Task Team Meeting - Rolling Agenda and Minutes 2019

From CCMDB Wiki
Jump to navigation Jump to search

Also see Task Team Meeting - Rolling Agenda and Minutes 2018, Task Team Meeting - Rolling Agenda and Minutes 2020

ICU Database Task Group Meeting – December 11, 2019

  • Present: Allan, Barret, Con, Joanna, Julie, Tina, Trish
  • Absent:
  • Minutes prepared by: AG
  • Action items in BOLD

1. Followup about working to reduce collector workload

  • Regarding obtaining CBS TraceLine for all transfusion data. Allan will follow up with Margaret Ring (margaret.ring@blood.ca) and Tony Loewen (anthony.loewen@blood.ca)
  • Regarding obtaining RIS data for radiology tests. Allan will follow up with RIS administrators (Angela Charbonneau 926-9874; Randy Roels 926-9871, rroels@sharedhealth.mb.ca).
  • Regarding obtaining automated ABG data at HSC and St. B:
    • Julie reported on direct comparison of ABG/VBG data from DSM vs. our collectors. Identification of individuals who got blood gases was 84% for VBG and 95% for ABG. Identification of the number of test was 82% for VBG and 92% for ABG. We agreed to move ahead now with obtaining these data (again, just for HSC and St. B; we will still get these data manually at Grace) from DSM.
  • Regarding consolidating some of the “what was done” components of CCI therapeutic interventions:

2. Followup on having the ICU nurses do all TISS coding --

  • We have top administrator agreement at HSC. Dan is working on getting similar meetings at St. B and Grace. Goal is to get their agreement and to then:
    • Begin having the nurses routinely and uniformly start doing all TISS sheets 1/1/2020.
    • For a couple of weeks to do audit by also having our data collectors do their own, separate TISS scoring ---> have Julie compare results ---> If adequate (>90% or so), then cease having our collectors do TISS, BUT to repeat such fidelity testing about 1 week every 6-12 months, indefinitely; see 2020 TISS audit

3. Followup on trying to get hospital-level data elements from EPR. Tina discovered that there is an application called Cognos which has capability to extract data elements from existing databases.

  • Allan left a message with Phil Jarman (926-8036) to pursue this further.
  • Tina has opened incident 3845870 and is working with Alex Omsen and Mike Ocko on getting this setup; see Cognos#Phone call with Mike Ocko

4. Follow up on how to code -- Allan will explore these:

  • Compartment syndrome other than abdominal -- it turns out that this is already included and explained in Muscle, ischemic infarction
  • Mesenteric vascular injuries
  • Diagnostic sampling of pericardial fluid or the pericardium -- to be discussed further at next Task meeting
    • for therapeutic pericardial drainage combine (T) Pericardium with Drainage, Evacuation
    • for diagnostic pericardiocentesis -- we currently do not have the components to create this item. We could add it as a picklist item, and indeed have a couple of choices:
      • we could call it "Diagnostic pericardiocentesis" with code 2.HA.52
      • or we could include both diagnostic fluid drainage AND diagnostic pericardial biopsy together, calling it "Diagnostic pericardial sampling (fluid or tissue)" with code 2.HA.13.

5. New issue of pulling in certain lab results as well as current counts.

  • The plan is to re-parse all DSM data from 1/1/2019, and to use this opportunity to pull in time/dates and results for specified tests.
    • Barret with help from Tina will generate a draft list of which tests to do this for, which we’ll discuss at the next Task meeting. It should include the tests needed to calculate APACHE 2. Care must be taken to balance usefulness with data storage issues.
  • As part of this, we will work so that the labs values that are part of APACHE 2 no longer need to be dealt with by the data collectors.

