Labels each cluster by its likely transmission route, from the earliest collected isolate among all patients in the cluster (the "index isolate") and whether the cluster's other isolates have an overlap explanation. The categories are:
"patient-to-patient": if the index isolate is admission-positive and there is overlap explanation for all other converts in the cluster.
"weak-index-patient-to-patient": if the index isolate is not admission-positive but is the first surveillance for the patient after admission and there is overlap explanation for all other isolates in the cluster.
"missing-intermediate": if the index isolate is admission-positive but at least one other convert in the cluster has no overlap explanation.
"false-negative-index": if the index isolate is not admission-positive but there is overlap explanation for all other converts in the cluster barring one (deemed to be the false negative index).
"missing-source": if the index isolate is not admission-positive and there is no overlap explanation for more than one convert in the cluster.
"multiply-colonized-index": if the index isolate is not in the cluster but is admission-positive, this is a "multiply-colonized index" if there is overlap explanation for all other isolates in the cluster.
"multiply-colonized-index-missing-intermediate": if the index isolate is not in the cluster but is admission-positive, and at least one other convert in the cluster has no overlap explanation.
"all-admission-positive": if all isolates in the cluster are admission positive, this is a special category that may indicate a common source or multiple colonization events, but is not explained
"inexplicable": catch-all category for cases that do not fit into the other categories. Currently, index and overlap explanations are not enough to explain these clusters.
Arguments
- isolate_lookup
A lookup table for isolates and their cluster assignments with other relevant epidemiological information. See
get_isolate_lookup().- cluster_overlap_df
A data frame of per-isolate cluster overlap, from
cluster_isolate_overlap().- surv_df
A data frame of surveillance cultures, with columns
patient_id,genome_id,surv_dateandresult.