Table 2.

Methods for sampling key populations (KP) for viral load monitoring

Sampling methods	Advantages	Disadvantages
Key population programme service delivery	KP‐identifiable and viral loads returnable using programmatic resources Blood draws can be collected in the community where KP are more easily reached	Sample includes those engaged in services and underrepresents those not engaged in care Data may include duplicates if biometrics are not utilized
Clinic data and national registries	Data longitudinal, assuming individuals are retained in care	Difficult and often impossible to identify KP through clinic data or national registries Sample includes those engaged in services and under‐represents those not engaged in care
Social network‐based recruiting, such as respondent driven sampling or snowball sampling	Methods can reach those not engaged in care Results may be generalizable to underlying population of interest If serial cross‐sectional studies are conducted, can ascertain insight into changes over time in terms of KP viral suppression	Difficult to verify that individuals truly belong to KP Need to account for recruitment methods, which may not be possible for subanalyses such as viral suppression due to breaks in chains since not all individuals enrolled will be living with HIV
Venue based sampling	Efficient recruitment method Community‐based viral load monitoring may reach those not engaged in care	May be difficult at certain venues to verify that individuals truly belong to KP Individuals who do not attend venues are not represented and may be substantively different from those that do In the absence of point‐of‐care diagnostics, returning results to individuals may be challenging Individuals may be recruited at multiple sites, potentiating duplicate enrolments if biometrics are not utilized