HIV seroprevalence among all TB patients
Surveillance of HIV prevalence among TB patients will give information about the epidemics of both TB and HIV. In particular, it gives an indication of the degree of overlap in the epidemics in any given setting and, when compared with HIV prevalence in the general population, an indication of the contribution that HIV is making to the TB epidemic.
Total number of newly registered TB patients (registered over a given period of time) who are HIV-positive.
Total number of newly registered TB patients (registered over the same given time period) who were tested for HIV and included in the surveillance system.
Selecting the appropriate strategy for HIV surveillance among TB patients will depend mainly on the existing surveillance system and the underlying HIV epidemic state in a country. There are three main methods for surveillance of HIV among TB patients:
Routine HIV testing data can form the basis of a reliable surveillance system at all levels of HIV epidemic (low-level, concentrated, generalized2), provided that high coverage is achieved (more than 80% of all TB patients giving consent and being tested). These routine data can be calibrated by periodic (special) or sentinel surveys.
Sentinel surveillance collects information in a regular and consistent way from a predetermined number of persons from specific sites and population groups that are of particular interest or are representative of a larger population. The difficulty with sentinel surveillance is in determining how representative the sampled population is of the population from which they are taken and also how representative they are of the general population of TB patients. Sentinel surveillance systems are usually based on unlinked anonymous testing methods, often using blood specimens that have been collected for other purposes and stripped of all identifying markers.
Periodic special surveys have a specific role where the prevalence of HIV among TB patients has not been previously estimated and are an essential part of the initial assessment of the situation. Surveys using representative sampling methods and appropriate sample sizes can provide accurate estimates of the burden of HIV in TB patients. This information may alert TB programmes to a potential HIV problem and enable action to be taken, which may include the institution of more systematic surveillance.
Ideally surveillance of HIV prevalence should include all newly registered TB patients, diagnosed according to international standards.3 However, if periodic special surveys or sentinel methods are used and resources are limited, countries may choose to include only adult smear-positive pulmonary patients, i.e. those with a definitive diagnosis of TB. Countries with scarce resources and an HIV epidemic state that is either low or concentrated may also choose to include only a smaller subgroup of TB patients, e.g. adults aged 1559 years. Unless relapse cases are identified as such and the results analysed separately, they should be excluded from surveillance systems because of the risk of surveying the same patient twice. However, relapse cases may be included and need not be identified as such if surveillance is based on survey methods and these surveys are undertaken over a short period of time, ideally less than 23 months. All countries with a generalized HIV epidemic state should aim to ensure that HIV counselling and testing are offered and actively promoted to all TB patients, in conjunction with ART where possible.
Every 2-3 years
Geographic location: N/A
Pregnancy status: N/A
Time period: N/A
Type of orphan: N/A
Vulnerability status: N/A
Measuring HIV seroprevalence among TB patients can inform the targeting of resources, the planning of prevention, care and support for people with HIV and TB, and monitoring the effectiveness of these activities over time. It can raise awareness among policy-makers and health care workers about HIV-related TB and the need for a collaborative approach to the problem. It is also helpful to corroborate surveillance data on HIV prevalence in the general population obtained from other sources. In low HIV epidemic states it will provide an early indication of changes in the HIV epidemic, alerting policy-makers to the need for joint strategies. In concentrated or generalized HIV epidemics it will help in assessing the impact of HIV upon TB and monitor the effectiveness of joint strategies to reduce the burden of HIV and TB. Even if more than 80% of patients are tested, using routine programme data from TB patients tested as part of their care carries the risk of bias if those who are tested are very different from those who are not tested. The use of unlinked anonymous surveys to derive HIV prevalence data is increasingly criticized because of the advantages to patients of knowing their status and the ethics of carrying out HIV testing in patients not offered VCT.