Task Team Meeting - Rolling Agenda and Minutes 2022: Difference between revisions

From CCMDB Wiki
Jump to navigation Jump to search
Line 58: Line 58:
*A: High platelets are an extremely common laboratory finding.  It's only rarely due to [[Essential thrombocythemia/thrombocytosis]] and instead is almost always a reactive change due to bone marrow stimulation from any number of stimuli.  Reactive thrombocytosis is rarely a clinical issue of importance except in the very rare case that the platelet counts >1 million when blood viscosity may rise leading to thromboembolism, strokes, TIA, and other problems of vascular congestion.  There is an ICD-10 code for "Reactive thrombocytosis" (D75.81) that we aren't currently using and don't want to add since as above it's rarely a clinical problem of relevance.  And our code [[Disease of blood or blood-forming organ, NOS]] ''seems'' like it might work but again reactive thrombocytosis isn't really a disease.  SO, the thrombocytosis not due to [[Essential thrombocythemia/thrombocytosis]] should only be coded at all if it is directly causing a clinical problem (as above) and then code the clinical consequence (such as a DVT) combined with [[Disease of blood or blood-forming organ, NOS]].  
*A: High platelets are an extremely common laboratory finding.  It's only rarely due to [[Essential thrombocythemia/thrombocytosis]] and instead is almost always a reactive change due to bone marrow stimulation from any number of stimuli.  Reactive thrombocytosis is rarely a clinical issue of importance except in the very rare case that the platelet counts >1 million when blood viscosity may rise leading to thromboembolism, strokes, TIA, and other problems of vascular congestion.  There is an ICD-10 code for "Reactive thrombocytosis" (D75.81) that we aren't currently using and don't want to add since as above it's rarely a clinical problem of relevance.  And our code [[Disease of blood or blood-forming organ, NOS]] ''seems'' like it might work but again reactive thrombocytosis isn't really a disease.  SO, the thrombocytosis not due to [[Essential thrombocythemia/thrombocytosis]] should only be coded at all if it is directly causing a clinical problem (as above) and then code the clinical consequence (such as a DVT) combined with [[Disease of blood or blood-forming organ, NOS]].  


13.  Report on the JALT meeting items that occurred prior to today’s Task Group meeting:
13.  JALT issues -- including leftover ones from before, and new ones:
*Apparently it is not uncommon (up to 10-15% in ICUs) for the accept time to be before the prior discharge time.  This doesn’t occur within Cognos, but can occur when collectors alter incorrect accept or discharge time based on chart data.  As the current rule is that collectors only alter such times if they’re off by >30 minutes, most of the resulting “negative” intervals found by Pagasa are <30 mins.   
*Apparently it is not uncommon (up to 10-15% in ICUs) for the accept time to be before the prior discharge time.  This doesn’t occur within Cognos, but can occur when collectors alter incorrect accept or discharge time based on chart data.  As the current rule is that collectors only alter such times if they’re off by >30 minutes, most of the resulting “negative” intervals found by Pagasa are <30 mins.   
**The biggest problem here is postoperative patients with times in OR not being appropriately included in Cognos.
**The biggest problem here is postoperative patients with times in OR not being appropriately included in Cognos.

Revision as of 14:01, 2022 August 24

List of items to bring to task meeting

Add to this by adding the following to the article where the problem is documented:

{{DiscussTask | explanation}}

Toggle columns: Last modified

wiki page question Last modified
wiki page question Last modified
Iatrogenic, mechanical complication/dysfunction, internal prosthetic device or implant or graft NOS At our last TASK meeting the decision was made to exclude spontaneous rupture of an ETT cuff or cuff leak, but we are wondering if this is correct based on what is listed in the includes section of this page and what is in Iatrogenic, mechanical complication/dysfunction, internal orthopedic prosthetic device or implant or graft or bone device and Iatrogenic, mechanical complication/dysfunction, cardiac or vascular prosthetic device or implant or graft, NOS Lisa Kaita 12:37, 2024 March 20 (CDT) 2024-03-20 5:37:32 PM
STB ICUs VAP Rate, CLIBSI Rate Summary
  • IIRC we collected CAM positive (TISS Item) specifically for this, right? If so, can we stop collecting it? And can we make sure a stoppage like this in the future results in reviewing what we collect? Ttenbergen 10:02, 2024 March 20 (CDT)
    • Delirium rate per 1000 days per unit is being reported in the OIT reports. ---JMojica 11:49, 2024 March 20 (CDT)
      • As in Delirium days is reported in Critical Care Program Quality Indicator Report? But that doesn't mention anything about per-1000-days. Ttenbergen 17:00, 2024 March 20 (CDT)
      • The rate is mentioned in the succeeding definition with the delirium days as numerator. Your proposal here is to stop collecting TISS item CAM positive which I disagree because that TISS item is being used and reported as rate in OIT report. Besides, the reason why it was dropped in in the STB VAPCLI report is because the requestor has changed. Brett Hiebert who used to request this was involved in the VAP group and another Delirium group so he asked to have both as one request. Brett had left and the VAP group filled up a new request to continue the VAP data and not on the delirium data. --JMojica 13:58, 2024 March 25 (CDT)
2024-03-25 6:59:38 PM

Also see Task Team Meeting - Rolling Agenda and Minutes 2021

_

_

ICU Database Task Group Meeting – August 24, 2022

  • Present: Allan, Julie, Tina, Lisa, Mailah, Gail, Pam, Pagasa, Val, Brynn, Michelle
  • Minutes prepared by: AG
  • Action items in BOLD

1. Allan reported having heard nothing more yet about Shared Health helping us move hosting by Digital Health forward.

2. Update on moving to automated identification of the five APACHE II comorbid conditions. Status is Julie to email Allan a final set of questions, before Julie implements it, backdating to 1/1/2019.

