- User Guide and Online Help
- Retrieve and review data mining results
- View results graphs
- Graphs for 2D results
- View a sector map
View a sector map
- In the left navigation pane, hover on the Data Analysis
icon (
), then click Data Mining Runs.
- Perform one of the following: To open the Select Criteria
page either:
- Click the Row Action menu (
) for the run, and then click View Results.
- Hover on the Data Analysis icon
(
), in the left navigation pane then click Data Mining Results.
The Select Criteria page appears. - Click the Row Action menu (
- From the Run name drop-down list, select the two-dimensional MGPS run you want to view. (Your user preference, Default run determines which run, if any, is selected by default.) You can also click Browse next to the Run name field to Find and select a data mining run.
- Type at least one, but no more than five, drug names.
- Click Choose Graph.
- Select a sector map graph.
- On the
Sector Map page, specify the following display options:
Column Description Subset
Available only if you selected All on the Select Criteria page. The available subsets are all those selected during run creation.
Color controlled by
Name of the statistic whose values will be ranked and used for coloring the sector map tiles. Available values are those computed by the data mining run and include EBGM, EB05, ERAM, ER05, Chi-statistic (signed), and PRR.
Term box size controlled by
The factor that controls the size of PT tiles:
-
Relative importance of term —Bases the size
of PT tiles on an algorithm that determines the
relative public health impact of PTs. Using a recent
version of AERS data, the application computes the
public health impact for a PT as: (number of times
the PT occurs in serious cases) * (proportion of
cases with that PT that are serious-or-fatal). A
higher Public Health Impact score corresponds to a
larger tile in the sector map. Because size is based
on the number of cases of the PT alone (not the
number of cases that have the PT and a particular
drug), the tile size is stable over sector maps for
different drugs.
Note:
For PTs that are new in the MedDRA version associated with the AERS data used to assess public health impact, the application cannot determine the tile size. The application represents these PTs as small tiles. - All terms have equal size —All PT tiles have the same size.
Display maximum intensity at signal score of __
Displays all tiles for which scores are above this value at the maximum-intensity color, as shown in the color key below the graph.
Use gray-scale instead of colors
Use shades of gray instead of colors in the graph.
Show low scores in green for color graph
Check to show low scores in varying shades of green. Clear to show low scores in black.
Note:
This setting has no effect if you display the sector map in gray-scale.Omit rare terms used fewer than __ times in AERS
Number indicating how many times a PT must be in the AERS database (the same one used for Relative importance of term) for that PT to be shown in the sector map.
Note:
The notes for a sector map indicate any rare terms that are omitted because of this restriction.Group by HLT
Within a SOC tile, group in the same rectangular area the tiles for PTs that are in the same HLT. If you also group by HLGT, the tiles for PTs are grouped by HLT within each HLGT.
Group by HLGT
Within a SOC tile, group in the same rectangular area the tiles for PTs that are in the same HLGT.
List __ top scores
Number indicating how many (up to 1000) terms with the top scores are ranked and listed below the sector map. The score is the statistic selected for Color controlled by.
If multiple terms have the same score, the application counts each term separately. For example, you enter 20 as the limit. If the 20th row has a score that other terms after that also have, the application shows those additional terms as well.
Show top score indexes
If checked, the application numbers tiles in the sector map to show their rank, up to the number of scores specified in List __ top scores. For this option to take effect, there must also be a value specified for List __ top scores.
-
Relative importance of term —Bases the size
of PT tiles on an algorithm that determines the
relative public health impact of PTs. Using a recent
version of AERS data, the application computes the
public health impact for a PT as: (number of times
the PT occurs in serious cases) * (proportion of
cases with that PT that are serious-or-fatal). A
higher Public Health Impact score corresponds to a
larger tile in the sector map. Because size is based
on the number of cases of the PT alone (not the
number of cases that have the PT and a particular
drug), the tile size is stable over sector maps for
different drugs.
- Click Display.
- If you point to a PT tile, the MedDRA PT, HLT, HLGT, and SOC appear. The
following information for the combination of the drug and the term that the tile
represents also appears:
- N—Observed number of cases with drug-event combination.
- EBGM—Empirical Bayesian Geometric Mean. A more stable estimate than RR; the so-called shrinkage estimate.
- EB05—A 5% probability value that the true Relative
Ratio lies below it.
Note:
The ERAM and ER05 are available only if the data mining run was set up to compute RGPS. - PRR—Proportional Reporting Ratio. For more information, see PRR computations.
- Chi-statistic—Chi-statistic associated with PRR. (If
EBGM < 1, the Chi-statistic is expressed as a negative number,
indicating that the drug-event combination is more rare than expected).
The Chi-statistic is the signed square root of Chi-square.
Note:
The PRR and Chi-statistic are available only if the data mining run was set up to compute PRR.
- To zoom in on a particular SOC, click anywhere on the SOC tile and select Zoom from the menu.
- If you click a tile for which N is at least 1, you can then drill down to a list of cases for the tile or view statistics for the tile.
- To save the graph as an attachment to a topic, click Save to Topic. (This option is available if Links is checked and the topics feature has been set up.) The graph is attached in a PDF file.
See Work with results graphs for information about links below the graph. For best results, print the sector map graph in landscape mode.
Parent topic: Graphs for 2D results