6. Item we didn't discuss yet -- unconfirmed diagnoses with priority 0. We'll discuss next time.

7. New items:

  • Barret raised the issue of CPEs = carbapenemase producing enterobacteraciae, and more generally the rising rate of carbapenem resistant bacteria. He suggested that in the list of resistant bacteria, we split out carapenem resistance as a separate category. Allan spoke to micro lab and ID personnel who also suggested tracking these separately. Thus a new code has been added Resistance to carbapenem antibiotics. Previously this was included in 82.8 Resistance to antimicrobials, antibiotic, resistance to other beta lactams; Allan has changed that Wiki page to remove it from there.
  • Allan will seek out how to code so-called autotransplantation of kidney --- turns out this is included in CCI, as 1.PC.83 which is "transfer" of kidney. We don't currently include "transfer" among the list of "what was done". So since this is a very rare procedure, to code it we should use 1.PC.94, i.e. combine therapeutic intervention on kidney with the what was done category of "NOS" (94).
  • Request for a code for bronchopleural fistula -- it turns out we had already decided on this, using the appropriate one of: Pneumothorax, traumatic OR Pneumothorax, tension, nontraumatic OR Pneumothorax, nontension, nontraumatic OR Pneumothorax, nontraumatic, NOS
    • IF, as is usually the case when these fistulas are spontaneous (rather than caused iatrogenically or by trauma), it is due to lung infection or abscess, then that code should also be combined with the appropriate one of these. And in this case also code Empyema (pyothorax) if present, combining it with the others.
  • Query about coding for the myraid of other fistulas out there -- Allan identified that there are separate codes for all of these when nontraumatic (J95.03 for T-E, K31.6 for stomach or duodenum, K60 for anorectal,, K63.2 for intestinal, K82.3 for gallbladder, M25.1 for joint, N32.2 for bladder, N82 for female genital tract, and other for less common ones (e.g. lacrimal duct)). What we have to decide is whether to include some or all of these additional codes, or just to code them in the appropriate "NOS" category. We'll discuss this at the next Task meeting.
  • Question was asked of how to code chronic subdural hematoma. It turns out that we previously dealt with this question at an earlier task meeting. ANSWER was that we already have I62 Subdural or epidural hematoma/hemorrhage, nontraumatic and that unless/until somebody specifically wants to distinguish these two related entities for a specific question/project, we will not subdivide them. Allan has added a note to that Wiki page saying so.
  • Question was asked of whether/how to code a flail chest resulting after rib fractures from CPR.
    • This is related to a prior question of how to code Rib fracture(s) due to CPR. We also already have a code for traumatic flail chest, i.e. Flail chest, injury/trauma. Given our general rule of NOT using trauma codes for iatrogenic injuries, Allan proposes here that when a flail chest occurs after CPR (which is rare) that we simply use Rib fracture(s) due to CPR. He has added this to the relevant Wiki pages.

Next meeting 1/2/2020 at 11am.

ICU Database Task Group Meeting – November 27, 2019

1. Followup about working to reduce collector workload

  • Blood Product Data - Allan reported that he connected with the CBS TraceLine administrator (Margaret Ring; margaret.ring@blood.ca) and Tony Loewen (anthony.loewen@blood.ca) about obtaining TraceLine automated data for all blood product transfusions. Allan will follow up on this if he doesn't hear back in the next couple of weeks.
  • Direct Data Access for RIS/PACS - Allan reported that he connected with RIS administrators (Angela Charbonneau 926-9874; Randy Roels 926-9871, rroels@sharedhealth.mb.ca) about seeking automated radiology data. He has now sent an email to the manager) Allan will follow up on this if he doesn't hear back in the next couple of weeks.
  • ABG Data - Regarding obtaining automated ABG Data -- it appears this info IS included (for HSC and St B, but not Grace) in the DSM data stream we receive.
    • Allan has parsed the listing to identify the elements we want.
    • Tina has done a comparison between DSM data and CCI for ABGs, see ABG_Data#Tina.27s_comparison
  • Regarding consolidating some of the “what was done” components of CCI therapeutic interventions:
    • Agreed to combine Excision, Resection, Excision, Resection, total and Excision, Resection, radical into a single item CCI “what was done” items which we’ll call Excision, Resection and give it the existing code of ‘87’.
      • Note that giving it this existing code is different than we discussed at today’s Task Group meeting (where we thought we’d make up a new code), but Allan realized that adapting an existing code is much more consistent with the way these “what were done” codes were originally chosen/assigned.
      • Note that in making this change, Julie will map any of the original 3 codes to the '87' code.
      • Tina has made the changes to the codes.

2. Regarding TISS coding -- based on the recent meeting of Dan, Jodi, Allan and Trish:

  • At HSC Jodi will work (starting after 1/1/2020) to ensure that the ICU nurses do the TISS.
  • At the other 2 hospitals, Jodi provided names of the nursing personnel that we will need to meet with to work towards our goal. Dan will work on planning those meetings.
  • The goal is that the nurse-completed TISS forms are of uniformly acceptable quality that the data collectors will no longer need to check them. But before we can assume that we need to do audits. We discussed two kinds of audits:
    • Have the collectors do a random sample of completely separate (duplicate) TISS sheets to compare “side-by-side”
    • There are some audits we can do with only the nurse-completed TISS forms -- i.e. #missing days; items where we have crosschecks (e.g. dialysis)

3. We made major decisions regarding the modality of data collection and the interface of data collection with the Minimal Data Set.

