Direct Data Access for RIS/PACS: Difference between revisions

From CCMDB Wiki
Jump to navigation Jump to search
← Older editNewer edit →
differences in counting imaging labs
Line 1: Line 1:
== Potential Change ==
== Potential Change ==
{{Potential Change}} We could stop having collectors enter imaging data available from IMPAX onto their laptops and use the combined data from dump and from imaging.mdb instead.  
{{Potential Change}} We could stop having collectors enter imaging data available from IMPAX onto their laptops and use the data from dump instead.  


To do that we would need to:  
To do that we would need to:  
* import all data received from PACS and not yet imported
* confirm that the data is of comparable quality
* confirm that the data is of comparable quality
* decide how to store this:
* come up with an import process (programming is almost done)
** if storing in Ed's DB, confirm that we have all required data to output the same way Delphic does, and that Ed can import our data
* ensure that we can generate output to support whatever this data is used for now (what is it used for? might want to put that into [[Lab Collection Process|Labs]]).
** if imaging.mdb is sufficient, ensure that we can generate output to support whatever this data is used for now (what is it used for? might want to put that into [[Lab Collection Process|Labs]]).


== Direct access to data (work in progress)==  
== Direct access to data (work in progress)==  
Line 14: Line 12:
=== Data Acquisition - RIS/PACS ===
=== Data Acquisition - RIS/PACS ===
Breanne Batters from the RIS/PACS team will provide us with a monthly report with columns (hospital|unit|phin|chart|image_DateTime|imaging_type).  
Breanne Batters from the RIS/PACS team will provide us with a monthly report with columns (hospital|unit|phin|chart|image_DateTime|imaging_type).  
This data will get imported into [[Regional Server]]\output\'''imaging.mdb''', table name '''images'''.
This data would get imported into [[Centralized_data.mdb]], details to be determined but possibly into [[L_Labs_Flowsheet table]] where it is stored now.
I have reports for July and August, but none after that... emailed Breanne [[User:Ttenbergen|Ttenbergen]] 16:28, 18 November 2011 (CST)


=== RIS/PACS ambiguous types ===
=== Importing ===
Things like "CT ANGIO ABDOMEN COMBINED" or "CT ANGIO ABDOMEN COMBINED" may be in the table s_aggregate_types twice to be counted as both Angios and CTs. {{discussion}} Is that how collectors would count it? Ttenbergen 15:37, 2015 August 19 (CDT)
Building importer for this, documentation pending. {{discussion}}
{{discussion}}
* No, we only count arterial angios done by arterial access in our lab counts as anglos. CT angios are counted as CT's in the lab counts.
** is that done consistent across town? Already identified with Laura that there might be inconsistencies in how some of these imaging items are counted, we are hoping to discuss at task meeting. Ttenbergen 09:28, 2015 August 24 (CDT) {{Discuss@task}}


=== Data Acquisition - Collectors ===
== Things still being worked out ==
Additionally, during [[sending]], '''complete patients' only''' (hosp|loc|serial|phin|chart|admit_datetime|disch_datetime) will be sent to the table '''patients'''; this information is required to link images to our database.
=== deciding on types ===
* hopefully fixed with update [[CCMDB.mdb_Change_Log_2011#ver_2011-11-18]] [[User:Ttenbergen|Ttenbergen]] 16:28, 18 November 2011 (CST)
There seem to be new types of images on every mew import. They get less but this will probably continue. To be able to categorize them they need to be added to table s_aggregate_types and categorized manually the first time they are encountered. This means it would probably be best if Trish or a collector did the imports. They take about 5 minutes once a month.


