View table: Discussions

• After 24 rows, open and closed assigning Pseudo Phin an error message pop up then I cannot assign Phin anymore. I closed CFE then open then I am good to go again. The second time the error message pop up not 24 rows it less than 24 like 15 rows then it will show again the error message. PTorres 16:09, 2022 June 14 (CDT)
  • Is it the "cannot open..." error or the "enter parameter..." error you get at this point? Emailed Pagasa... Ttenbergen 10:49, 2022 November 16 (CST)
    • Still shows "cannot open "so I clicked ok then it says run time error 3014 cannot open any more tables. Closed the CFE then log back in.
    • After I continue assigning Pseudo Phin after 25 rows "cannot open" showed up again I clicked ok then error message shows again. Closed CFE then log back in.PTorres 14:42, 2022 November 30 (CST)

• • • • Are you following the steps in Generating PseudoPHINs when this goes wrong? Which step in those instructions are you at when it stops responding and you need to restart the program? Ttenbergen 13:48, 2022 November 29 (CST)
    • Yes, I am. I am clicking and assigning the new Pseudo Phin and not moving. PTorres 14:47, 2022 November 30 (CST)

• Assigning Pseudo Phin or working on the queries if I worked long enough opened closed it then the error message pop up "Enter Parameter Value". PTorres 17:01, 2022 April 14 (CDT)
  • What is the specific action or button press after which this happens? I.e. which step in Generating PseudoPHINs? Ttenbergen 10:49, 2022 November 16 (CST)
    • When I click the pseudo button to the left of the PHIN field label, nothing is moving.PTorres 15:23, 2022 November 30 (CST)
      • I don't understand what you mean by that. What is the last thing you do before the "Enter Parameter Value" error happens? As in, what is the last button you click or last field you enter? Ttenbergen 12:15, 2022 December 7 (CST)

• Identified as something we should do to streamline data collection. I have made this page to document progress toward this import. Blood gas data is in DSM listing; need to compare to see if we can use it
  • Allan will revisit with Lab people whether this is obtainable now Ttenbergen 11:34, 2022 February 9 (CST)
    • Allan is negotiating inclusion of Labs data with Grace hospital as per Task_Team_Meeting_-_Rolling_Agenda_and_Minutes_2023#ICU Database Task Group Meeting – April 19, 2023, leaving this flagged to ensure update. Ttenbergen 14:00, 2023 May 17 (CDT)

• Do we ever include the age in the Charlson score? Should we? If not, we need to clearly document that, since others looking at this score would expect us to use the original definition. Ttenbergen 11:03, 28 January 2025 (CST)
  • No, the one I had provided with label =Charlson Score has no Age , just sum of all categories. The APACHE SCORE for sure has the AGE score and not Charlson Score. --JMojica 11:26, 28 January 2025 (CST)
• Did we include age as part of the data used to generate the Overstay model? We need to use whatever we used for that in our score and generator.
  • The problem is I did not see the final overstay model and I'm only relying on what was on the WIKI Overstay Predictor Project Algorithm showing the AGE and AgexADLmean. In Rodrigo's model, Age is included. I would presume Leigh Anne also included it.--JMojica 11:26, 28 January 2025 (CST)
    • I just re-read your original answer, and it actually addressed my question: any Charlson Score you gave them did not include that age factor. I am pretty sure the official Charlson score does include an age factor, I implemented it in CCMDB (not CFE for some reason) "L_ICD10_Charlson_Agepts" query. We had this conversation before, but then I didn't document what was the reason to omit the age from the score. It makes a significant difference to the score, so whoever uses our data (eg is it included in our dump to MCHP) would expect it to include the age portion. Ttenbergen 18:08, 29 January 2025 (CST)

• Do we need to consider 1010 Sinclair separately as well?
  • 1010 sinclair is an independent apartment living see https://tenten.mb.ca/housing/1010-sinclair-street/. so I would think this might be home or community facility with support? Lisa Kaita 12:49, 10 March 2025 (CDT)
    • 1011 supports ventilated patients (or used to anyway, so that's pretty heavy support... Ttenbergen 17:54, 10 March 2025 (CDT)

• Any others we might need to consider separately?

• Discussed this at JALT Meeting - Rolling Agenda and Minutes 2025#JALT 2025-03-11 but I don't remember if we came to an answer or next step. Just found a note to add that we will also need to decide if any of these are in-patient locations. This would make them collectable as Pre-admit Inpatient Institution, and is relevant as per Pre-admit Inpatient Institution field#Data Use / Purpose.