3. Regarding creating new table containing information (metadata) about individual medicine wards (including High Obs). Tina reports no new progress.

4. Update regarding the universal, unique patient identifier in Cognos, ClientGUID. Tina reports she is working on it.

5. Regarding transfer ready issues -- everything is done except some decisions about reporting. Julie, Allan, Tina to arrange a meeting to finalize that.

6. Update on new lab listings, which appear to duplicate prior listings. Nothing new to report, Allan to discuss with Dr. Sokoro.

7. Regarding the new decision (see July 13 minutes) to NOT allow someone with CRF Stage 5 not yet on dialysis to also be coded as having acute renal failure either as an ICD10 code or as regards APACHE II.

  • After further discussion it was agreed not to implement any cross-checks around this. Thus this should now be disseminated to all collectors and implemented.

8. Discussion about when to use Post COVID-19 condition vs. Past history of Covid-19 infection

  • It was clarified that Past history of Covid-19 infection is a comorbid diagnosis that we use to keep track of people who have had COVID-19 in the past, but it's currently inactive. While Post COVID-19 condition is used to indicate an ongoing sequelum of having had COVID-19 in the past (e.g. pulmonary fibrosis, brain fog, etc).

9. Discussion about whether we still want/need to code the Previous Service field, in light of the fact that since Randy Martens retired nobody has asked for this information.

  • We agreed that before removing it from collection, Allan will ask Bojan if he wants such info (email sent 8/24/22).

10. Other new questions:

  • Q: How should we code VEXAS Syndrome (Vacuoles E1 ubiquitin UBA1 gene on the X chromosome Autoinflammation Somatic mutations). It's a rare, predominantly-male, newly-described (in 2020), genetic abnormality that can cause inflammation in numerous different places along with a variety of hematologic abnormalities. It's diagnosed with genetic testing.
  • Q: Should we, and if so how should we, code elevated platelet counts not due to Essential thrombocythemia/thrombocytosis?
  • A: High platelets are an extremely common laboratory finding. It's only rarely due to Essential thrombocythemia/thrombocytosis and instead is almost always a reactive change due to bone marrow stimulation from any number of stimuli. Reactive thrombocytosis is rarely a clinical issue of importance except in the very rare case that the platelet counts >1 million when blood viscosity may rise leading to thromboembolism, strokes, TIA, and other problems of vascular congestion. There is an ICD-10 code for "Reactive thrombocytosis" (D75.81) that we aren't currently using and don't want to add since as above it's rarely a clinical problem of relevance. And our code Disease of blood or blood-forming organ, NOS seems like it might work but again reactive thrombocytosis isn't really a disease. SO, the thrombocytosis not due to Essential thrombocythemia/thrombocytosis should only be coded at all if it is directly causing a clinical problem (as above) and then code the clinical consequence (such as a DVT) combined with Disease of blood or blood-forming organ, NOS.

13. JALT issues -- including leftover ones from before, and new ones:

  • Apparently it is not uncommon (up to 10-15% in ICUs) for the accept time to be before the prior discharge time. This doesn’t occur within Cognos, but can occur when collectors alter incorrect accept or discharge time based on chart data. As the current rule is that collectors only alter such times if they’re off by >30 minutes, most of the resulting “negative” intervals found by Pagasa are <30 mins.
    • The biggest problem here is postoperative patients with times in OR not being appropriately included in Cognos.
    • We decided that: (a) the crosscheck for such “negative” intervals will be altered by Tina to only highlight those that exceed negative 30 minutes, (b) when those occur Pagasa will not just check them against EPR and change them based on that but instead query collectors for “the truth”, and (c) hoping that ‘a’ will greatly reduce the # of such crosschecks, we will have Pagasa report on this issue in about 6 weeks.
  • Julie reported that sometimes the Visit Admit Date/time in Cognos is DIFFERENT for records included within a single hospitalization. This is clearly an error. To figure it out, the next few times Julie sees such a problem she’ll forward them to Tina, who’ll analyze the reason(s). This may be an ADT/Cognos error and it may require reporting it to be fixed to Charity.
  • There was discussion about ambiguity in the ICD10-based algorithms Allan created for the APACHE II diagnoses. Allan will check them out.

14. Next Task meeting September 7 at 11 am

ICU Database Task Group Meeting – July 13, 2022

  • Present: Allan, Julie, Tina, Lisa, Mailah, Gail, Pam, Pagasa, Barret, Val, Gladys, Pam
  • Minutes prepared by: AG
  • Action items in BOLD

1. Allan reported that he asked Bojan today to ask Perry again about Shared Health helping us move hosting by Digital Health forward. Awaiting more information from Bojan/Perry -- Bojan indicated that Perry indicated at end of June that it’s being looked it “actively” (whatever that means).

2. Update on moving to automated identification of the five APACHE II comorbid conditions. Status is Julie and Tina are working through a final inconsistency related to immunocomprimise, before Julie implements it, backdating to 1/1/2019.

3. Regarding creating new table containing information (metadata) about individual medicine wards (including High Obs). Tina reports no new progress.

4. Update regarding the universal, unique patient identifier in Cognos, ClientGUID. Tina reports she is working on it. - 2022-08-09 now being collected; needs to be added to CFE and sending still.

5. Update regarding rule in template Template:ICD10 Guideline Transplant Failure in relation to ICD10 codes for transplant rejection. Specifically, to eliminate the guideline that says: If you use this code, you should not also code the comorbid diagnosis for past history of the transplant. Julie identified that in the past 3 yrs there have been 354 T86.x codes (i.e. transplant failure or rejection) -- 30 as acquired, 105 as admit and 219 as comorbids.

