Chain size analysis clarifies surveillance needs.

<p>A) Applying maximum likelihood estimation to simulated data shows the sensitivity of chain size analysis and contact tracing analysis for detecting a change in . Results show the probability of detecting a significant change between the monkeypox data from the 1980s and a simulated data set with (equal to the ML value for the 1980s data) and specified along the x-axis. Statistical significance was determined by setting a 95% confidence threshold on the likelihood ratio test (details provided in <a href="http://www.ploscompbiol.org/article/info:doi/10.1371/journal.pcbi.1002993#s3" target="_blank">methods</a> section). Curves represent different values for the number of simulated chains, . Results are depicted for inference from detailed contact tracing data (dashed line) or more readily available chain size data (solid line). B) The specificity for detecting a statistically significant change in (as compared to 1980s monkeypox contact tracing data) is shown when various values of are assumed during chain size analysis (as applied to the same chain size data simulated for panel A). The specificity is defined as the probability that a change is not detected for an assumed value of conditioned on our gold standard for a lack of change (e.g. a change is not detected when is allowed to be a free parameter during inference). The solid line corresponds to chains and the dashed line to .