We design record-linkage studies across health registers and administrative databases to ascertain key event-dates (such of hospitalization, incarceration, drug treatment, or sentencing) for national 'virtual' cohorts of hard-to-reach individuals. We then analyse cause-specific morbidity and mortality rates and make proposals for, or test, interventions to reduce identified major risks such as from carriage of undiagnosed Hepatitis C virus (HCV), or of overdose death soon after release from prison. The hard-to-reach populations we have focused on are caught up in addictions, criminal justice or epidemics of injection-related blood-borne viruses. We have needed to estimate how many current injectors there are by capture-recapture techniques; and to project their future drugs-related deaths. By record-linkage, we have quantified hazard-rates for ex-prisoners, HCV carriers, drug treatment clients, and ever-injectors: in the short-term (notably within 28 days of prison-release or hospital-discharge), by calendar era, and as individuals age or report misuse of alcohol. Our seminal Scottish record-linkage studies have been corroborated internationally and, in 2012, randomization began in our prison-based N-ALIVE Trial to determine if naloxone (opiate antagonist) on-release can reduce by 30% overdose-deaths soon after release from prison. In addition, our statistical work has determined the primary and secondary outcomes for how Scotland conducts 5-year before/after monitoring of whether its public health policy on take-home naloxone will have reduced opiate-related deaths when these were on an age-related increasing trajectory during the baseline period of 2006-10. During 2005-09, Sheila Bird was the Royal Statistical Society's (RSS) vice-president for external affairs and introduced both the RSS's statistical seminars for journalists and awards for statistical excellence in journalism: for which, in 2010, she was awarded the RSS's Chambers Medal. Following swine-flu, on which she wrote a series of articles for Straight Statistics, Sheila Bird has led the Royal Statistical Society's campaign for legislation to end the late registration of deaths in England and Wales where fact-of-death is not registered for at least six months for one in five premature deaths (aged 5-44 years). From 2006 to 2014, Sheila Bird and the late Colonel Clive Fairweather CBE, reported 20-weekly on military fatality rates in Afghanistan by nationality and cause. In 2011, Sheila was appointed OBE for her services to social science and, in 2012, was elected a Fellow of the Royal Society of Edinburgh.