Background:Since the onset of the COVID-19 pandemic, mathematical models have been widely used to inform public health recommendations regarding COVID-19 control in healthcare settings.Objectives:To systematically review SARS-CoV-2 transmission models in healthcare settings, and summarise their contributions to understanding nosocomial COVID-19.Methods:Systematic search and review.Data sources:Published articles indexed in PubMed.Study eligibility criteria:Modelling studies describing dynamic inter-individual transmission of SARS-CoV-2 in healthcare settings, published by mid-February 2022.Participants and interventions:Any population and intervention described by included models.Assessment of risk of bias:Not appropriate for modelling studies.Methods of data synthesis:Structured narrative review.Results:Models have mostly focused on acute care and long-term care facilities in high-income countries. Models have quantified outbreak risk across different types of individuals and facilities, showing great variation across settings and pandemic periods. Regarding surveillance, routine testing - rather than symptom-based testing - was highlighted as essential for COVID-19 prevention due to high rates of silent transmission. Surveillance impacts were found to depend critically on testing frequency, diagnostic sensitivity, and turn-around time. Healthcare re-organization was also found to have large epidemiological impacts: beyond obvious benefits of isolating cases and limiting inter-individual contact, more complex strategies such as staggered staff scheduling and immune-based cohorting reduced infection risk. Finally, vaccination impact, while highly effective for limiting COVID-19 burden, varied substantially depending on assumed mechanistic impacts on infection acquisition, symptom onset and transmission. Studies were inconsistent regarding which individuals to prioritize for interventions, probably due to the high diversity of settings and populations investigated.Conclusions:Modelling results form an extensive evidence base that may inform control strategies for future waves of SARS-CoV-2 and other viral respiratory pathogens. We propose new avenues for future models of healthcare-associated outbreaks, with the aim of enhancing their efficiency and contributions to decision-making.