  • We WILL move to collecting almost all data in medical records after discharge.
  • The new/updated Minimal Data Set will include only items available from EPR, specifically:
    • serial, location, names, PHIN, chart#, admission date/time, accept date/time, previous location, DOB and IF AVAILBLE yet, then disposition location and disposition date/time.
      • the expectation is that all of these will be rechecked and fixed if needed on final chart review

4. Regarding the IICU consult temporary project that was supposed to terminate in July of 2019. Allan asked Bojan/Jodi when we can cease it --- answer is stop at the end of 2019.

5. Regarding trying to get hospital-level data elements from EPR. Tina discovered that there is an application called Cognos which has capability to extract data elements from existing databases.

  • Allan left a message with Phil Jarman (926-8036) to pursue this further.
  • Tina is in contact with Alan Madick to find out if we can have a more suitable EPR Report that would give us some of this at entry time.

6. Regarding “separation” of iatrogenic injury and trauma codes. We had previously agreed (see Oct 24, 2019 minutes, item#4) not to code any trauma codes along with an iatrogenic injury. However, we realized today that it would be important -- for iatrogenic codes that don’t adequately specifiy what was injured -- to combine it with a non-trauma code identifying what was injured.

7. Queries for how to code -- Allan will explore these:

  • Compartment syndrome other than abdominal
  • Mesenteric vascular injuries
  • Diagnostic sampling of pericardial fluid or the pericardium

8. Item we didn't discuss yet -- unconfirmed diagnoses with priority 0. We'll discuss next time.

Next meeting 12/11/2019 at 11am.

ICU Database Task Group Meeting – November 14, 2019

  • Present: Allan, Barret, Joanna , Julie, Michelle, Tina, Trish
  • Absent: Con
  • Minutes prepared by: AG
  • Action items in BOLD

1. Allan notified the group that he has made a Wiki page for the upcoming December 17 First Annual Baking Event.

2. Followup about working to reduce collector workload

  • Blood Product Data - Allan connected with Margaret Ring (margaret.ring@blood.ca) and Tony Loewen (anthony.loewen@blood.ca) from CBS about obtaining TraceLine automated data for all blood product transfusions. At their request he sent them an email with the description of what we're seeking, and the existing PIA approval from HIPC. Allan will follow up on this if he doesn't hear back in the next couple of weeks.
  • Direct Data Access for RIS/PACS - Allan connected with the RIS administrator (Angela Charbonneau 926-9874) about seeking automated radiology data. He has now sent an email to the manager Randy Roels (926-9871; rroels@sharedhealth.mb.ca) Allan will follow up on this if he doesn't hear back in the next couple of weeks.
  • ABG Data - Regarding obtaining automated ABG Data --- Kim Webster from HSC Resp Therapy indicated that the hospitals do not maintain a database of results, but that this SHOULD be obtainable from DSM. To start Tina will check to see if it’s already included in what we get from DSM.
  • Regarding consolidating some of the “what was done” components:
    • proposal: combine all 4 current versions of “Excision” into a single item
    • possibility: combine “fixation” and “fusion”
    • possibility: combine “repair”, “reattachment”, and “construction or reconstruction”
    • we will discuss these at the next Task Group meeting
  • A major time savings would be to cease trying to get the data in real-time (except for the Minimal Data Set which we will need to further discuss in regards to both the quarterly reporting and Overstay)
    • At the next Task meeting we will work out Minimal Data Set issues, and then plan to pilot this in a few locations where we are currently trying for real-time data collection.
  • Regarding trying to have ICU nurses accurately did the TISS sheets. On 11/26/2019 a meeting was held including Dan, Jodi, Allan and Trish to address this. The result was that:
    • At HSC Jodi will work (starting after 1/1/2020) to ensure that the ICU nurses do the TISS. We will need to plan an auditing to ensure good quality before we cease having our data collectors check the TISS.
    • At the other 2 hospitals, Jodi provided names of the nursing personnel that we will need to meet with to work towards our goal. Dan will work on planning those meetings.

2.Leftover item: Michelle identified a problem related to kidney TP. Specifically, we have a current rule disallowing CRF Stage 5 AND any acute kidney problem --- but this should be EXCEPT in the presence of a transplanted kidney. Apparently this isn’t working correctly, possibly because there are 2 ways to identify a renal TP: Past history, transplanted kidney, or the CCI code. Tina to work to fix this.

3. New items:

  • Regarding the “X Medicine Service” at HSC. This is a separate attending internist who (usually temporarily) accepts admissions from ED to offload the admitting services. Usually, but not always, these patients end up on one of the regular ward services. Locating these patients, especially when they do NOT end up on a regular ward service, is challenging. Our access to the ADT system may eventually facilitate this but Tina points out that we’re just learning to incorporate that resource into our processes.
  • The IICU consult temporary project was supposed to terminate in July of 2019. Allan asked Bojan/Jodi when we can cease it --- answer is stop at the end of 2019.
  • In further discussion about the "set MRSA Colonized" button that Tina created for entering the various ICD10 and CCI items related to MRSA positivity -- we agreed that the CCI code for isolation should be an acquired procedure.
  • There was discussion about whether we could obtain hospital-level data elements (timing of hospital admit and discharge, discharge disposition or at least whether the patient survived hospitalization) from EPR. This would likely require us going back to e-Health to ask for a filtering capability in ADT in the form of a new web-based report in EPR. Allan and Tina will discuss this further.

