Diagnostic and Prognostic Value of ESR and CRP in HIV Positive Adults in Durban, South Africa.

Abstract

Purpose: To review the data of ESR and CRP for screening, diagnosing, and prognosticating adults who are co-infected with HIV and MTB. Additionally, we review the same data to assess MTB treatment response in the same patients. Methods: We conducted a retrospective study of patients with suspected smear negative pulmonary tuberculosis at King Edward VII hospital in Durban, South Africa from October 2007 to November 2009. Sputum culture for M. tuberculosis was the gold-standard diagnostic test for active pulmonary TB. Results were stratified by HIV-infection and CD4 count. Participants were over 18 years of age, had any 2 of 4 tuberculosis symptoms, (cough, fever, night sweats and loss of weight) for more than two weeks. Data from 90 patients was for ESR and CRP at baseline, after 2 weeks and monthly up to six months during treatment for MTB. Multiple statistical models including McNemar’s test, and paired Student’s t-tests as well as regression models were used to analyze the data using SAS version 9.2. Results: ESR and CRP had high diagnostic sensitivity for culture-positive TB. Therefore, a normal result may be valuable to “rule out” active disease. We showed that a combination of (ESR 100 or CRP 10) has a sensitivity of 100% and a specificity of 23% among culture confirmed TB patients. When comparing ESR and CRP between culture confirmed TB patients and those who tested negative for TB, at baseline ESR and CRP were both elevated. However, both ESR and CRP levels decreased during the course of treatment. Conclusions: ESR and CRP had high diagnostic sensitivity for culture-positive TB. Therefore, a normal result may be valuable to “rule out” active disease. ESR and CRP which are elevated at baseline among HIV/TB patients can potentially be valuable biomarkers to monitor response to therapy.