Background:
Despite increased social vulnerability and barriers to care, there has been a paucity of data on SARS-CoV-2 incidence among key populations in sub-Saharan Africa. We seek to characterize active infections and define transmission dynamics of SARS-CoV-2 among people who inject drugs (PWID) and their sexual and injecting partners from Nairobi and the coastal region in Kenya.
Methods:
This was a nested cross-sectional study of SARS-CoV-2 infection from April to July 2021 within a cohort study of assisted partner services for PWID in Kenya. A total of 1000 PWID and their partners (500 living with and 500 living without HIV) were recruited for SARS-CoV-2 antibody testing, of whom 440 were randomly selected to provide self-collected nasal swabs for real-time PCR testing. Whole genome sequencing (WGS) was completed on a limited subset of samples (Nf23) with cycle threshold values £ 32.0. Phylogenetic tree construction and analysis was performed using the Nextstrain pipeline and compared with publicly available SARS-CoV-2 sequences from GenBank.
Results:
A total of 438 (99.5%) participants provided samples for SARS-CoV-2 PCR testing. Median age was 37 (IQR 32-42); 128 (29.2%) were female; and 222 (50.7%) were living with HIV.The overall prevalence of SARS-CoV-2 infection identified by RT-PCR was 86 (19.6%). In univariate analyses, there was no increased relative risk of SARS-CoV-2 infection related to positive HIV status, frequenting an injection den, methadone treatment, unstable housing, report of any high-risk exposure, or having a sexual or injecting partner diagnosed with COVID-19 or who died from COVID-19 or flu-like illness.Eight samples were successfully sequenced via WGS and classified as WHO variants of concern: 3 Delta, 3 Alpha, and 2 Beta. Seven were classified into clades predominantly circulating in Kenya during 2021. Notably, two sequences were identical and matched identically to another Kenyan sequence, which is consistent with, though not indictive of, a transmission linkage.
Conclusions:
Overall, the risk of SARS-CoV-2 infection in this population of PWID and their partners was not significantly associated with risk factors related to injection drug use. At a genomic level, the SARS-CoV-2 strains in this study were consistent with contemporary Kenyan lineages circulating during the time and not unique to PWID. Prevention efforts, therefore, must also focus on marginalized groups for control given the substantial amount of mixing that likely occurs between populations.
Predictors of active SARS-CoV-2 infection among PWID and their partners in Kenya & representative phylogenetic tree using Nextstrain