To the Editor.—
While the efforts of Gorbea and colleagues1 to define safe infection control practice is laudable, I found their application of statistical methods faulty. Their study1 examined the safety of changing intravenous (IV) tubing containing burettes at 48 hours instead of the usual 24 hours. Since contamination was the end point studied, interest is in contamination rates at 24 hours in the 24-hour group and contamination rates at 48 hours in the 48-hour group. Even if the McNemar test were capable of showing independence between the 24- and 48-hour specimens in the 48-hour group, the 24-hour specimen results do not contribute to the determination of safety at 48 hours and are irrelevent. In fact, they only inappropriately inflate the denominator.2 Using the data in Table 2 and the text,1 I calculate that there were five positive 48-hour cultures. If all 118 burettes had a 48-hour culture,
Josephson A. Changing Intravenous Tubing Containing Burettes. JAMA. 1985;253(1):42. doi:10.1001/jama.1985.03350250050016
Customize your JAMA Network experience by selecting one or more topics from the list below.