AEGiS-15IAC: Predicting the CD4 count from the result of dinitrochlorobenzen skin test using linear regression analysis.

15th International AIDS Conference


Bangkok, Thailand - July 11-16, 2004

DonateNow
Print this article

Predicting the CD4 count from the result of dinitrochlorobenzen skin test using linear regression analysis.

Int Conf AIDS 2004 Jul 11-16; 15:(abstract no. B10437)

Nguyen NL
Faculty of Medicine, Hue, Vietnam

BACKGROUND:The CD4 count is being used to evaluate the immune status of HIV- infected person, but it is too expensive to become popular in developing countries. The DNCB (Dinitrochlorobenzen) skin test has been used by immunologist to evaluate the cell-mediated immune response. Some investigators proposed some blood parameters such as, absolute count of lymphocyte, the total count of leukocyte, to replace CD4 count for this aim with too large confident interval. I tried to predict the CD4 count from the results of DNCB skin test using linear regression analysis. MethodsA control case, double blind randomized trial was implemented in 140 persons in Hue and Hai phong City during 3 years (1999-2003). It consists of 4 groups, 35 each, for the group HIV (-), in stages 1 and 2 of HIV infection, in stage 3, in AIDS stage, respectively (criteria of WHO classification). Each person was tested with the DNCB skin test and CD4 count simultaneously. I use the simple linear regression analysis to calculate the regression equation. ResultThe slope of the regression line = 11.37; 95% CI = 9.82 to 12.91 ; t = 14.55; P< 0,0005. Y (CD4 count) = 166.06 + 11.37X (diameter of the DNCB skin reaction); r square = 0.60. ConclusionThe slope of the regression line is significantly greater than 0, indicating that we could predict the CD4 count from the result of DNCB skin test (cheaper, simpler, easier to become popular than CD4 counter). This suggests a high correlation between CD4 and the diameter of the DNCB skin reaction which may allow financially constrained HIV care services to substitute the DNCB test for the more expensive CD4 counts

Keywords: AEGIS, CD4 Lymphocyte Count, Regression Analysis, Linear Models, HIV Infections, Acquired Immunodeficiency Syndrome, Skin Tests, HIV Seropositivity, Antigens, CD4, Tuberculosis, immunology

040711
B10437

Copyright © 2004 - International AIDS Society (IAS). Reproduction of this abstract (other than one copy for personal reference) must be cleared through the IAS.