Question special
Faculty Member

Among the 1954 patients assessed for eligibility, roughly 20% were randomized. Therefore, the findings of this trial apply to only a small percentage of the patients. As a result, external validity (generalizability) is a major issue.
In addition, in true RCTs all intervention groups must be treated identically except for the experimental treatment. However, in this trial all patients in the IV group remained at the hospital; those in the oral group were discharged. The median length of stay in the hospital was 19 days for the IV group versus 3 days for the oral group. In the United States, the majority of these patients would have been discharged with Home IV abx. A prolonged hospital stay by itself is associated with poor outcomes due to nosocomial /iatrogenic complications. This situation could have potentially created a bias by favoring the oral group. Therefore, internal validity is also a potential concern.