  • Julie reports the software is made and for the approximately 100 records that need fixing, Pagasa is doing them manually.

6. There was more discussion about the Transfer Delay issues. Julie reports that there are still some differences and some collectors are looking at them. Julie indicated that she has some questions about how to report this information, and it was decided that this will be discussed at the next JALT meeting.

7. Update on new lab listings, which appear to duplicate prior listings. Allan reported that he emailed Dr. Sokoro for a name of someone at DSM who can clarify how these 11 new listing may relate to those labs we currently collect/count.

8. We revisited the question of whether someone with Chronic kidney disease (end-stage renal/kidney disease, ESRD), Stage 5, GFR LT 15 not yet on dialysis should be allowed to also have a coding of acute renal failure.

  • As we previously discussed, such people are perilously close to needing dialysis and if they live long enough all will progress to needing it. Thus, since a majority of such people who have deterioration of renal function don’t have ARF, but rather just progression of their chronic disease, we generally shouldn’t also code ARF in these cases. BUT, this issue not only comes up in diagnosis coding, but also in APACHE II scoring, where the presence of ARF doubles the number of renal points. And we really need these two aspects of coding to be the same.
  • Accordingly, we tentatively agreed to change the rule such that a person with CRF Stage 5 isn’t allowed to have ARF either as regards ICD-10 coding OR APACHE scoring. But there are possibly other crosschecks or codings that this may affect, so before implementing this rule, Tina will check on such potential complications.

9. New question: How to code foot drop?

  • Answer consistent with ICD-10 is to include it as part of Joint disorder, nontraumatic NOS, rather than add the code specific for this entity. Allan altered that Wiki page to indicate that it includes foot drop and wrist drop.

10. New question: If a patient has a UTI with one organism, and then grows another bug in the urine from a later urine culture, should this be coded as a different UTI?

  • Answer: Especially in patients with Foleys, growing bugs in the urine may or may not indicate the presence of a UTI. The way urine colonization is usually differentiated from UTI is from presence of WBC in the urine, and the organism colony counts. Thus the ultimate answer here is that clinical judgement (on the part of the clinical team, and the data collector) is required.

11. New issue: Code for Disorder of vein, NOS (I87.9) lists alternative diagnoses which have their own codes and so shouldn’t be included here. It was requested that Allan identify them, so they can be listed in the Wiki page for I87.9.

12. New issue: Interplay between regional ischemia and the corresponding regional atherosclerosis. Example is Limb ischemia (upper or lower) and Atherosclerosis of arteries of arms or legs

  • Answer goes beyond just the limbs, and should incorporate any regional ischemia.
  • Ischemia is an acute manifestation of arterial problems that can be from atherosclerosis or other causes (e.g. arterial throboembolism, compartment syndrome, physical compression, pressor agents, and other things.
  • Thus an ischemia code should be used if ischemia exists but coding atherosclerosis (which is a chronic condition) may or may not go along with it.
  • Allan has updated the Template:ICD10 Guideline ischemic gut and inserted these ideas into the new Template:ICD10 Guideline Regional Ischemia

13. Report on the JALT meeting items that occurred prior to today’s Task Group meeting:

  • Apparently it is not uncommon (up to 10-15% in ICUs) for the Service tmp entry to be before the prior Dispo DtTm. This doesn’t occur within Cognos, but can occur when collectors alter incorrect accept or discharge time based on chart data. As the current rule is that collectors only alter such times if they’re off by >30 minutes, most of the resulting “negative” intervals found by Pagasa using link_suspect_negative_transit_time query are <30 mins.
    • The biggest problem here is postoperative patients with times in OR not being appropriately included in Cognos.
    • We decided that:
      • (a) the crosscheck for such “negative” intervals will be altered by Tina to only highlight those that exceed negative 30 minutes - done in Centralized data front end.accdb Change Log 2022 #2022-07-13
      • (b) when those occur Pagasa will not just check them against EPR and change them based on that but instead query collectors for “the truth”, and
      • (c) hoping that ‘a’ will greatly reduce the # of such crosschecks, we will have Pagasa report on this issue in about 6 weeks.
  • Julie reported that sometimes the Visit Admit DtTm in Cognos is DIFFERENT for records included within a single hospitalization. This is clearly an error. To figure it out, the next few times Julie sees such a problem she’ll forward them to Tina, who’ll analyze the reason(s). This may be an ADT/Cognos error and it may require reporting it to be fixed to Charity.
  • There was discussion about ambiguity in the ICD10-based algorithms Allan created for the APACHE Acute Diagnoses
    • Allan checked and fixed them, and sent the updated file to Julie and Tina
    • Notes on some choices made in doing this -- so we can remember why it was done this way:
      • For the postop diagnoses which include an admit diagnosis and a CCI code: Reasoning that since these cases must be postop (i.e. preICU location is OR or PACU), the choice was made for procedures only to require the 1st 3 digits (i.e. organ) and not "what was done on it". While this choice may overcount it seems likely that allowing only certain of digits 4/5 would undercount even more.
      • Example for craniotomy for malignancy: our version of CCI doesn't include digits 6/7 (which are the ones which specify craniotomy), so it seems sufficient to require a brain tumor admit diagnosis, patient to be postop, and the organ of the admit procedure done be brain components.
  • Over several meetings we discussed whether or not to include "wasted bed-days in ED while on our service" in the calculations of wasted bed-days, defined as bed-days while in ICU on our service but after being designated as transfer ready to a lower level of care (remembering that level of care while in ED is taken as the level of the service caring for them).
    • There are arguments on both sides of this: (i) is we seek only to count wasted hospital bed days, then we'd exclude ER days, (ii) but on the other hand, in using wasted bed-days as an argument that we need more hospital beds, it would seem relevant to include ER days.
    • Julie did some analysis which showed that: (a) in ICU such wasted bed-days were very few, and (b) while in Medicine they were somewhat more common, they were a small fraction of the total wasted bed-days.
    • Based on this, we decided to INCLUDE such wasted bed-days.
  • Question arose that there are some patients (ICU>Medicine) with lengths of say of just a few minutes. We hypothesized that these are mainly people who die very soon after arriving in ICU or ward. Indeed a quick check of a handful of them showed that all were of this type.
    • Before going further, Julie is going to collate, for 2019-2021, the number of such people, separately in Medicine and ICU, and identify the fraction of them with a dispostion of death. We expect almost all of these are deaths very quickly after admission.
  • Question arose of if/how to include lost/discontinued records in our various reports and analyses. As there are only a handful of these per year, we decided to EXCLUDE them from all reports/analyses.
  • There was discussion about the Minimal Dataset:
    • As the reason for creating this preliminary data is "early reporting", we agreed that: (a) there is no need to go beyond Cognos in obtaining this information, and (b) there is no need to do any data checks on it, or for it to be vetted.
      • Tina will change the Wiki in this regard
      • Lisa will notify collectors
    • We also agreed to add previous location to the Minimal Dataset, and again, to just take for this whatever is in Cognos. If it's blank in Cognos, we'll leave it blank for the Minimal Dataset.

