Zambia Population-based HIV Impact Assessment 2021

Households were the highest-level observation. Each sampled household has been represented as a single record (row) on the household dataset, including sampled households that were ultimately determined to be ineligible (i.e. vacant, not a dwelling, or dwelling destroyed) or were non-responding households. Each household has been identified by a unique householdid. Households that participated in the household survey were indicated as eligible respondents by the variable hhstatus = 1.

Roster records were the next level observation. Each individual rostered has been represented by a single record (row) on the roster dataset. All individuals who slept in the household the night before or who were usual residents were included on the household roster; this information was provided by the head of the household. Roster data contains individual-level roster data collected during the household interview for individuals of all ages, including those who were not eligible, who did not consent or who were not interviewed. Individuals who participated in the individual interview were indicated as eligible respondents by the variable indstatus = 1. Individuals who participated in biomarker testing and had valid laboratory test results were indicated by the variable bt_status = 1. Each person on the Roster dataset has been identified by a unique personid.

The adult interview records have been contained within the adult individual interview data. Within each responding household, eligible responding individuals as identified by the variable indstatus = 1 have been represented as a single record (row) on the interview datasets. ID variables were identical to those in the Roster dataset.

Individuals who were interviewed and consented to the biomarker testing as identified by the variable bt_status = 1 have been represented as a single record (row) on the Biomarker dataset. ID variables were identical to those in the Roster dataset.