ICU Database Task Group Meeting – November 5, 2019

  • Present: Allan, , Barret, Joanna , Julie , Tina, Trish, Dan
  • Absent: Con, Laura
  • Minutes prepared by: AG
  • Action items in BOLD

1. We spent the entire meeting discussing ideas about reducing collector workload, in light of the fact that due to a combination of factors (including personnel vacancies, vacations, sick leaves and increased patient throughput in both ICUs and wards) we are weeks to months behind in data collection. Aspects discussed were:

  • Allan to continue working to get automated data:
  • AG will work to consolidate the “what was done” component
    • proposal: combine all 4 current versions of “Excision” into a single item
    • possibility: combine “fixation” and “fusion”
    • possibility: combine “repair”, “reattachment”, and “construction or reconstruction”
  • A major time savings would be to cease trying to get the data in real-time (except for the “minimal data set” which we will need to further discuss in regards to both the quarterly reporting and Overstay)
    • There was general agreement that with exception of TISS, this will have little impact on data correctness
    • Though doing this would almost certainly delay data collection completeness by about a month or so, we see no major difficulty with it for reports or for research.
    • We agreed that once we work out the minimal data set issues, we should pilot this in a few locations where we are currently trying for real-time data collection.
  • It could save time to come up with a more automated and complete way of identifying admissions. This is needed because there is currently no “1-stop-shopping” approach to this. The paper logs in wards and ICUs are very imperfectly kept up to date. The approach that seems most promising is accessing the ADT system in each hospital -- we attempted this about 5-7 years ago without success, but we could try again.
  • Another MAJOR time savings would be if all the ICU nurses accurately did the TISS sheets. Since some do so, we know that it is possible and in fact does not require more than a few minutes per day. Several years ago the Steering Committee and Jodi tried to get the individual ICU nurse managers to get this done, without success. Plan is to set up a meeting for Dan, Jodi, Allan and Trish to address this again.
  • Finally, Dan led the group on a quick calculation of workload:
    • The ICU data collectors believe that the average ICU chart takes them 75 minutes to abstract fully. With about 6000 ICU records per year, this amounts to 4625 hrs of work. With 5 ICU-related EFTs, at 1920 hrs/year/EFT, this means that we have 9600 ICU collector hours/year, which SHOULD be plenty to get through all the charts. THUS, there appear to be some inefficiencies in play.

2.Leftover item: Michelle identified a problem related to kidney TP. Specifically, we have a current rule disallowing CRF Stage 5 AND any acute kidney problem --- but this should be EXCEPT in the presence of a transplanted kidney. Apparently this isn’t working correctly, possibly because there are 2 ways to identify a renal TP: Past history, transplanted kidney, or the CCI code. Tina to work to fix this.

3. Leftover item: How to handle transfer delays when the patient dies before the transfer happens. We recognized that before we can answer this we need to talk to Julie about her reporting on it.

ICU Database Task Group Meeting – October 24, 2019

  • Present: Allan, Con, Joanna , Tina, Trish
  • Absent: Barret, Julie, Laura
  • Minutes prepared by: AG
  • Action items in BOLD

1. Leftover item: Michelle identified a problem related to kidney TP. Specifically, we have a current rule disallowing CRF Stage 5 AND any acute kidney problem --- but this should be EXCEPT in the presence of a transplanted kidney. Apparently this isn’t working correctly, possibly because there are 2 ways to identify a renal TP: Past history, transplanted kidney, or the CCI code. Tina to work to fix this.

2. Continued discussion about ways to try and reduce the CCI workload.

  • Tina reported on analysis she did of the # of CCI codes per chart. She found out that mean and SD is about 3 and 3 -- thus 95% of records have <11 CCI codes, and that those with a lot of codes are primarily those with very long LOS. Furthermore, a lot of the replicate CCI coded charts have lots of transfusions and x-rays.
  • We decided to:
    • NOT now work to consolidate the # of body parts
    • Reinvigorate attempt to get CBS data directly (Allan to continue to try and contact anthony.loewen@blood.ca)
    • Work on trying to obtain PACS radiology data directly
    • Seek out whether each hospital has an ABG database we could get dumps from
    • AG will work to consolidate the “what was done” component
      • proposal: combine all 4 current versions of “Excision” into a single item
      • possibility: combine “fixation” and “fusion”
      • possibility: combine “repair”, “reattachment”, and “construction or reconstruction”

3. Leftover item: How to handle transfer delays when the patient dies before the transfer happens. We recognized that before we can answer this we need to talk to Julie about her reporting on it.