• These should be made consistent with Template:PCH Riverview Deer Lodge and should probably use that template to enforce consistent use. Ttenbergen 14:23, 2024 October 16 (CDT)
  • and I have lost track of the details Ttenbergen 10:45, 2024 November 1 (CDT)
    • There might be an easy way out of this question. We now only have one awaiting code; if waiting for any of them qualifies for this code then we may not need to know about the level of care. But then again, some of those sites also have acute/LAU type settings so we may need to figure this out after all...Ttenbergen 20:27, 2024 December 11 (CST)

Here is the breakdown

• • Riverview-
    • PCH- C and D wings floors 1-4 42 beds
    • PCH- Aand B wings -locked unit for alzheimer/dementia 30 beds
    • 2E is LTV 30 beds
    • 3E is palliative care 30 beds
    • 4E is stroke rehab 30 beds
    • 4W is Acute brain injury- 10 beds
    • 3W is PCH- 29 beds
    • 2W respiratory ward (no vents)- 30 beds
  • Deer Lodge
    • Rehab wards are Lodge 2W, 4E and 4W
    • TCU is Lodge 2E
    • chronic care is Lodge 5E, 5W, 6E, 6W, 7E and 7W (clients with comples medical issues who can't be cared for in PCH ie. tracheostomy, ostomy, enteral and parenteral nutrition, complex wound care, blood transfusions and PIV and Central lines.
    • PCH is Tower 3N, 3S, and 4,5,6,7 and also Lodge 3W, 3E

Lisa Kaita 13:03, 10 March 2025 (CDT)

• Lisa, any idea why this might be happening from a collection perspective? Is the problem in what we receive the right info from Cognos, or is there an issue in the collection process that makes this happen? Ttenbergen 09:52, 2024 December 5 (CST)
• Julie, do we need to flag anywhere further that this data is suspect? Ttenbergen 09:52, 2024 December 5 (CST)

• Do we need similar instructions or mechanisms as for MRSA for this one? eg coding for Resistance to carbapenem antibiotics? Ttenbergen 11:18, 18 February 2025 (CST)
• I'm not clear on what you are asking Lisa Kaita 20:57, 10 March 2025 (CDT)
• Sorry, I meant the "set MRSA Colonized" button, and something like Template:ICD10 Guideline MRSA for how we code it when it's an actual infection rather than just a colonization. Ttenbergen 00:51, 11 March 2025 (CDT)

Just a placeholder for now because the idea of how we define dispo to community (or for that matter, re-admit Previous Location) in data came up re. things like Readmission to MedWard and others. We have the obvious "Home" but if someone is discharged to something like Dialysis, would that also count? How do we define? Ideally by a column in s_dispo table such as s_dispo.loc_type, but that one uses "non-patient" which it also uses for Deceased patients (should we just split that out?). There is probably even more to this. Likely Julie has more than one approach in reporting. This came up because we were looking to define this for LAU collection readmission data.

• This is actually just as much regarding to admitted from community, so maybe this should just be renamed to "outpatient sites in s_dispo table"?

• This will need to be concistent with the concept of Pre-admit Inpatient Institution field ie they should not overlap, and possibly an entry's inclusion in that (ie s_dispo.inpatient unchecked) could be a way to define this.

If this is actually done then why don't we just make the nursing stations available? We already have s_dispo table entries for them. 35 entries since 2016, 6 in the last 12 months. Are there other things for which this entry is used? Checked for reasons for this decision but they don't appear documented Ttenbergen 19:42, 13 March 2025 (CDT)

I thought we had decided at JALT to collect this as presented by EPR... do I remember this wrong? I had already added it in CCMDB.accdb Change Log 2025#2025-03-11-1. Ttenbergen 22:52, 11 March 2025 (CDT)

• Yes, I saw that, come to think of it I don't think we decided, not in my notes, but we can use it and I will change the wiki instructions Lisa Kaita 11:25, 13 March 2025 (CDT)

That more detailed info included things like HSC Lennox Bell; we discussed whether or not we need to document the details of this. If so, the following is likely a starting point:

• Home
• Home with support
• Institution NOS
• TRSF Group/Supportive housing
• TRSF to Correctional Facility
• any Hospices
• GH-Transitional care
• HSC - Transitional care
• Misericordia - Transitional care
• STB - Transitional care
• VIC - Transitional care
• HSC Lennox Bell
• Manitoba Adolescent Treatment Center - unknown/other