14. Next Task meeting August 10 at 11 am

ICU Database Task Group Meeting – June 22, 2022

  • Present: Allan, Julie, Tina, Lisa, Stephanie, Joyce, Mindy, Brynn, Mailah, Gail, Pam, Pagasa
  • Minutes prepared by: AG
  • Action items in BOLD

1. Allan reported that he asked Bojan today to ask Perry again about Shared Health helping us move hosting by Digital Health forward. Awaiting more information from Bojan/Perry.

2. Change for Apache Chronic to ICD10 from separate variable - Update on moving to automated identification of the five APACHE II comorbid conditions. Tina reported that she is working on this. We agreed that this will be backdated to 1/1/2019, i.e. when we began using ICD10 coding.

3. Location metadata storage - Regarding creating new table containing information (metadata) about individual medicine wards (including High Obs). Tina reports no new progress.

4. ClientGUID field - Update regarding the universal, unique patient identifier in Cognos, ClientGUID.

  • Tina reports she needs to give this more thought.

5. Update regarding rule in template Template:ICD10 Guideline Transplant Failure in relation to ICD10 codes for transplant rejection.

  • We want to eliminate the guideline that says: If you use this code, you should not also code the comorbid diagnosis for past history of the transplant. Julie identified that in the past 3 yrs there have been 354 T86.x codes (i.e. transplant failure or rejection) -- 30 as acquired, 105 as admit and 219 as comorbids.
  • Before moving forward on it and fixing the template Allan and Julie will soon chat about the programming logic to fix these going backwards.

6. Transfer Delay - There was more discussion about the transfer delay issues. Tina and Julie are working on this.

7. We reviewed and dealt with a number of outstanding questions.

8. Update on new lab listings, which appear to duplicate prior listings. Allan will contact Dr. Sokoro at DSM to clarify how these 11 new listing may relate to those labs we currently collect/count.

9. Next Task meeting July 13 at 11 am

ICU Database Task Group Meeting – May 17, 2022

  • Present: Allan, Lisa, Julie, Pam, Tina, Gail, Barret, Val, Mindy, Stephanie, Mailah, Pagasa
  • Minutes prepared by: AG
  • Action items in BOLD

1. Allan updated the group about asking Shared Health to help us move hosting by Digital Health forward. Awaiting more information from Perry.

2. Change for Apache Chronic to ICD10 from separate variable - Update on moving to automated identification of the five APACHE II comorbid conditions. Tina is working on this.

3. Location metadata storage - Regarding creating new table containing information (metadata) about individual medicine wards (including High Obs). Tina is working on this.

4. ClientGUID field - Update regarding the universal, unique patient identifier in Cognos, ClientGUID.

  • Would be convenient for us and we want to use it to replace both pseudoPHIN and PersonID, except that we don’t have it for records entered manually by data collectors.
  • Tina will inquire with Chastity about this.

5. Update regarding rule in template Template:ICD10 Guideline Transplant Failure in relation to ICD10 codes for transplant rejection.

  • We want to eliminate the guideline that says: If you use this code, you should not also code the comorbid diagnosis for past history of the transplant. Julie identified that in the past 3 yrs there have been 354 T86.x codes (i.e. transplant failure or rejection) -- 30 as acquired, 105 as admit and 219 as comorbids.
  • Before moving forward on it and fixing the template Allan and Julie will chat about the programming logic to fix these going backwards.

6. Controlling Dx Type for ICD10 codes - Barret updated the group on his first path through all our ICD10 codes to identify those that should NOT be allowed in the 3 bins (comorbids, admit dx, acquired dx). He’s about 40% through the listings.

7. There was more discussion about the transfer ready issues. There will be a smaller working group meeting on this June 8.

8. Update on new lab listings, which appear to duplicate prior listings. Allan reported that Dr. Sokoro, his contact at DSM, is away until next week, after which Allan will contact him to clarify how these 11 new listing may relate to those labs we currently collect/count.