4. Leftover item: Uncertainty about if/how iatrogenic injury codes intersect with trauma codes.

  • After discussion we agreed NOT to try and ensure that they are combo codes. While this would provide a lot of detail for those users savvy enough to understand this nonstandard approach, in the setting of the usual way of generating reports, it lead to incorrect double counting of some things -- e.g. combo of an iatrogenic performation with a trauma code with mechanism of stab would be counted both as the iatrogenic thing and a stab wound. THUS, we agreed that we will accept the fact that iatrogenic codes generally do not delineate the body part involved. Tina has added to Template:ICD10 Guideline Iatrogenic an explanation that we do NOT want to use trauma codes for iatrogenic injuries.
  • The one exception desired is to be able to code rib fracture(s) due to CPR. Allan proposed and Tina added a new code for this: Y65.81, Rib fracture(s) due to CPR. He has already added to the page Rib, fracture, injury/trauma.

5. New items

  • Question was if/how to code a history of polio but with no sequelae (i.e. no physical deficits, no post-polio syndrome) -- answer is that this is like most (but not ALL, e.g. TB) completely resolved infections -- do NOT code it.
  • Question was how to code Lynch Syndrome, one of many genetic disorders conferring an elevated risk of cancer (in this case colon CA) -- answer is to add a wastebasket US ICD10 code: Z15.0, “Genetic susceptibility to malignant neoplasm”. Tina has added this and included in the page that it includes Lynch Syndrome, and also that “If a malignant neoplasm has occurred, code that separately”.
  • Issue about need to make 4 total entries for someone coming in with MRSA colonization
    • Tina will has made "set MRSA Colonized" button that creates all 4 entries: Staph aureus + Resistant to methicillin + colonized + CCI code for isolation, and indicate it’s existence on the relevant wiki pages.
    • Current guidelines and data check don’t allow an individual to be both Colonized with organism (not infected) and infected. But clearly this IS possible so this data check will be changed to a “soft” check instead. Tina was not able to find a cross check for this; are we sure one exists?

ICU Database Task Group Meeting – October 3, 2019

  • Present: Allan, Con, Michelle, Tina, Trish
  • Absent: Barret, Joanna, Julie, Laura
  • Minutes prepared by: AG
  • Action items in BOLD

1. Leftover item: Michelle identified a problem related to kidney TP. Specifically, we have a current rule disallowing CRF Stage 5 AND any acute kidney problem --- but this should be EXCEPT in the presence of a transplanted kidney. Apparently this isn’t working correctly, possibly because there are 2 ways to identify a renal TP: Past history, transplanted kidney, or the CCI code. Tina to work to fix this.

2. Leftover item: Ideas for reducing workload relating to nonradiologic, therapeutic interventions CCI coding:

    • Condense/combine some body parts -- e.g. all segments of the upper extremity
    • Reduce the number of “what was done” items -- e.g. combine partial, complete and radical resections
    • For very complicated, multifactorial procedures (e.g. multiple tissues and/or multiple procedures done on same REGION of the body) to create picklist items
    • IF we implement these, we’ll want to, as much as possible, remain true to the organization of CCI
    • Allan to contemplate how to implement these. Discussed with Tina, and she has pulled up some numbers (CCI Volumes 2019) to learn more about what would have biggest impact.

3. Leftover item: Blood Product Data - Regarding seeking to obtain all CBS data to merge in an ongoing way with our databases. Allan reports he’s continuing to try and contact Anthony Loewen (anthony.loewen@blood.ca) at CBS to pursue this.

4. Leftover item: Possible need for additional abscess locations: retroperitoneal, psoas. It was decided to add 2 new codes, from the US version: K68.1 = retroperitoneal abscess and K68.12 =psoas abscess. Tina has added these.

5. Leftover item: After discussion we agreed that we do not currently require to split out I62 Subdural or epidural hematoma/hemorrhage, nontraumatic into acute vs. non-acute.

6. Long discussion about if/how to split up the different types of beds at Riverview and Deer Lodge. Specifically, there are PCH beds and non-PCH beds. At Riverview the latter include chronic vent beds, rehab beds, and palliative care beds. After discussion we agreed that the only distinction we will have in those 2 places will be between the PCH beds and all other non-PCH beds, which we will call “Chronic Health Facility” beds. Likewise, St Amant, Selkirk Mental Health Centre and hospices will be considered to be Chronic Health Facility beds. Chronic Health Facility was updated to reflect this.