• I have re-updated Created_Variables_Common_maker_2021 query, for some reason the change I had made was not reflected in the master version. Ready to test. Ttenbergen 13:25, 2022 June 28 (CDT)
  • emailed Tina some inconsistencies found in ER Delays Aug 15,2022. --JMojica 13:21, 2022 August 29 (CDT)
    • are these still an issue? Ttenbergen 11:15, 2024 May 1 (CDT)
      • I will re-check again. Can't remember if have been resolved. --JMojica 14:40, 2024 October 2 (CDT)

• I just had a look at that sas file (they open as text files) to see how you define transfer delay. If that file is still being used we may have a problem, it still defines tdelay different if a pt goes to a higher level of care, goes AMA or dies, and we changed that some time ago. So is this still the reference of how you calculate this? Ttenbergen 22:50, 2024 November 16 (CST)

• This data is problematic before 2011-Q2,the only reason there are any is because it derives them for EMIPs. Some data may be available in Moves for Medicine, but that would still leave a gap. Ttenbergen 23:09, 2024 November 16 (CST)
  • OK, will be working on this getting data from Moves for Medicine from period Sept 2007 to June 3, 2011. will give to Pagasa for upload to Arrive DtTm field. I will update the WIKI as soon as done. Conclusion: for Medicine, prior Sept 2007, no ER delay while for Critical Care, no ER Delay prior July 1, 2016 -- these are treated as missing. --JMojica 16:35, 2024 December 11 (CST)

• I have a hunch that this error may be related to a send failing for one collector and a second collector then also sending, leading to problems with the lock file.

So, when the error happens, please have a look at Regional Server\​/output\\ad.wrha.mb.ca\WRHA\REGION\SHARED\ICU_DATA_COLLECTION\/output and see if there is a second centralized_data file there with an extension .lccdb. See Lccmdb and ldb files for details. Please log here about what you find. Ttenbergen 10:41, 2024 November 30 (CST)

• Thanks for updating, but I don't understand... I thought the whole idea with having them in L_Problems_D_ID would be that that is the actual data used to do the translation from one ID to the other. Do you mean you actually hard-code it in SAS instead? I ask because you say you need to modify the SAS program to incorporate the change. Ttenbergen 21:47, 14 March 2025 (CDT)
  • That is correct, I add a line in my SAS code to change the unmatched D_ID between the DSM dumped data versus the L_LOG D_IDs. Pagasa was instructed to list down in the L_Problems_D_ID those D_ID she had changed during her data processing task if these D_IDs had already been for DSM request. There will be no entry in the L_Problems_D_ID if any change in D_ID happens before sending a DSM request. So far, SAS finds zero of such case. If in case, SAS finds any record in the requestor not in L_LOG, that is only the time the SAS code assigns a new D_ID to the dumped DSM D_ID so as to be consistent with the L_Log D_ID. --JMojica 10:38, 17 March 2025 (CDT)

• This means the entries will need to be available as Previous Location and Dispo, and we can't restrict those by site, so they will be available everywhere. Lisa's email mentioned that Nick would want them available to be collected that way anyway. We will need to inform collectors of the change and update Previous Location and Dispo.

• We need to decide if we will use the criteria set out in Readmission to MedWard The criteria may be too restrictive for the LAU project.
  • waiting for Dan to weigh in
  • also see Discharged to community
  • decided not to use the Readmission to MedWard criteria instead use criteria of readmission within 7 days of previous inpt admission (any inpt location) or readmission within 7 days after current admission.

• We'll need to fill these in as they become available; that may well be as links to other documents, since the unit documentation would not be on this wiki. If files are generated we should link to them at their home location, but also maintain files on this wiki, since link rot is inevitable for this sort of thing. Ttenbergen 15:45, 23 February 2025 (CST)

ER Delay and Overstay

• Need to make sure I do things this way and include this in the right pages, possibly Change from Service Location to Service, Boarding Loc and Transfer Ready DtTm tmp entry but a lot of other pages touch on this, so maybe it needs to be a template? It may not be consistent everywhere:

Note from email from Julie 2024-20-24: I do not use any cut-off e.g. 2020-10-15 because the tmp has been populated by boarding data since July 2018 for ICU and Sept 2019 for Med. The cut-off Oct 2020 Patient Follow only applies for tmp Service but not for tmp Boarding loc. By the way. the patient follow started early at Grace Oct 1, 2020 (they piloted it), while HSC and STB started Oct15, 2020. Similar with EMIPs, I have two sources of ER-delay namely 1. from Accept and Arrive 2. from Tmp : first boarding dttm and second boarding dttm. If both have values, I use the one from the tmp source.

