Quantifying the roles of vomiting, diarrhea, and residents vs. staff in norovirus transmission in U.S. nursing home outbreaks

Carly Adams; David Young; Paul A Gastañaduy; Prabasaj Paul; Zach Marsh; Aron J Hall; Benjamin A Lopman

In addition to regression analyses, we also used the The original dataset consisted of 209 lab-confirmed and probable cases from six separate 200 outbreaks. One case was excluded from the estimations of R Ei and all further analyses because 201 he/she was missing an illness onset date. After the estimations of R Ei , four additional cases were 202 excluded from the regression analyses because they were missing information on diarrhea, 203 vomiting, and/or resident vs. staff. Lastly, 9 more cases (4.3% of all cases with onset date 204 information) had symptom onset dates on the last day the outbreak and thus did not produce any 205 reported secondary cases. Therefore, they had estimated R Ei 's of zero. Because log R Ei could not 206 be taken for these cases, they were excluded from all regression analyses. Sensitivity analyses 207 were performed by adding 0.01 to these R Ei estimates to examine the influence of these cases on 208 model estimates.