7. New items:

  • Regarding options for previous location: There are 2 options that never got implemented: Not known; Known but not listed. Tina has added these.
  • Regarding how to handle transfer delays when the patient dies before the transfer happens. We recognized that before we can answer this we need to talk to Julie about her reporting on it.
  • There is uncertainty about if/how iatrogenic injury codes intersect with trauma codes.

Next Meeting -- October 24, 2019 at 11 am

ICU Database Task Group Meeting – September 18, 2019

  • Present: Allan, Con, Joanna, Julie, Michelle, Tina, Trish
  • Absent: Barret, Laura
  • Minutes prepared by: AG
  • Action items in BOLD

1. Revamping of definitions for Admit and Comorbid diagnoses -- Allan and Tina reworked this, and the group approved the new wording.

2. Michelle identified a problem related to kidney TP. Specifically, we have a current rule disallowing CRF Stage 5 AND any acute kidney problem --- but this should be EXCEPT in the presence of a transplanted kidney. Apparently this isn’t working correctly, possibly because there are 2 ways to identify a renal TP: Past history, transplanted kidney, or the CCI code. Tina to work to fix this.

3. Regarding admit procedure coding in relation to whether an organ transplant done prior to admission to the current location should be included as an Admit Procedure. The issue here relates to the current 48 hour limitation. We agreed that for organ transplants we will remove that time limit. Done.

4. Leftover item: Ideas for reducing workload relating to nonradiologic, therapeutic interventions CCI coding:

    • Condense/combine some body parts -- e.g. all segments of the upper extremity
    • Reduce the number of “what was done” items -- e.g. combine partial, complete and radical resections
    • For very complicated, multifactorial procedures (e.g. multiple tissues and/or multiple procedures done on same REGION of the body) to create picklist items
    • IF we implement these, we’ll want to, as much as possible, remain true to the organization of CCI
    • Allan to contemplate how to implement these.

5. Leftover item: Blood Product Data - Regarding seeking to obtain all CBS data to merge in an ongoing way with our databases. Allan reports he’s continuing to try and contact Anthony Loewen (anthony.loewen@blood.ca) at CBS to pursue this.

6. New collector items:

  • Need for additional abscess locations: retroperitoneal, psoas, muscle.
    • Retroperitoneal, psoas abscess: In the US version of ICD10, K68.1 is retroperitoneal abscess and K68.12 is psoas abscess. In the Canadian versioin there are no such specific codes for these, and they’re included under K65.0 which is Peritonitis, infectious. So we need to discuss how specific we need to be here.
    • For generic muscle abscess, we already have Muscle, infection (pyomyositis, infective myositis)
  • Need to be able to code a chronic subdural hematoma/hemorrhage --> ANSWER: we already have I62 Subdural or epidural hematoma/hemorrhage, nontraumatic which we could use to cover both acute, subacute and chronic. This could be subdivided as I62.0 (acute) and I62.9 (not acute, including chronic and subacute). So we need to discuss how specific we need to be here.
  • Need to be able to code a renal hematoma/ hemorrhage --> ANSWER: This is one of those diagnoses for which there isn’t a single ICD10 code, but it can be created by combining Kidney or ureter, disorder NOS and Hemorrhage, NOS
  • Whether to distinguish ward bed borrows as “stabilize” vs. “contingency”. Answer is No.
  • We currently have no way to identify the “prior inpatient location” as the rehab section of Riverview. Tina will add this.
  • After discussion we agreed that all WRHA ICUs (including CCU) EXCEPT IICU will be considered to be the same level of care.

Next Meeting -- October 3, 2019 at 11 am

ICU Database Task Group Meeting – August 28, 2019

  • Present: Allan, Con, Joanna, Julie, Michelle, Tina, Trish
  • Absent: Barret, Laura
  • Minutes prepared by: AG
  • Action items in BOLD

1. Revamping of definitions for Admit and Comorbid diagnoses -- Allan and Tina reworked this, and the group approved the new wording.

2. Michelle identified a problem related to kidney TP. Specifically, Query check ICD10 ESRD vs ARF disallows CRF Stage 5 AND any acute kidney problem --- but this should be EXCEPT in the presence of a transplanted kidney. Apparently this isn’t working correctly, possibly because there are 2 ways to identify a renal TP: Past history, transplanted kidney, or the CCI code. Tina to work to fix this. Tina has emailed Michelle for some further info Ttenbergen 22:33, 2019 October 2 (CDT)

3. Regarding admit procedure coding in relation to whether an organ transplant done prior to admission to the current location should be included as an Admit Procedure. The issue here relates to the current 48 hour limitation. We agreed that for organ transplants we will remove that time limit. Done.

