Ventilator-associated pneumonia and mortality

摘要: Ventilator-associated Pneumonia (VAP) is one of the most common nosocomial infections among patients admitted to intensive care unit. It generally believed that VAP increases mortality patients, however exact determination attributable challenging. This thesis demonstrates can be investigated in several ways: 1. The use observational studies at risk for VAP, whom some will develop (and not) and subsequent comparison crude rates those with without VAP. obviously flawed by many variables - aside from developing or not differ between both patient groups also influence mortality. 2. same studies, but now additional adjustments differences are equally distributed Yet, these methods overestimate because – inevitable incomplete adjustment unmeasured baseline, as well changes clinical condition during ICU stay. 3. One could theoretically ethically randomize receive count deaths each group. 4. Using data randomized controlled trials on prevention. provides benefit randomization, which excludes effects measured groups. 5. novel statistical methods, such competing survival analyses including a time dependent exposure. From described this it concluded estimated around 10% (9% chapter 3 13% 4), much lower than estimates provided previous studies. Subgroup revealed an absence trauma medical low high severity illness admission. Highest found surgical intermediate score subgroups critical importance design interpretation prevention trials. Many were hugely underpowered demonstrate improvements outcome. new knowledge underpins need large (>1000 per study group) whether improves Furthermore, rational focus subgroup highest rates, considering subgroups. Further methodological needed determining especially concerning limit confounding. multistate models might improve As enable various events, thus offer more flexibility better capture (chronological) complexity factors influencing