• Do we want that to be the first time during a medicine program admission, or only the first time during a hospitalization? Ttenbergen 08:45, 26 February 2025 (CST)
• Also, we define Transfer Ready DtTm tmp entry to be able to do this per level of care to enable reporting separately for HOBS. I think there are problems with that concept since I doubt users of the data keep the detailed definition in mind, but we still need to decide if we want to be able to report this the same way. If so, we need the the same connector integer in the instructions. Ttenbergen 08:51, 26 February 2025 (CST)
  • emailed DR about this Ttenbergen 11:36, 28 February 2025 (CST)

• How do we want to define where to get this? Unless we get physicians to put this information into a specific spot we are right back to all the definition difficulties in Transfer Ready DtTm tmp entry, only it will be worse because collectors will need to keep conflicting but similar instructions in mind. Ttenbergen 08:49, 26 February 2025 (CST)
  • emailed DR about this Ttenbergen 11:36, 28 February 2025 (CST)

• ADLmean_nh - in the CCMDB Script it is written -- Adlmean_nh = (ADLScore - 12)*(from a nursing home), what does 12 mean?
  • Very good question. The function adl_sc used to get the individual scores just generates a score of 0, 3 or 6, or -1 if given bad input. That doesn't explain why -12. And I can't imagine why this would be. But it's consistent between two functions, so unlikely to be just a finger fumble. Ttenbergen 19:15, 25 January 2025 (CST)
    • looks like ADLmean does not imply average of the 6 individual ADL categories (e.g. total ADL SCore/6) but (TOTAL ADL Score -12). I'm guessing here, maybe 12 is the Grand Mean of total ADL and Leigh Anne is trying to evaluate the deviation of the individual data point from the mean of the dataset of those from Nursing Home . --JMojica 09:29, 27 January 2025 (CST)

• Charlson categories are provided but the CCMDB Script model only includes MI, Pulmonary, Connective, Renal
  • It also contains the full score as part of Charlson_nh. Ttenbergen 19:15, 25 January 2025 (CST)
    • correct, there are selected subcategories and also FULL CHARLSON SCORE for those from Nursing Home. --JMojica 09:29, 27 January 2025 (CST)
  • I think "connective" became "Rheumatic disease" in ICd10, right? Ttenbergen 16:41, 24 January 2025 (CST)
    • Yes it refers to Connective Tissue Disease-Rheumatic Disease. --JMojica 09:29, 27 January 2025 (CST)

• OutsideWPG - I have used Previous Location because Leigh Anne specified the patient came from a facility outside of Winnipeg.
  • MB - I would think Leigh Anne had used the Province because the data set has no Postal Code yet
  • I just checked the query check_from_out_of_town I used to derive this and it only considered Pre-admit Inpatient Institution having geog like "out-of*"! Ttenbergen 16:57, 24 January 2025 (CST)
    • Will ask Dr. Roberts what this was all about to better understand what to use now. Specifically, if we use previous location, would we only care about hospitals, or also e.g. nursing stations, PCH, hospice, "unknown/other" loc_types. Ttenbergen 16:51, 24 January 2025 (CST)

• I think this is the process where you said you are having problems with copy/pasting. Copy pasting isn't even mentioned here, so maybe update the process to show how you actually do this, so that someone like Sheila Rusnak would be able to follow the instructions. Ttenbergen 15:45, 2022 March 17 (CDT)
  • Do you create that query each time? Would we be able to update the NDC query that finds these in the first place to include the info you need? We can discuss at our next meeting. Ttenbergen 16:23, 2022 March 17 (CDT)
    • Taking this off Pagasa's list for now, since if we can get this data from DSS we won't need to do this any more. Ttenbergen 15:36, 2022 March 24 (CDT)