4. Leftover item: Ideas for reducing workload relating to nonradiologic, therapeutic interventions CCI coding:

  • Condense/combine some body parts -- e.g. all segments of the upper extremity
  • Reduce the number of “what was done” items -- e.g. combine partial, complete and radical resections
  • For very complicated, multifactorial procedures (e.g. multiple tissues and/or multiple procedures done on same REGION of the body) to create picklist items
  • IF we implement these, we’ll want to, as much as possible, remain true to the organization of CCI
  • Allan to contemplate how to implement these.

5. Leftover item: Blood Product Data - Regarding seeking to obtain all CBS data to merge in an ongoing way with our databases. Allan reports he’s continuing to try and contact Anthony Loewen (anthony.loewen@blood.ca) at CBS to pursue this.

6. New collector items:

  • Need for additional abscess locations: retroperitoneal, psoas, muscle. Answers:
    • Retroperitoneal, psoas abscess: In the US version of ICD10, K68.1 is retroperitoneal abscess and K68.12 is psoas abscess. In the Canadian versioin there are no such specific codes for these, and they’re included under K65.0 which is Peritonitis, infectious. So we need to discuss how specific we need to be here.
    • For generic muscle abscess, we already have Muscle, infection (pyomyositis, infective myositis)
  • Need to be able to code a chronic subdural hematoma/hemorrhage --> ANSWER: we already have I62 Subdural or epidural hematoma/hemorrhage, nontraumatic which we could use to cover both acute, subacute and chronic. This could be subdivided as I62.0 (acute) and I62.9 (not acute, including chronic and subacute). So we need to discuss how specific we need to be here.
  • Need to be able to code a renal hematoma/ hemorrhage --> ANSWER: This is one of those diagnoses for which there isn’t a single ICD10 code, but it can be created by combining Kidney or ureter, disorder NOS and Hemorrhage, NOS
  • Whether to distinguish ward bed borrows as “stabilize” vs. “contingency”. Answer is No.
  • We currently have no way to identify the “prior inpatient location” as the rehab section of Riverview. Tina will add this.
  • After discussion we agreed that all WRHA ICUs (including CCU) EXCEPT IICU will be considered to be the same level of care. Confirmed that this is what we have in Transfer_Ready_DtTm_field#Hierarchy_of_levels_of_care

Next Meeting -- October 3, 2019 at 11 am

ICU Database Task Group Meeting – August 28, 2019

  • Present: Allan, Barret, Con, Joanna, Julie, Tina, Trish
  • Absent: Laura
  • Minutes prepared by: AG
  • Action items in BOLD

1. Continued discussion of reducing workload relating to CCI coding.

  • We will not, at this time, eliminate nonradiologic diagnostic procedures, as the collectors relate that these generally are quick to do.
  • Ideas for doing this for nonradiologic, therapeutic interventions:
    • Condense/combine some body parts -- e.g. all segments of the upper extremity
    • Reduce the number of “what was done” items -- e.g. combine partial, complete and radical resections
    • For very complicated, multifactorial procedures (e.g. multiple tissues and/or multiple procedures done on same REGION of the body) to create picklist items
    • IF we implement these, we’ll want to, as much as possible, remain true to the organization of CCI
    • Allan to contemplate how to implement these

2. Followup of the large errors in coding the number of CXRs.

  • Further forensic work indicates that this looks like an uploading/programming issue, as the CCMDB backups seem correct while the final database numbers often incorrectly have a count of ONE.
  • Tina fixed the underlying cause in CCMDB.accdb_Change_Log_2019#2019-08-28
  • Julie is going to try and reconstruct the correct values from saved backups from July 1, 2019.
  • To deal with the values, starting 1/1/2019 that we cannot reconstruct, we agreed that Julie will put a large negative number in them to “signal” that they are incorrect, e.g. -99999
    • Julie re-claimed these from backups

3. In light of the fact that “Homelessness” is one of the items in Pre acute living situation, we will retire the ICD10 code Z59.0, i.e. Homelessness. Tina has taken care of this.

4. Leftover item: Blood Product Data - Regarding seeking to obtain all CBS data to merge in an ongoing way with our databases. Allan reports he’s continuing to try and contact Anthony Loewen (anthony.loewen@blood.ca) at CBS to pursue this.

Next Meeting -- September 18, 2019 at 11 am

ICU Database Task Group Meeting – August 22, 2019

  • Present: Allan, Julie, Michelle, Tina, Trish
  • Absent: Barret, Con, Laura
  • Minutes prepared by: AG
  • Action items in BOLD

1. Followup of bed borrowing and EMIP, i.e. “boarding”.

  • Julie and Tina have started to implement Boarding Loc.
  • Plan is to extend this machinery to Medicine and in particular plan a 1 month test where we do it for both moves BEFORE AND AFTER arrival at the expected service location -- start date to be 9/1/2019.