9. Next Task meeting June 16 at 10:30 am

ICU Database Task Group Meeting – May 4, 2022

  • Present: Allan, Lisa, Brynn, Michelle, Julie, Pam, Tina, Gail, Val, Gladys, Pagasa
  • Minutes prepared by: AG
  • Action items in BOLD

1. Allan updated the group about asking Shared Health to help us move hosting by Digital Health forward. At a Zoom meeting about this on May 2, including Allan, Shelley Irvine Day, others from SH and DH, it was again recognized that we need Perry Gray to take this to the SH Executive. To that end, Bojan Paunovic asked Perry about the status. Perry replied that he will address it this week. It is not completely clear what form this will take, so we’ll have to wait and see. Awaiting more information from Perry.

2. Update on replacing Accept DtTm with first Service tmp entry, and Arrive DtTm with first Boarding Loc. Tina reports this is now essentially completed.

3. Update on moving to automated identification of the five APACHE II comorbid conditions. Tina reported that she is working on this.

4. Regarding creating new table containing information (metadata) about individual medicine wards (including High Obs). Tina reports no new progress.

5. New item. After discussion we agreed that we no longer need the DC Treatment field and will stop collecting it. Tina has removed it from collection

6. New item: Unique patient identifiers.

  • Manitobans have PHIN; non-Manitobans have a pseudoPHIN. In addition, everyone has a PersonID. These latter two are created for the database and thus have no external existence, i.e. could not be used to link our data to other data. Also, it requires both a consistency check and Pagasa to ensure that the pseudoPHIN and PersonID are consistent across the entire database.
  • Cognos has a universal, unique code called the ClientGUID and as such it could be a very convenient replacement for both pseudoPHIN and PersonID, and allow for linkage to other datasets -- except that we don’t have it for records entered manually by data collectors (instead of using Cognos data as the starting point). This latter occurs mainly for all IICU patients (as they’re not on the “patient follow” method of assigning collectors) and for a smattering of others (likely <25/year). We agreed that we will replace pseudoPHIN and PersonID with the ClientGUID if we can secure a reliable method of filling it in for records entered manually.
    • To this end, Tina will inquire with Chastity.
    • If it turns out we can do it then:
      • Tina has in mind a workaround for the IICU patients
      • Tina will implement automated entry of ClientGUID from Cognos -- and then its absence will indicate to Pagasa those we need to fill in manually.
      • Tina/Julie will update the ClientGUID backwards to the start of its existence, and to the extent possible connect it to pseudoPHIN and PersonID prior to that.
  • We’ll discuss this more next time.

7. New item. We recognized the inconsistency of an item in the template Template:ICD10 Guideline Transplant Failure regarding ICD10 codes for transplant rejection. Specifically we agreed to eliminate the guideline to: If you use this code, you should not also code the comorbid diagnosis for past history of the transplant. In order to help guide us whether to try and fix this going backwards, Julie will obtain a count of the number of persons with codes for transplant rejection.

8. Update on new lab listings, which appear to duplicate prior listings. After the regular meeting today, Allan, Julie and Tina discussed these. Allan has identified 11 new lab listings that may (or may not) relate to the lab counts we collect.

  • Tina has sent Allan the official DSM codes for the labs we collect -- and thereafter Allan will contact DSM to clarify how these 11 new listing may relate to them.

9. Next Task meeting May 18 at 11 am

ICU Database Task Group Meeting – April 19, 2022

  • Present: Allan, Lisa, Brynn, Julie, Pam, Tina, Sherry, Gail, Mindy, Val, Mailah
  • Minutes prepared by: AG
  • Action items in BOLD

1. Update on new lab listings, which appear to duplicate prior listings. Tina will send Allan an updated list of these (done) and he’ll contact the lab this week.

2. Update on asking Shared Health to help us move hosting by Digital Health forward. There is a Zoom meeting about this scheduled May 2 including Shelly Irvine Day to talk about it further, but prior to that Allan will ask Bojan to send to Perry a request for an update (done). Awaiting more information from Perry.

3. Updates on two data change issues: (a) moving to automated identification of the five APACHE II comorbid conditions, and (b) replacing Accept DtTm with first Service tmp entry, and Arrive DtTm with first Boarding Loc.

  • The status of making these changes are similar, i.e. that there are many Wiki pages and other data-related issues that have to be identified and modified before going live with these 2 changes. Tina is working on them.
  • Allan reported that after talking with Carmen and Bojan, they desire to have reporting on transfer delays (including out of ED) altered to be as follows:
    • Reporting on all 3 of these items (transfer out of ED, transfer out of ICU, transfer out of ward) to be the same.
    • At today’s meeting we made some decisions, but on April 27 Julie and Allan had additional discussion and these will be done as indicated herein, in which (unlike the prior decision, we will include, not exclude, delays <30 mins)
    • For transfer out of ED delays report as follows: total # of patients who left alive and admitted to hospital; distribution of TRUE transfer delays [i.e. from transfer-ready until they left]; average value of the TRUE delay; cumulative true delays
    • For delays in transfer out of ICU or ward, report both of the following:
      • Total # of patients who left alive to a lower level of care -- among these provide distribution of true transfer delays [from transfer-ready until they left to lower level, considering leaving to lower level without a noted transfer ready date/time as 0 delay]; average value of true delay; cumulative true delays AND also report #of people who had a transfer ready date/time to lower level but in fact went to a higher or same level instead AND also report #of people who had a transfer ready date/time to lower level but in fact died before going anywhere
      • same as above except restricted to those with transfer delays >2 hrs (ICU) and >4hrs (ward)
  • Sherry identified that the ward attendings at Grace are generally not putting a date/time alongside their notes (and orders) indicating a patient is transfer ready to a lower level of care. Allan let Nick and Ebi know this April 21 in hope they can get it to change, so that they can then have more accurate transfer delay data -- Nick indicated he will try and work with them on this.

