The reported survival times for the virus in different matrices was subjected to meta-analysis to account explicitly for the effect of censoring, that is, studies which were interrupted while the agent was still detectable in the matrices investigated.

ONLY STUDIES WITH TEMPERATURE BETWEEN 0 AND 30 DEGREES CELSIUS (materials not frozen or heated) are eligible, as comparisons among studies with larger temperature variation would not be meaningful.

Studies confirming virus detection and studies in which detection was based on genetic material (DNA/RNA) were evaluated separately.

Plots show the parametric confidence intervals for the mean survival time when at least 4 study groups are reported. The plots are then repeated presenting interquartile ranges (IQ, Q1-Q3). Following the plots, a table gives the exact number for these intervals for each of the matrices reported. For the raw data (all study groups reported individually), see download link on the left panel.

lower.CI and upper.CI stand for the lower and upper control limit of a 95% confidence interval accounting for censoring (lack of information about true maximum when experiments ended while the agent was still detectable). q1 and q3 refer to the first and third quartiles (25% and 75% of values).

N.rows is the number of experiments from which pathogen survival data was available.