DSM data testing: Difference between revisions

From CCMDB Wiki
Jump to navigation Jump to search
← Older edit
m Text replacement - "[[Category: " to "[[Category:"
 
(One intermediate revision by the same user not shown)
Line 1: Line 1:
currently in "DSM labs comparison" on Tina's desktop
We are currently comparing our third batch of one month of test data from DSM.


# open DSM_Labs_Consistency_check
Follow [[Instructions for importing a batch of DSM Data]] to start.
# open module DSM_Processor
# change public const path if necessary
# run Sub DSM_Data_prepper


# run query 1_requester_maker to populate Labs_requester; too slow to do with query
== Queries/Tests ==
# run query 2_labs_reformed_maker to deal with date formats and to split out MRNs etc
We requested data for 463 critical care patients.


== Findings ==
=== 4_1_compare ===
A count per site of comparison between how many records we requested and how many we received at least some data back for.
Site request_count found_count
CON 26 25
GRA 36 36
HSC 142 137
OAK 32 29
STB 194 178
VIC 33 31


We sent a list of patients for CC for one month, they sent use whom they had for the period.
=== 4-2_list admissions w nothing_reported_back ===
A list of records for which we requested but did not receive labs. There were 27 out of 463, 5.8%.


We would not be able to find any tests in their data that we don't have, by definition of how we requested.
Emailed Trish/Gail/Pagasa to check a few to find out if legit or what's wrong. Ttenbergen 18:16, 2015 October 15 (CDT)


We requested 430 CC patients(rest med)
=== 6_1_count_compare ===
Comparison of what is in import and in CFE for the following:
* CBC, PTT, Creatinine, magnesium, sgpt, troponin
Dr Garland suggested these would be a good subset to compare.


=== comparing number of patients ===
=== 6_3_pivot_CBC ===
430 entries in requester, but some are duplicates: .
Comparison of accuracy for CBC across sites
SELECT [1_requester_cc].D_ID, Count([1_requester_cc].D_ID) AS CountOfD_ID
FROM 1_requester_cc
GROUP BY [1_requester_cc].D_ID
HAVING (((Count([1_requester_cc].D_ID))>1))
ORDER BY Count([1_requester_cc].D_ID) DESC;
7 D_IDs show up twice in requester. Those patients have different instances.


====Query 4_1_compare====
=== 6_3_pivot_PTT ===
* 13 - NO results
Comparison of accuracy for PTT across sites
** checked 3, there were labs in our data
** confirmed that at least two have results in EPR
** sent email to Alun Carter to see if he knows what's up
* 392 - results exist but ''not necessarily during stay''
* 383 - have results ''during'' the stay (during column = -1)


(what did I mean by)
=== 6_4_outliers ===
No results for 40 of them.
Listing of records where counts don't match.  


=== comparing lab counts ===
Emailed Trish/Gail/Pagasa to check a few to find out if legit or what's wrong. Ttenbergen 18:16, 2015 October 15 (CDT)
6-2 lab count compare
ONLY 423 ?! not 427?


6-3 pivot
430??


==== HSC lab import issue ====
[[Category:DSM Labs Extract]]
Ran into problems with HSC labs being wrong after import, need to fix before continuing comparison.
Have question out to Julie/Trish how to reconcile that... Ttenbergen 21:16, 2015 July 14 (CDT)
 
 
== For next data request ==
* ask if they can include an Arrived time with the arrived date; expect differences if not. Ttenbergen 13:00, 2015 July 22 (CDT)
 
[[Category: DSM Labs Extract]]

Latest revision as of 10:32, 30 July 2025

We are currently comparing our third batch of one month of test data from DSM.

Follow Instructions for importing a batch of DSM Data to start.

Queries/Tests

We requested data for 463 critical care patients.

4_1_compare

A count per site of comparison between how many records we requested and how many we received at least some data back for. 

Site	request_count	found_count
CON	26	25
GRA	36	36
HSC	142	137
OAK	32	29
STB	194	178 
VIC	33	31

4-2_list admissions w nothing_reported_back

A list of records for which we requested but did not receive labs. There were 27 out of 463, 5.8%.

Emailed Trish/Gail/Pagasa to check a few to find out if legit or what's wrong. Ttenbergen 18:16, 2015 October 15 (CDT)

6_1_count_compare

Comparison of what is in import and in CFE for the following:

  • CBC, PTT, Creatinine, magnesium, sgpt, troponin

Dr Garland suggested these would be a good subset to compare.

6_3_pivot_CBC

Comparison of accuracy for CBC across sites

6_3_pivot_PTT

Comparison of accuracy for PTT across sites

6_4_outliers

Listing of records where counts don't match.

Emailed Trish/Gail/Pagasa to check a few to find out if legit or what's wrong. Ttenbergen 18:16, 2015 October 15 (CDT)

Retrieved from "https://ccmdb.kuality.ca/index.php?title=DSM_data_testing&oldid=166104"