4. Regarding creating new table containing information (metadata) about individual medicine wards (including High Obs). Tina reports no new progress.

5. Final discussion about how to code if someone has a limb or organ removed during this stay but before they become our patient, in light of the fact that we already have a group of “Past history of” ICD10 codes for this. See April 6 minutes for decisions on this item.

  • Remaining is to modify the guidelines for Admit Procedure. This has now been done.

6. After further discussion, we agreed that instead of creating new location items of hospital-specific procedure locations, we already have a location item ("Other Procedure Location") that we will use for both previous locations and disposition locations. Tina has made this happen.

7. New item: whether to allow/disallow duplication of diagnoses. This arose with a patient who had two separate skin infections during a single admission. After discussion we agreed to allow this, and remove the crosscheck disallowing it. Tina has disabled query check_ICD10_duplicates.

8. New item: when should Liver enzymes, elevated (liver function tests) be used, as opposed to Disorder of liver, NOS? Discussion highlighted:

  • In general, coding abnormal laboratory findings is optional when the specific cause is known and coded.
  • In the absence of knowing the specific cause, the choice of using the code for abnormal lab test versus a nonspecific diagnosis code (e.g. Disorder of liver, NOS) should mainly be guided by whether or not the other coded diagnoses are present.
    • Example: COVID is associated with elevated LFTs in about 30% of cases. Thus if that is the only recognized reason for the elevated LFTs, then it makes most sense to code and link the COVID diagnosis and the abnormal labs together (instead of using the Disorder of liver, NOS code).

9. Next Task meeting May 4 at 11 am.

ICU Database Task Group Meeting – April 6, 2022

  • Present: Allan, Mindy, Gladys, Mailah, Stephanie, Julie, Tina, Val, Pam, Pagasa
  • Minutes prepared by: AG
  • Action items in BOLD

1. Instructions_for_importing_a_batch_of_DSM_Data#Check_for_new_labs - Update on new lab listings, which appear to duplicate prior listings. Nothing new to report.

2. Update on what Shared Health is planning with regards to the databases. Allan reported that he resent to Perry Gray the most recently requested document about this. Awaiting more information from Perry. Tina reported that through the ticket about this she put in to Digital Health awhile ago, apparently they are deciding on who should be the “owner” of the database.

3. Update on agreement to move to automated identification of the five APACHE II comorbid conditions. So, collectors will no longer be specifically identifying these once Julie/Tina implement this change.

4. Regarding creating new table containing information (metadata) about individual medicine wards (including High Obs). Tina reports that Chastity gave her a data sample from Cognos and that Tina/Julie to go over it.

5. More discussion about how to code if someone has a limb or organ removed during this stay but before they become our patient, in light of the fact that we already have a group of “Past history of” ICD10 codes for this. We agreed on the following:

  • As per the existing guidelines, those “Past history of” codes will remain restricted to comorbid diagnoses.
  • In coding an admission which is directly related to a prior procedure:
    • We will expand the guidelines for Admit Procedure. Specifically, to the existing 48 hour rule we will add “OR -- irrespective of whether it's within 48 hours (this could be outside of current hospitalization), if the association of that procedure with the admission is abundantly obvious (as per data collector discretion)”
    • We recognize and accept that since there is no linking of diagnoses with procedures, and in light of restricting “Past history of” diagnosis codes to comorbid diagnoses, such a relationship will be suggestive, not explicit.
  • Example: Patient had a BKA before ICU admission (either before hospital admission, or during this hospital admission), and then develops a wound infection from the amputation site with sepsis. Past history, loss of limb(s) should be a comorbid diagnosis. The CCI code for the amputation should be an Admit Procedure, and the diagnosis would be Iatrogenic, infection, following a procedure or surgery, NOS .

6. More discussion about replacing Accept DtTm with first Service tmp entry, and Arrive DtTm with first Boarding Loc

  • A canvassing of collectors indicates that the Service tmp and Boarding Loc entries are likely MORE accurate that are the Accept DtTm and Arrive DtTm fields.
  • Julie informed the group that per old guidelines, before calculating statistics on discharge/transfer delays, she substracts 2 hrs for ICU and 4 hrs for wards.
    • Apparently, this was in order to account/adjust for procedural delays (e.g. room cleaning, transport, etc) in moving patients.
    • As many/most of those sorts of delays ARE modifiable, it’s not clear if this practice should continue.
    • Allan queried Bojan Paunovic, Carmen Hrymak and Ebi Renner to ask about how they prefer this in the future. The consensus answer is to report both:
      • (a) total # of patients who left alive; the distribution of TRUE transfer delays [i.e. from transfer-ready until they left]; and the average value of the TRUE delay
      • (b) same as 'a' except restricted to those with transfer delays >2 hrs (ICU) and >4hrs (ward)
    • Tina will implement an automatic data check (soft check) for such delays <30 mins, and in calculating the statistics on such delays, Julie will ignore them if <30 mins.
  • Once all this is done, we will make the switchover, ceasing to collect Accept DtTm and Arrive DtTm fields.

7. Further discussion about recording Drug-induced immunosuppression, particularly as part of the APACHE comorbidity

  • As per before, this is now going to be collected exclusively under ICD10 code Z92.25 Past history of immunosuppressive drugs or corticosteroids. Lisa to notify collectors about this.
  • Of note, there’s no real reason to convert over the prior incarnations of collecting this as a CCI/TISS item, as in any case it exists as a created variable, not a hard-coded variable.

