Background: Tuberculosis and HIV are major causes of morbidity and mortality worldwide, and TB/HIV co-infection is a major barrier to global TB and HIV control efforts. Successful interventions exist to control TB and HIV individually, but they are less effective in areas where TB/HIV co-infection is prevalent partly because TB is difficult to diagnose in immunosuppressed persons. The World Health Organization recommends a symptom-based clinical screening tool to help clinicians rule out TB in HIV-infected patients and initiate appropriate treatment. To date, the performance of this screening tool has not been evaluated across a range of immune status.
Methods: We conducted a secondary analysis of data from a cross-sectional study of HIV-infected patients seeking care in Cambodia, Thailand, and Viet Nam to evaluate the performance of the WHO screening tool using a reference standard approach and mycobacterial culture status as the referent. Logistic regression modeling with piecewise linear splines was used to estimate the sensitivity and specificity of the screening rule compared to mycobacterial culture across a continuous range of CD4+ cell counts, comparing the effect of immune status on the tool's performance in persons with CD4+ cell counts above and below 400.
Results: Of 1,988 participants who underwent TB evaluation, 276 (13.8%) had TB and 1,514 (76.2%) had at least one of the 4 symptoms comprising the WHO screening tool. Among patients with CD4 counts below 400, the sensitivity (AOR 0.41) and (1- specificity) (AOR 0.765) decreased with increasing immune status. There is no evidence of a statistically significant effect of immune status above 400 cells on the sensitivity or specificity of the screening tool.
Conclusions: Our results indicate that among persons infected with HIV in South-East Asia, the WHO screening tool for TB is more sensitive with falling immune status. The screening rule is most effective among those with more severe immunosuppression, so it may improve TB case-finding and treatment initiation among those most susceptible to infection and adverse health outcomes.
Table of Contents
1. Introduction - 1 2. Literature Review - 5 2.1 Introduction and TB/HIV Epidemiology - 5 2.2 TB Diagnostics and the HIV Context - 8 2.3 TB/HIV Epidemiology and Control in S.E. Asia - 11 2.4 The Need for Screening Tests and How to Evaluate Their Effectiveness - 15 3. Methods and Results - 18 3.1 Methods - 18 3.1.1 Study population and specimen collection - 18 3.1.2 Statistical analyses - 20 3.2 Results - 22 4. Discussion, Conclusions, and Recommendations - 31 4.1 Review of major findings - 31 4.2 Discussion - 34 4.4 Conclusions and Recommendations - 38 References - 40 Appendices - 45 Appendix 1. Acronyms and Abbreviations - 45 Appendix 2. IRB Approval - 46 Appendix 3. Regression Models - 47 Appendix 4. Sensitivity and Specificity plots - 53 Table 1. Flowchart of participants with HIV infection - 23 Table 2. Frequency of selected characteristics by CFSW symptom screen status - 24 Table 3. Frequency of selected characteristics by MTB status - 24 Table 4. Frequency of potential confounders by MTB and CFSW symptom screen status - 26 Table 5. Presence of symptoms by CD4 cell count - 26 Table 6. Regression models for CFSW Positivity Among Patients with HIV - 28 Table 7. Multivariate Logistic Regression Models for CFSW Positivity Among Patients with HIV - 29 Table 8. Estimated Sensitivity and (1-Specificity) at Intervals of CD4 Count in HIV-infected Patients - 30
About this Master's Thesis
|Committee Chair / Thesis Advisor|
|The Effect of Immune Status on the Performance of a WHO-recommended Screening Tool for Tuberculosis in HIV-infected Patients in South-East Asia ()||2018-08-28||