• We should set up a Record definition query in CFE to provide a clear way in CFE to treat historical, pre PatientFollow Project, records in a way that is consistent with our current definition. This is a good idea but not trivial, so we should decide how badly we need this, but definitely document the concept.
  • JM email from 2025-03-18: "Perhaps, creating a query to link those records from old definition to become similar to patient follow is needed in ACCESS to generate at least the LOS, first admit date and last discharge date, first previous location and last dispo location. In this way, SAS will not generate it and I just get it from ACCESS query. Is this hard to do? "
    • The query is probably the easiest part, but this would take some analysis to figure out what all is actually affected by this, ie which other queries would need to be updated. Having said that, we should still probably do it. I don't know who would be in a position to prioritize this, I don't have bandwidth for this as a side project right now. Don't think it's a task question because I don't know if anyone there would have anything to add to this. Ttenbergen 14:17, 18 March 2025 (CDT)

• What would be the priority for this?
• Technical because governance and permissions are a separate issue we would need to address; not SH's, but our service providers for any solution.
• Let's talk about this; this is where SH data will live, and if we can't interact with it in the future the utility of our data will be greatly reduced. Ttenbergen 14:33, 17 March 2025 (CDT)

• I don't think you write from SAS, right? Ttenbergen 14:33, 17 March 2025 (CDT)
  • I export the ACCESS table using SAS and then perform all the analysis by programing in SAS. --JMojica 13:35, 27 March 2025 (CDT)
• I made this a (3) because the ability to export above would provide a work-around of copying data to local; open to discuss. Ttenbergen 14:33, 17 March 2025 (CDT)

difficult to say, right now we have wifi access everywhere, used to be sketchy at the death desk but we no longer are able to review charts there, SBGH does not have wifi in the basement where med records is located would we be able to use the database offline? like we currently do Lisa Kaita 20:59, 23 March 2025 (CDT)

• in s_dispo table HSC Lennox Bell is listed as loc_type=unknown/other ; many in that location would be most similar to "home", so should these be grouped as "non-patient" instead? Wherever we deal with "home", are we already including these? For example, for Readmission to MedWard, would we include anyone who has been discharged to a "non-patient" location, or do we use more explicit groupings? Ttenbergen 15:24, 11 March 2025 (CDT)
• also in s_dispo table HSC Lennox Bell is listed as site=HSC; it kind of is, but wherever that column is used to group, would this be a location we would expect in that group? Or should it be blank? Ttenbergen 15:24, 11 March 2025 (CDT)

• I have linked this in from several reports based on those having "L2" in them. There might be others where I just don't know that they also use this building location because they word it through stb icu or cardiac unit or something. It would be worth updating and linking those reports so that, when a location related to them changes, we will be easily able to find it on the wiki and know the underlying SAS might need updating. Ttenbergen 12:35, 20 March 2025 (CDT)

• While discussing Visit Admit DtTm differences within same admission at JALT Meeting - Rolling Agenda and Minutes 2025#JALT 2025-03-11 I realized we don't have any part of your "cleaning" process documented. We should, even if it is a rudimentary notice of the SAS files you use and what you check for. Ttenbergen 21:51, 11 March 2025 (CDT)
• If there is linking beyond Populate linking pairs, or if you use a different linkage, we need to document that as well; do you? Ttenbergen 21:51, 11 March 2025 (CDT)

Moved here from Service tmp entry to avoid duplication, will need validating and integrating into this page:

• at StB, the Service tmp date and time for ICU patients coming from ER is taken from the orders. If the time is written after the patient has already arrived to the ICU, the Service tmp date and time is the same time as the Boarding Loc time
• if the patient is coming from the Cath Lab, or the OR, the time is obtained from the ICU flowsheet

• met with Andie, she would be happy if we could start as a temp project for 3 months, would like the breakdown to include, back/coccyx, heels and other for all of the decubitus ulcer stages, need to add the new ICD -10 dx deep tissue injury as well Lisa Kaita 10:28, 12 March 2025 (CDT)
  • If this is to be temporary, would it make more sense to put these into tmp entries? That would also mean we can make them exactly what they want, without any limitations of what ICD10 gives us. Ttenbergen 02:45, 13 March 2025 (CDT)
• that is what I had in mind, I will send her the ICU request form Lisa Kaita 06:02, 20 March 2025 (CDT)

z "It was decided that Allan with contact Dr. Embil after COVID is over and see if we can obtain this data from Infection Control. If so, we could import it into the database, and have our data collectors cease obtaining it." - did anything come of that?

• I have sent an email to John Lisa Kaita 20:49, 10 March 2025 (CDT)