8. Next Task meeting April 20 at 11 am

ICU Database Task Group Meeting – March 9, 2022

  • Present: Allan, Julie, Lisa, Tina, Pam, Pagasa, Barret, Pam, Stephanie, Mailah
  • Minutes prepared by: AG
  • Action items in BOLD

1. Update on new lab listings, which appear to duplicate prior listings. Nothing new to report.

2. Update on what Shared Health is planning with regards to the databases. Allan reported that as requested, he provided Perry with a document (from Tina) a technical description of what is needed. Awaiting more information from Perry.

3. Update on agreement to move to automated identification of the five APACHE II comorbid conditions. So, collectors will no longer be specifically identifying these once Julie/Tina implement this change. See also point (7)

4. Location metadata storage - Regarding creating new table containing information (metadata) about individual medicine wards (including High Obs). Tina/Julie to implement this, but Tina recently recognized that Cognos contains this information and before implementing it using info supplied by collectors, she has asked Chastity about obtaining it from Cognos instead.

5. Question from list: How to code if someone has a limb or organ removed during this stay but before they become our patient?

6. Question from list: Change to replace Accept DtTm with first Service tmp entry, and Arrive DtTm with first Boarding Loc

  • We want to finally retire arrive time and accept time, replacing them with info from the Boarding Loc and Service temp. But Julie notes that we use these data to calculate and report on ER delays, and that in comparing the old with new way methods she has found some inconsistencies. So, before we completely make the changeover,
    • At the next task meeting Julie will show us histograms comparing the two sets of data
    • Lisa will look at the inconsistencies and see if she can figure out why it’s happening

7. Question from list Drug-induced immunosuppression.

  • Tina made the point that the way we’re currently recording this entity for purposes of Chronic Health APACHE is as a TISS element, but that this is awkward and suboptimal.
  • After discussion we agreed to instead collect this information under existing code Z92.25 Past history of immunosuppressive drugs or corticosteroids, and for that purpose Allan has changed the definition of that ICD-10 code, which is applicable to both ICU and wards. Lisa to notify collectors of this. Tina to update the created variables queries to reflect this.

8. Next Task meeting April 6 at 11 am

ICU Database Task Group Meeting – February 17, 2022

  • Present: Allan, Julie, Sherry, Lisa, Tina, Pam, Pagasa, Mindy, Brynn, Barret
  • Minutes prepared by: AG
  • Action items in BOLD

1. Update on new lab listings, which appear to duplicate prior listings. Allan has begun working on this. Nothing new to report.

2. Update on what Shared Health is planning with regards to the databases. Allan reported that as requested, he provided Perry with a document (from Tina) a technical description of what is needed. Awaiting more information from Perry.

3. Change for Apache Chronic to ICD10 from separate variable - We agreed to move to automated identification of the five APACHE II comorbid conditions. So, collectors will no longer be specifically identifying these. Julie/Tina to implement this change.

4. Update on the approximately 200 database questions that have been assigned for assessment but not addressed.

  • Tina reports this list is down to about 120 now.
  • After discussion, we agreed to delay aggressively addressing these until our data collectors come back from their COVID assignments.

5. Location metadata storage - Update on creating new table containing information (metadata) about individual medicine wards (including High Obs). Tina/Julie to implement this. It will require that data collectors notify Julie and/or Tina and/or Pagasa when a specific ward is altered in such ways.

6. New questions:

7. Next meeting March 9 at 1pm.

ICU Database Task Group Meeting – February 3, 2022

  • Present: Allan, Julie, Sherry, Lisa, Tina, Val, Pam, Pagasa, Mindy
  • Minutes prepared by: AG
  • Action items in BOLD

1. Update on new lab listings, which appear to duplicate prior listings. Allan has begun working on this. Nothing new to report.

2. Update on what Shared Health is planning with regards to the databases. Allan reported that as requested, he provided Perry with a document (from Tina) a technical description of what is needed. Awaiting more information from Perry.

3. Update on plan (4/15/2021 Task meeting) to implement automatic coding of APACHE comorbidities

  • Allan is working on refining the ICD10 codes.

4. Further consideration of T19 - Spontaneous breathing via ETT without PEEP/CPAP, with or without supplemental O2 (TISS Item) and T20 - Supp O2 through any device, delivered via nose, mouth, ETT or trach (TISS Item).

  • The original TISS28 item (CCM 24(1):64-73,1996) of supplemental ventilatory support, worth 2 points, is either of: (a) Breathing spontaneously via ETT without PEEP/CPAP, or (b) Supplemental O2 by any method without mechanical ventilation (invasive or noninvasive)
    • So there’s some overlap here (which is OK since it’s a single TISS item) in that spontaneous breathing via ETT without PEEP/CPAP but with supplemental O2 fits into both ‘a’ and ‘b’
  • On 6/15/2021 we changed TISS19 and TISS20 to:
    • TISS19 = Spontaneous breathing via ETT without PEEP/CPAP, with or without supplemental O2 -- e.g. includes trach mask, T-tube
    • TISS20 = Supplemental O2 though any device, delivered via nose, mouth, ETT or trach [which must be without invasive or noninvasive mechanical ventilation]
    • However, at that time we incorrectly said that TISS19 is just 1 point, while in fact it should be 2 points (and TISS20 is 2 points also)
  • There is still overlap between these versions of TISS19 and TISS20, specifically: breathing via ETT with O2 but without PEEP/CPAP fits both -- but this is not a problem, as when one computes the TISS score for a given domain (e.g. respiratory support), only the single item with the highest point score is counted.
  • Allan suggests that we:
    • Clarify TISS20: Supplemental O2 though any device, delivered via nose, mouth, ETT or trach, without invasive or noninvasive mechanical ventilation
    • Modify TISS19: Spontaneous breathing via ETT without PEEP/CPAP, with or without supplemental O2, without invasive or noninvasive mechanical ventilation
    • Ensure that TISS19 and TISS20 both are 2 points

5. Update on the approximately 200 questions that have been flagged for assessment but not addressed.

  • Lisa dealt with approximately 40 of these, and referred approximately 70 of them to Julie and Pagasa.
  • The remainder were either assigned to others, or unassigned. Lisa will categorize them into into 3 groups: Need to be addressed; Can be deleted from further consideration; Unclear.

