Interesting to see that WHO SEARPO has published a document on key populations that 'crosses lines' by including data from countries that are not in the Southeast Asia region.

https://www.aidsdatahub.org/sites/default/files/publication/WHO_Are_Key_Populations_Really_the_KEY_to_Ending_AIDS_2018.pdf

The sex worker HIV prevalence graph on page 18 demonstrated a sampling bias in Myanmar. Early sentinel surveillance was closer to a convenience sample: women were tested when the police brought them in. As the size of the sample and its representativeness increased, the true prevalence was found to be low and slowly decreasing. Deaths contributed to this. The impact of programming is unknown.

Jamie