Title

Population size estimation within a biobehavioral survey among people who inject drugs and men who have sex with men in the survey cities of the Kyrgyz Republic, 2021

Presenter

Nazira Usmanova

Authors

N. Usmanova¹, B. Kalmyrzaev¹, D. Biibosunova¹, A. Mereke², A. Mclntyre², J. Neal², B. Sheralieva³, L. Yanbukhtina³, C. Zhumalieva³, A. Solpueva³, A. Bekbolotov³, M. Tursunbekov⁴, A. Karipova⁴, G. Esenkeldieva⁴, V. Ivakin⁵, A. Deryabina⁵, P. Nadol^1,1

Institutions

¹Centers for Disease Control and Prevention, DGHT, Bishkek, Kyrgyzstan, ²Centers for Disease Control and Prevention, DGHT, Atlanta, United States, ³Republican Center for Control of Blood Born Viral Hepatitis and HIV, Bishkek, Kyrgyzstan, ⁴ICAP at Columbia University, Bishkek, Kyrgyzstan, ⁵ICAP at Columbia University, Central Asia Regional Office, Almaty, Kazakhstan

BACKGROUND: The Kyrgyz Republic (KR) is a country with an HIV epidemic concentrated among key populations (KP). However, updated, high-quality population size estimates (PSE) for KP are lacking. Accurate KP PSE, particularly at a sub-national level, are necessary to design and evaluate programs and policies to prevent new HIV acquisitions and to monitor the dynamics of the epidemic. Incorporating empirical PSE methods within biobehavioral surveys may offer an efficient opportunity to generate valid PSEs for select populations.
METHODS: Using respondent-driven sampling (RDS), a biobehavioral survey (BBS) was conducted among PWID in Bishkek, Karabalta, Karasuu, Osh, Tokmok, and Sokuluk and among MSM in Bishkek and Osh in 2021. The BBS incorporated multiple empirical PSE methods – successive sampling PSE (SS-PSE), unique object multiplier, 3-source capture - recapture (3-SCR) and service multiplier. A Bayesian-based model was used to synthesize the various estimates to generate point estimates and plausibility bounds by location.
RESULTS: The results of the PSE exercise suggest that Bishkek had 6,100 MSM, constituting 2.0% of the adult male population (plausibility bounds [PB], 1.7%-2.4%); Osh – 860 (1.0%; 0.8%-1.1%). The number of PWID was 4,100 (1.4%; PB, 1.1%-1.6%) in Bishkek; 110 (1.1%; 0.7%-1.4%) in Kara-Balta; 380 (1.4%; 1.2%-1.63%) in Karasuu; 1,150 (0.9%; 0.7%-1.2%) in Osh; 710 (1.3%; 1.0%-1.6%) in Sokuluk, and 220 (1.3%; 1.0%-1.6%) in Tokmok.
CONCLUSIONS: These updated estimates generally fell within the globally recognized ranges. The triangulation of survey and programmatic data over time along with use of synthesis models could help refine PSE and support the optimal allocation of limited resources and improve target-setting and program monitoring.

Download the e-Poster (PDF)