Graphs for 2D results

For two-dimensional results (for example, Drug+Event), the graph types are described in the following table. The graph types are available for MGPS runs. Some are available for logistic regression runs.

Graph Description

Bar graph

Available for MGPS and logistic regression run results. A horizontal bar graph with one bar for each combination. Use this graph type to plot EBGM, PRR, ROR, ERAM, and LROR scores.

Note:

If the run includes subsetting, the graph label includes (for a Single Subset).
  • If looking for link name EBGM, use run type MGPS.
  • If looking for link name PRR, use run type MGPS with PRR.
  • If looking for link name ROR, use run type MGPS with ROR
  • If looking for link name ERAM, use run type MGPS with RGPS.
  • If looking for link name LROR, use run type Logistic regression.

Confidence interval bar graph

Available for MGPS and logistic regression run results. A horizontal bar graph with one bar for each combination, showing confidence intervals based on the EB05-EBGM-EB95, ROR05-ROR-ROR95, ER05-ERAM-ER95, or LR05-LROR-LR95 values. Suitable for studying scores and related confidence intervals.

Note:

If the run includes subsetting, the graph label includes (for a Single Subset).
  • If looking for link name EB05-EBGM-EB95, use run type MGPS.
  • If looking for link name ROR05-ROR-ROR95, use run type MGPS with ROR.
  • If looking for link name ER05-ERAM-ER95, use run type MGPS with RGPS.
  • If looking for link name LR05-LROR-LR95, use run type Logistic regression.

Discrete map graph

Available for results of MGPS runs that use discrete (non-cumulative) subsets. A map graph (discrete) displays a colored grid showing combinations of items (drugs or events) with a drug or event, where each column of the grid reflects the data in a different subset. The subsets were defined as non-cumulative (discrete) in the data mining run.

This graph is suitable for comparing strength of association between different subsets. For example, to compare scores for the individual years of 1998, 1999, and 2000, or for the genders Female, Male, and Unknown.

Link name: Map Graph showing Multiple Discrete Subsets.

Cumulative map graph

Available for results of MGPS run results that use cumulative subsets. A map graph (cumulative) displays a colored grid showing combinations of items (drugs or events) with a drug or event, where each column of the grid reflects the data in a subset. The subsets were defined as cumulative in the data mining run.

This graph is suitable for studying the change in association strength over time. For example, to observe scores for 1998, 1998 through 1999, and 1998 through 2000.

Link name: Map Graph showing Cumulative Subsets.

Sector map graph

Available for MGPS runs that include PT as an item variable and for which you have selected drugs to view results. A sector map graph for data mining results is a visual presentation of data for a particular drug across all System Organ Classes (SOCs). Sector map graphs are available if you select up to five drugs on the Specify Criteria page. The application ignores event selections for the sector map graph.

Link name: Sector Map for Specified Drugs(s) and All Events.

Note:

If the run includes subsetting, the link name includes (for a Single Subset).