Coverage of HIV Testing

Percent distribution of women and men eligible for HIV testing

Definition

Percent distribution of women and men age 15-49 eligible for HIV testing who were tested, refused to provide blood, were absent at the time of blood collection, or had other/missing results, by completeness of individual interview.

Coverage:

Population base: Women and men age 15-49 eligible for HIV testing (PR file, AR file)

Time period: Current status at time of survey

Numerators:

Number of de facto women (or men) age 15-49 eligible for HIV testing who:

1)      Were DBS tested (women: ha63 = 1 & hiv03 ≠ 8, na, men: hb63 = 1 & hiv03 ≠ 8, na)

2)      Refused to provide blood (women: ha63 = 3, men: hb63 = 3)

3)      Were absent at the time of blood collection (women: ha63 = 2, men: hb63=2)

4)      Had other/missing results (women: ha63 in 4, 5, 6, 9 or hiv03 in 8, na, men: hb63 in 4, 5, 6, 9 or hiv03 in 8, na)

By completeness of individual interview (women: ha65 = 1, men: hb65 = 1)

Numerator 1 includes women (or men) whose blood specimen completed the HIV testing algorithm at the lab for the survey with a final HIV test result of positive, negative, or inconclusive.

Denominator: Number of de facto women (or men) age 15-49 eligible for HIV testing (household selected for HIV testing and hv103 = 1). In some surveys, HIV testing is conducted in a sub-sample of households, usually those selected for the men’s survey (hv027 = 1)

Variables: PR file, AR file.

Calculation

The HIV test results data (AR file) should first be merged to the household members dataset (PR file) by cluster number (hv001/hivclust), household number (hv002/hivnumb) and line number (hvidx/hivline).

Numerator 1) includes all blood specimens tested at the lab and for which there is a final result, i.e., positive, negative, or inconclusive. If the woman or man provided a blood specimen for HIV testing (ha63 = 1 or hb63 = 1) but either there was no laboratory test result (hiv03 = na) or there was insufficient sample to complete the testing protocol (hiv03 = 8) then the result of testing is updated from numerator 1 (blood specimen tested) to numerator 4 (other/missing result).

Numerator 4) includes: 1) other results of blood collection (e.g., technical problem in the field), 2) lost specimens, 3) non-corresponding bar codes, and 4) laboratory results such as blood not tested for technical reason or not enough blood to complete the algorithm.

Numerator divided by the denominator, multiplied by 100.

Handling of Missing Values

Individuals who consented to the HIV test and provided a blood specimen for testing, but who do not have a corresponding final HIV test result are considered “not tested”. They are excluded from the denominator and numerator of the HIV prevalence estimate. Thus missing data on blood specimen tested (ha63/hb63) are included in numerator 4 (other/missing result).

Notes and Considerations

HIV prevalence estimates for surveys conducted by The DHS Program are derived from centralized laboratory testing on blood samples collected from respondents during the survey fieldwork. Response rates are calculated among all persons listed in the household roster as de facto household members in the age range defined as eligible for the survey interview and HIV testing. HIV testing is often conducted in a subsample of selected households. The population eligible for survey interview and HIV testing in this response rate is restricted to the population in households selected for the subsample with HIV testing. This is often the same group of households in the subsample for male interview. As a general rule, only those who agree to be interviewed, and successfully complete the interview, are offered HIV testing.

Technical reasons for the absence of an HIV test result include a poor quality or insufficient blood specimen that could not be tested or inability to link the individual tested with their HIV test result due to loss of the specimen or incorrect labeling of the specimen.

The coverage of HIV testing indicators are produced unweighted as they reflect the coverage of the survey, not a population estimate.

An overall HIV test response rate can also be calculated by multiplying the individual HIV response rate by the response rate for the household interview.

The indicators are presented separately by sex and for both sexes combined. For both sexes combined the data are restricted to women and men age 15-49.

References

UNAIDS/WHO Working Group on Global HIV/AIDS and STI Surveillance. 2005. Guidelines for measuring national HIV prevalence in population-based surveys. http://apps.who.int/iris/handle/10665/43349

World Health Organization (WHO). 2015. Consolidated Guidelines on HIV Testing Services. Geneva: WHO. https://www.who.int/publications/i/item/978-92-4-155058-1

Resources

DHS-8 Tabulation plan: Tables 14.1 and 14.2

API Indicator IDs:

Women:
HA_CHVT_W_DBI, HA_CHVT_W_DBN, HA_CHVT_W_RFI, HA_CHVT_W_RFN, HA_CHVT_W_ABI, HA_CHVT_W_ABN, HA_CHVT_W_OTI, HA_CHVT_W_OTN
(API link, STATcompiler link)

Men:

HA_CHVT_M_DBI, HA_CHVT_M_DBN, HA_CHVT_M_RFI, HA_CHVT_M_RFN, HA_CHVT_M_ABI, HA_CHVT_M_ABN, HA_CHVT_M_OTI, HA_CHVT_M_OTN
(API link, STATcompiler link)

Both sexes:

HA_CHVT_B_DBI, HA_CHVT_B_DBN, HA_CHVT_B_RFI, HA_CHVT_B_RFN, HA_CHVT_B_ABI, HA_CHVT_B_ABN, HA_CHVT_B_OTI, HA_CHVT_B_OTN
(API link, STATcompiler link)