6. A clarification was made in the Wiki regarding Medical Assistance In Dying (Z51.81). Specifically that it should not be coded in a record if the patient was transferred to a different institution to undergo MAID -- instead it should be coded at that other institution.

7. It was agreed to stop the manual notation made in the Boarding Loc comment field for COVID patients or those on COVID wards.

8. Clarification of coding for iatrogenic pneumothorax.

  • The issue was that in coding iatrogenic PTX, the instructions are to combine: Iatrogenic, puncture or laceration, related to a procedure or surgery NOS along with one of the nontraumatic PTX codes. BUT all of the latter indicated not to use them for iatrogenic or traumatic causes --- and this was WRONG. The relevant Wiki pages have now been changed to say not to use them for traumatic pneumothorax, but that can be used for iatrogenic PTX.

9. After extensive discussion it was agreed to create a new table containing information (metadata) about individual medicine wards (including High Obs). This additional layer will be used to keep track over time of: bed count, ward type (specifically general ward vs. high obs) and other information. We recognize this will increase the work that Julie has to do for reporting on unit-specific information, but it provides flexibility for future. See Location metadata storage for further info. Tina/Julie to implement this. It will require that data collectors notify Julie and/or Tina and/or Pagasa when a specific ward is altered in such ways.

10. An example was raised that a patient went medicine ward --> endoscopy --> OR --> surgery ward; but there was no appropriate Dispo location from the medicine ward. After discussion, we agreed to add to Dispo locations for each hospital a listing such as: HSC, Procedure location, NOS. Tina to implement this; this is also mentioned at Medical_Assistance_In_Dying#Collection_Instructions_for_MAID so any update needs to fix that as well..

ICU Database Task Group Meeting – January 6, 2022

  • Present: Allan, Julie, Sherry, Lisa, Stephanie, Tina, Val, Pam, Mailah, Mindy
  • Minutes prepared by: AG
  • Action items in BOLD

1. Update on new lab listings, which appear to duplicate prior listings. Allan has begun working on this. Nothing new to report.

2. Update on what Shared Health is planning with regards to the databases. Allan reported that as requested, he provided Perry with a document (from Tina) a technical description of what is needed. Awaiting more information from Perry.

3. Update on plan (4/15/2021 Task meeting) to implement automatic coding of APACHE comorbidities. Allan is working on refining the ICD10 codes.

4. Update on the decision from 11/30/2021 meeting to revert to the old system, of starting a new record (with new APACHE score and new admit diagnoses) with service changes. This has been done, with relevant changes to the Wiki and “unwinding” of the 10 records that contain such MICU to SICU or SICU to MICU transfers into separate records.

5. Allan reported that he is working on drafting a Data Sharing Agreement that all users of line-level (i.e. individual record) data will need to complete to obtain such data. Draft version is in Database_Request_Process#Data_Accessibility_and_Responsibility. Work is ongoing on this. Also, a decision will need to be made about whether we need a similar agreement for users of cumulative/aggregate data.

6. Discussion about the Minimal Data Set and reporting. We agreed:

  • We will retain the 11 elements of the “true” minimal dataset, all of which come from Cognos.
  • Going forwards, all reporting by Julie will consist of details only from completed records, along with information about what % and # of records for the time interval are incomplete as of the time of the reporting.
  • Lisa will let all collectors know of this change
  • Tina has updated Minimal Data Set regarding it
  • Allan will let Bojan know -- done.

7. There is a new ICD10 code U07.5 Past history of Covid-19 infection which is meant to be used just like the other “Past history of…” codes. Lisa to let all collectors know about it.

8. After discussion we agreed that there is no further need for the Wiki page "HSC Boarding Locations". Tina has deleted it.

9. The question was raised about definition of Emergency Surgery (concept). After discussion we agreed to maintain it as is, i.e. admitted from Operating Room or Recovery AND surgery was classified as E1. Tina has updated wiki Emergency Surgery (concept), Emergency Surgery (TISS Item) and Admit Type for APACHE II to make sure this is defined consistently.

10. There is confusion about TISS elements T19 - Spontaneous breathing via ETT without PEEP/CPAP, with or without supplemental O2 (TISS Item) and T20 - Supp O2 through any device, delivered via nose, mouth, ETT or trach (TISS Item) .

  • This issue tracks back to the 6/15/2021 and 7/8/2021 minutes where indeed the current versions of TISS#19 and TISS#20 were chosen, both as 2 points, in order to correlate correctly to the original/actual delineation of TISS. So, no changes needed for these.

11. Tina identified that there are approximately 200 database questions that have been assigned for assessment but not addressed. Lisa will go over this list and categorize them into 3 groups: Need to be addressed; Can be deleted from further consideration; Unclear. We’ll discuss this further next meeting.

12. Next meeting February 3, 2022 at 11 am.

2021...

Also see Task Team Meeting - Rolling Agenda and Minutes 2021