| Search | |
|---|---|
Survival analysis of herd incidence data
Analyse time-to-outbreak data data using survival analysis methods. Data must include:
- one row for each cluster (farm or village) in the survey;
- a column ("Time") for the time to the last event, either a disease outbreak or censoring time (extent of recall);
- a column ("Censored") to indicate whether the data is censored or not. Use TRUE or 1 to indicate censored data (can't recall last outbreak) and FALSE or 0 to indicate an outbreak;
- a column ("Group") to divide the data into two groups for separate analysis and comparison between the groups; and
- an optional column ("Weight") weightings on the data if RGCS sampling was used.
Note: Columns can be in any order but must be named "Time", "Censored", "Group" and "Weight", as indicated above. The "Group" column is optional if analysing the entire dataset as a single group and "Weight" is not required unless the analysis is to be adjusted for weightings. Download and check out the example data if you are unsure what is required.
Download and check out the example data if you are unsure what is required.
Results are presented as a summary of the data, Kaplan-Meier survival plots (with confidence limits for single-group analyses) and a summary of survival times. A log-rank test is used for comparison of survival times for two groups, if appropriate.