Z I M - Example 2005-02 | February 2008 |
Visualisation of Coding Intensity
|
Wolfram Fischer
Zentrum für Informatik und wirtschaftliche Medizin
CH-9116 Wolfertswil SG
(Switzerland)
http://www.fischer-zim.ch/
|
|
|
|
Introduction |
"Has there been a change in the coding intensity in the course of time?" The answer to this question can help when we judge changes of the DRG casemix index. The graphic presented provides an overview of volumes of codes and changes in the average number of codes per case applied in several departments. |
||
Table 1: |
|||
Source: Fischer [Krankenhaus-Betriebsvergleiche, 2005]: 33. |
|||
Data |
The necessary data comprises the number of medical and procedural codes for each inpatient case during a sequence of years (e. g.: five years from one hospital). |
||
Methods |
A mosaic plot is capable of showing proportions of numbers of several variables. We constructed a plot based on four variables: (1) type of code (diagnosis/procedure), (2) main department, (3) year, (4) number of codes. on the upper ribbon of the plot, we see the coding of diagnoses, on the lower ribbon the coding of procedures. The plot shows the portions of cases in the departments by the widths of the main graphic blocks. Within each block, the change in the number of cases in each department from year to year can be seen by the widths of the vertical strips. Within each strip the portions of cases with one, two, three, or more diagnoses is shown by the height of the rectangles which are coloured from orangered to blue. |
||
Results |
Slow and steady changes as well as irregular coding habits can be detected. For instance, it can be seen in most blocks of the example plot that in most cases,the number of codes increased in the course of time (from right to left). The three eye catching orangered rectangles in the lower left block of the example plot "tell" that the number of coded procedures for cases in the medical department which was generally one during the first two years, changed continuously down to zero by the fifth year. |
||
Discussion |
This kind of graphic is well adapted to data from not too many departments and not too many years. Yet if these numbers increase a great deal, the plot can easily be split up. – When we compare coding habits of several hospitals, the names of the departments could be replaced by the names of the hospitals. |
||
Conclusions |
Mosaic displays that show coding intensities are capable of giving an informative overview on coding habits within a hospital, a group of hospitals, or a whole country. |
||
|
|||
|
References |
||
Fischer W. Neue Methoden für Krankenhaus-Betriebsvergleiche. Ein Werkstattbuch zur Visualisierung DRG-basierter Daten. Wolfertswil (ZIM) 2005: 160 pp. Auszüge: http:// www.fischer-zim.ch / studien / KBV-0506-Info.htm. | |||
Fischer W. Statistische Grafiken zur Beurteilung von Patientenklassifikationssystemen. dargestellt am Beispiel der pädiatrischen Sicht auf das APDRG-System. Wolfertswil (ZIM) 2008: 169 pp. Internet: http:// www.fischer-zim.ch / studien / Grafiken-PCS-Beurteilung-0804-Info.htm. | |||
Friendly M. Visualizing Categorical Data. Cary NC (SAS Institute) 2001: 436 pp. | |||
Hartigan JA, Kleiner B. Mosaics for contingency tables. In: Eddy WF [Ed.]. Computer Science and Statistics. Proceedings of the 13th Symposium on the Interface, New York (Springer) 1981: 268–273. |
Z I M
-
Zentrum für Informatik und wirtschaftliche Medizin
|
Previous Example: | Tables of Contents: | Next Example: |
Hospital Similarity Maps | Example Graphics | Difference Diagram |
Navigations-Tabellen | ||
Z I M - Hauptseite |
© Z I M
Source =
http://www.fischer-zim.ch/graphics/Codes-Mosaic-0802.htm
( latest compilation:
16.01.2014
)