIV Natural History Study. Setting. —Military medical center that performs annual evaluation all HIV-infected personnel. Patients. —A 828 consecutive patients no prior history zidovudine use, evaluated from January 1985 through July 1991. For multiple observations over time, single point within each 6-month interval was included (N=2866). Measurements Main Results. —The sensitivity, specificity, likelihood ratio (LR) TLC, range 1.00×10 9 /L to 2.00×10 /L, predicting an less than 0.20×10 or 20% were calculated. In addition, LR pretest probability significant immunosuppression used calculate posttest probabilities for given TLC value. The increased 2.4(95% confidence interval, 2.2 2.5) TLCs 33.2 (95% 24.1 45.7) /L. specificity this prediction 57% 97% range. also 1.4(95% 1.3 1.6) 9.7 7.1 13.1) 20%. Conclusions. between appears be useful predictor as measured by CD4+ persons. value can individual patients. example, patient 1.50×10 16%, increases 53%. contrast, greater 30% will decrease 4%. Physicians should find these help predict risk opportunistic infection among who present syndromes are potentially compatible but have not had recent analysis. ( JAMA . 1993;269:622-626)