=== combining ===
==== RIS/PACS ambiguous types ====
combine patients and images: see [[Imaging.mdb]] query '''combiner'''
Things like "CT ANGIO ABDOMEN COMBINED" or "CT ANGIO ABDOMEN COMBINED" may be in the table s_aggregate_types twice to be counted as both Angios and CTs. {{discussion}} Is that how collectors would count it? Ttenbergen 15:37, 2015 August 19 (CDT)
 
* No, we only count arterial angios done by arterial access in our lab counts as angios. CT angios are counted as CT's in the lab counts.
=== Aggregating ===
There will be one line per image, so we will need to aggregate the lines into counts image_type. This will be done by a query '''combiner_pivoter''' which will provide a count for lines of the same combination of (Serial/PHIN /MRN/location/admit_dt_tm/discharge_dt_tm/imaging_type).
The query will start to pick up records from the image file only once a line is present in the patients table, i.e. after a patient has been sent.
Further, the query will of course only pick up images that have been imported.


=== Where to store ===
=== HSC and STB data problem ===
Data can be stored in the imaging.mdb program. I think it would be a long time before we would hit any size restrictions with this, and by then we will hopefully have moved on to a different data repository that can handle the volume.
We didn't get much for HSC/STB, email is out to Breanne Batters to find out more


There is another option for long-term storage. We are currently importing some values for HSC into TMSX. Do we have any documentation for the importer? If I stream the reports from RIS into the right format, would we be able to import into Ed's program directly? [[User:Ttenbergen|Ttenbergen]] 11:30, 4 August 2011 (CDT)
=== comparison ===
*In the Labdownloading program, the study and chart#, admit datetime and disch datetime are entered manually by batch to download the lab dump file from New Delphic. Then Ed's interface program picks up the corresponding match, sums it up and upload to the master database.  I don't think we have the documentation of the importer. [[User:JMojica|JMojica]] 12:13, 4 August 2011 (CDT)
review queries 5_compare_* with Trish


=== Backups ===
== Backups ==
eHealth backs up the server, so we should not have to back this up.  
eHealth backs up the server, so we should not have to back this up.  
Considering that we could get new dump files generated when needed we don't need to keep or back those up either.  
Considering that we could get new dump files generated when needed we don't need to keep or back those up either.  
Line 47: Line 37:
[[Category: IT Instructions]]
[[Category: IT Instructions]]
[[Category: RIS/PACS]]
[[Category: RIS/PACS]]
[[Category: Labs]]

Revision as of 08:52, 2015 August 24

Potential Change

Template:Potential Change We could stop having collectors enter imaging data available from IMPAX onto their laptops and use the data from dump instead.

To do that we would need to:

  • confirm that the data is of comparable quality
  • come up with an import process (programming is almost done)
  • ensure that we can generate output to support whatever this data is used for now (what is it used for? might want to put that into Labs).

Direct access to data (work in progress)

We are going to test accessing this data directly from RIS. Once we know it is feasible we will ask the Database Steering Committee if we can stop manual collection.

Data Acquisition - RIS/PACS

Breanne Batters from the RIS/PACS team will provide us with a monthly report with columns (hospital|unit|phin|chart|image_DateTime|imaging_type). This data would get imported into Centralized_data.mdb, details to be determined but possibly into L_Labs_Flowsheet table where it is stored now.

Importing

Building importer for this, documentation pending. Template:Discussion

Things still being worked out

deciding on types

There seem to be new types of images on every mew import. They get less but this will probably continue. To be able to categorize them they need to be added to table s_aggregate_types and categorized manually the first time they are encountered. This means it would probably be best if Trish or a collector did the imports. They take about 5 minutes once a month.

RIS/PACS ambiguous types

Things like "CT ANGIO ABDOMEN COMBINED" or "CT ANGIO ABDOMEN COMBINED" may be in the table s_aggregate_types twice to be counted as both Angios and CTs. Template:Discussion Is that how collectors would count it? Ttenbergen 15:37, 2015 August 19 (CDT)

  • No, we only count arterial angios done by arterial access in our lab counts as angios. CT angios are counted as CT's in the lab counts.

HSC and STB data problem

We didn't get much for HSC/STB, email is out to Breanne Batters to find out more

comparison

review queries 5_compare_* with Trish

Backups

eHealth backs up the server, so we should not have to back this up. Considering that we could get new dump files generated when needed we don't need to keep or back those up either.

Retrieved from "https://ccmdb.kuality.ca/index.php?title=Direct_Data_Access_for_RIS/PACS&oldid=45518"