Clinical signs and symptoms in the assessment of immunodeficiency in men with subtype C HIV infection in Harare, Zimbabwe

HIV Clin Trials. 2002 Mar-Apr;3(2):148-54. doi: 10.1310/EHME-C6EA-U79L-H9Y6.

Abstract

Purpose: Providing low-cost interventions such as co-trimoxazole as prophylaxis against opportunistic infections among HIV-infected individuals depends on the identification of those at risk. This article describes the prevalence of self-reported signs and symptoms and CD4 cell counts in a cohort of 447 HIV seropositive men. A scoring system using self-reported signs and symptoms was developed and tested in the prediction of low CD4 cell counts. This approach may allow health care providers in low-resource settings to predict severe immunodeficiency and to provide care.

Method: Data on clinical manifestations of HIV infection and blood samples for HIV serology were collected prospectively from an ambulatory cohort of men seen at their workplace at enrollment and every 6 months thereafter. CD4+ cell counts were obtained on samples testing positive on ELISA. Using data reduction techniques and logistic modeling, we developed a prognostic score system.

Results: 20% of the men had CD4+ cell counts below 200. All reported signs and symptoms were more frequent in men with less than 200 CD4+ cell counts compared to men with CD4+ cell counts greater than 200. History of malaria, fever, lymphadenopathy, persistent diarrhea, persistent cough, and skin infections robustly predicted low CD4+ count. A scoring system equation was developed based on the coefficients of the multivariate logistic regression: 1x(tuberculosis) + 3.2x(herpes zoster) + 4.5x(malaria) + 5.7x(fever) + 5.8x(cough) + 8.2x(lymphadenopathy) + 8.5x(skin infection). Setting the score cutoff value greater than or equal 5, the model had moderately high sensitivity of 61% and specificity of 72%. The scoring system had an overall classification error rate of 30%.

Conclusion: By using this simple scoring system, physicians can correctly identify 72% of patients who do not require immediate intervention, thereby channeling scarce resources to those who have both low CD4+ cell counts and symptoms and are most likely to benefit from prophylactic and antiretroviral interventions.

Publication types

  • Research Support, Non-U.S. Gov't
  • Research Support, U.S. Gov't, P.H.S.

MeSH terms

  • AIDS-Related Opportunistic Infections / physiopathology
  • CD4 Lymphocyte Count
  • Cohort Studies
  • HIV / classification
  • HIV Infections / complications
  • HIV Infections / immunology
  • HIV Infections / physiopathology*
  • HIV Infections / virology
  • Humans
  • Male
  • Prospective Studies
  • Zimbabwe