2Literature search strategy and information retrieval

Studies were identified using PubMed, EMBASE, Web of Science, Google Scholar and the Cochrane Library databases. Proceedings and abstracts from international conferences (CROI, IAS, ICAAC, World AIDS Conference) were searched.

Selection criteria

Studies were included if:

  • randomized and quasi-randomized controlled trials, including historically controlled trials or observational and cohort studies;
  • participants being presumed or confirmed TB cases or/and partners and family members of those who turn to be HIV positive;
  • comparison being addressed was: routine HIV testing vs. no HIV testing.
Image annex2fu1

Most data that can be found on this topic is circumstantial evidence, most likely because it is deemed unethical to compare groups in which HIV testing is offered against one in which it is not. This would deprive people of the right to then make choices about treatment.

The only study which has a comparison group was carried out in Malawi and compared patients after the introduction of voluntary counselling and testing (VCT) and co-trimoxazole prophylaxis therapy (CPT) to patients before (1); unfortunately there was no stratification and therefore it is not amenable to GRADE appraisal.

It is now clear that people with HIV-associated TB fare a lot worse than those with HIV or TB separately. Treatment for HIV is becoming increasingly available, and evidence shows that even just CPT is beneficial in decreasing morbidity and mortality. Knowledge of HIV status is also beneficial for sexual partners and unborn children. Thus it is hard to imagine the need for evidence to be collected to prove that testing patients with TB for HIV is beneficial.

A large number of studies show that testing for HIV in confirmed TB patients (mostly) and in presumptive TB patients and their contacts (occasionally) yields a high number of new diagnoses of HIV. The assumption is that this allows newly diagnosed HIV-positive people to receive CPT and ART, thus hypothetically decreasing their morbidity and mortality and preventing HIV transmission. These same descriptive studies also show that often after diagnosis, either the HIV treatment is not offered or available, or it is not taken up. According to the HIV/AIDS universal access report, only 17% of the estimated number of TB patients living with HIV were receiving ART, a figure considerably lower than the estimated coverage rate of antiretroviral therapy for all HIV patients in low- and middle-income countries (2).

Table 1 summarizes the 24 studies retrieved from the systematic literature search. It shows that in these studies HIV testing in TB patients can yield positive results anywhere between 6.3% and 77% depending on the epidemiologic setting. Although 6.3% appears low, consideration should be given to the fact that HIV is a transmissible, 100% fatal disease without available treatment and prevention measures. A few of the studies addressed the issue of contacts or cases with presumptive TB, although they yield very little data. In one study carried out in Kenya on presumptive TB patients, 61% diagnosed with TB were also HIV-positive, whereas 63% without TB were positive (3). Another study in Uganda (4) found that 39% patients diagnosed with TB were also HIV-positive, whereas 49% of those who had no active TB were HIV-positive. The yield of HIV-positive results among those with presumptive TB was 39% in Guinea-Bissau and 61% in Zimbabwe (5, 6).

Only one study addressed the issue of testing close contacts of TB patients and on doing so found that 13.8% of contacts of TB patients living with HIV were HIV-infected, whereas 2.5% of contacts of HIV-negative TB patients were HIV-infected (7). One study attempted to quantify the benefit of testing by looking at relative risk of death of a cohort of patients in one programme before and after the introduction of VCT (1). It found the adjusted relative risk of death to be 0.81 (p<0.001).

In all studies, the prevalence of HIV was found to be higher in TB patients, those with presumptive TB and possibly even among TB contacts than the expected national adult HIV prevalence. One study highlighted higher rates of mortality among HIV-positive TB patients (29%) compared to those with TB only (8%) (24). Interestingly, 34% of those who refused testing died, but a possible bias is that anybody very unwell and suspicious of having HIV is possibly more likely not to test.

Table 1. Summary of studies on testing for HIV infection in TB patients, people with suspected TB and contacts of TB patients.

Table 1

Summary of studies on testing for HIV infection in TB patients, people with suspected TB and contacts of TB patients.

From: Annex 2, Voluntary counselling and testing for HIV benefits patients with diagnosed and presumptive tuberculosis – summary of findings and evaluation of the quality of the evidence

Cover of WHO Policy on Collaborative TB/HIV Activities
WHO Policy on Collaborative TB/HIV Activities: Guidelines for National Programmes and Other Stakeholders.
Copyright © 2012, World Health Organization.

All rights reserved. Publications of the World Health Organization are available on the WHO web site (www.who.int) or can be purchased from WHO Press, World Health Organization, 20 Avenue Appia, 1211 Geneva 27, Switzerland (tel.: +41 22 791 3264; fax: +41 22 791 4857; e-mail: tni.ohw@sredrokoob).

Requests for permission to reproduce or translate WHO publications – whether for sale or for noncommercial distribution – should be addressed to WHO Press through the WHO web site (http://www.who.int/about/licensing/copyright_form/en/index.html).

NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.