There is a hypothesis among epidemiologists that presence of bacteria in the blood, or bacteraemia, is a risk factor for serious health conditions such as cardiovascular disease. Combine this with the fact that cardiovascular disease is the leading cause of death globally and the motivation behind this thesis becomes clear. We begin by establishing the methods available for likelihood estimation in a Cox proportional hazards model in the context of a case-cohort study, along with any necessary theory required to understand the subject matter. An investigation into how such methods work, as well as a comparison of said methods is achieved by the use of simulations, giving a deeper understanding of the material and the mechanisms involved. Everything learned up until that point will then be implemented when we investigate the Oslo II subcohort and we explore the predictability of bacterial blood sample readings on cardiovascular disease mortality when adjusting for known risk factors and age as a confounding variable. The simulation study proved that the use of the Lasso for a case-cohort study will not identify all true predictors but will equally identify very few false ones. This, along with the adaptability of the Kalbfleisch and Lawless estimator and a rule designed to help computation with rare covariates, were the fundamental aspects of the simulation study that were applied in the practical analysis of the Oslo II subcohort. The various results obtained indicate that presence of the Kocuria and Enhydrobacter genera are predictive of cardiovascular disease mortality; while to a lesser extent the absence of Bacteroides, Streptococcus, Paracoccus and Stenotrophomonas are also predictive of the same result. The background risk factors to which the bacteria genera were adjusted also require confirmation: history of diabetes, use of blood pressure medication, history of myocardial infarction, history of heart failure, systolic blood pressure, a self-health evaluation, level of tough physical activity and excessive coffee consumption were all deemed significant predictors of cardiovascular disease mortality in this subcohort.