This study investigated the prevalence and genomic epidemiology of SARS-CoV-2 variants as part of the TQT-COVID-19 Project, which aimed to expand COVID-19 testing, quarantine, and telemonitoring strategies. A total of 177 samples were analyzed between July 2022 and April 2023 from patients of 17 Primary Health Care (PHC) units in Salvador, Bahia, Brazil. Sequencing was done using Oxford Nanopore MinION technology and bioinformatics analysis for variant typing and phylodynamic spread assessment. Most participants were women (68.5%), under 39 years old (48.9%), used public transportation (66.4%), had no comorbidities (66.1%), presented mild COVID-19 symptoms (40.8%), and had received a second booster dose of a COVID-19 vaccine (88.2%). Nevertheless, 11.8% had never received a COVID-19 vaccine. The most prevalent Omicron subvariants were BA.5 (32.6%), BA.4 (21.7%), B (18.8%), and BQ.1 (26.3%), with other variants like XBB also identified. Vaccinated and unvaccinated individuals had different spike gene mutations. While multiple variant introductions occurred over time, SARS-CoV-2 geographic spread was mainly localized, with occasional rapid spread associated with inter-city travel or containment failures. These findings underscore the need to track viral spread and mutations to develop targeted strategies for breaking transmission chains.