2. Continued discussion of the workload relating to CCI coding.

  • The issue arose with the question of if indeed, as discussed on July 16, every level of GI biopsy needs to be done (e.g. if during EGD they biopsy esophagus, stomach and pylorus). Our initial answer was “yes”, but this spawned a discussion about if and how diagnostic testing information contained in the database will be used.
  • For now, Allan will explore in CCI whether we should code fluid sampling as clearly different from “biopsy”.
  • But next time we will also have a larger discussion about the benefit/workload of recording diagnostic tests in general.

3. Followup: regarding up to 15 different ICD10 codes are being linked together. Allan, Trish and Julie met and recognized that in general this IS being done correctly, and in any case there are few consequences to doing it incorrectly (except for the “mandatory” combinations of bug with infection, mechanism with trauma). Allan has created a new section about this issue in the Wiki page Combined ICD10 codes, called "Q&A: Just How Far Should You Go in Linking Cause and Effect Diagnoses?".

4. Followup: Eliminate Depressive episode without major depressive disorder, and effectively fold it into Mood (affective) disorder, NOS. Tina has made this change.

5. Followup: There is no body part for therapeutic interventions (e.g. amputation) on fingers or toes. Allan will look these up this ---> Answer: Tina has added to CCI:

  • 1.UJ = Therapeutic Intervention on Finger --> (T) Finger
  • 1.WL = Therapeutic Intervention on Toe --> (T) Toe

6. Followup: Last time we agreed on a new rule that in coding a Cardiac arrest, the INITIAL observed rhythm should be coded and linked to it. This requires having a code for each arrest rhythm, and we currently have ones for: Ventricular fibrillation, Ventricular tachycardia, and all the supraventricular rhythms. But it turns out that ICD10 does not have codes for asystole or PEA, i.e. they’re included under Cardiac arrest. So in order to make this rule feasible, Tina has added new (invented) ICD10 codes for:

7. New items:

  • There continues to be large errors in coding the number of CXRs.
    • Julie/Pagasa did some work in ICUs indicating that: (a) of 474 records with an ICULOS>4 days and just one recorded CXR, this was incorrect in 88% of them, (b) this phenomenon is common in MICU, SICU, ICMS, Con but not in ICCS, ACCU, Gra and Oaks.
    • In discussion we narrowed the problem down to either collector issues OR being related to different approaches that different collectors are taking on the laptops to recording the number of CXRs.
    • We agreed that the first step will be for Julie, Trish, Tina and Pagasa to look into how different collectors (some at MICU/SICU and others) are recording this information.

8. Leftover item: Blood Product Data - Regarding seeking to obtain all CBS data to merge in an ongoing way with our databases. Allan reports he’s continuing to try and contact Anthony Loewen (anthony.loewen@blood.ca) at CBS to pursue this.

Next Meeting -- August 27, 2019 at 11 am

ICU Database Task Group Meeting – July 30, 2019

  • Present: Allan, Barret, Con, Laura, Tina, Trish
  • Absent:Julie
  • Minutes prepared by: AG
  • Action items in BOLD

1. Followup of bed borrowing and EMIP, i.e. “boarding”.

  • Julie and Tina will look at Boarding Loc listed in the minutes from 6/26/2019 to ensure what is written there makes sense.
  • Plan is to extend this machinery to Medicine and in particular plan a 1 month test where we do it for both moves BEFORE AND AFTER arrival at the expected service location.

2. Continued discussion of the workload relating to CCI coding.

  • The issue arose today with the question of if indeed, as discussed on July 16, every level of GI biopsy needs to be done (e.g. if during EGD they biopsy esophagus, stomach and pylorus). Our answer last time was “yes”, but this spawned a discussion about if and how diagnostic testing information contained in the database will be used.
  • For now, Allan will explore in CCI whether we should code fluid sampling as clearly different from “biopsy”.
  • But next time we will also have a larger discussion about the benefit/workload of recording diagnostic tests in general.

3. Followup: regarding up to 15 different ICD10 codes are being linked together. Allan, Pagasa and Julie met and there is to be followup.

4. Followup: Tina has retired the following 6 codes of care of various artificial orifices,

5. Followup: Continued discussion of coding “depression”. We decided to eliminate Depressive episode without major depressive disorder, and effectively fold it into Mood (affective) disorder, NOS. Tina has made this change.

6. New items:

  • (a) There is no body part for therapeutic interventions (e.g. amputation) on fingers or toes. Allan will look these this up.
  • (b) Discussion of arrythmias in cardiac arrest. After discussion, we agreed to: (i) add separate codes for asystole and PEA, (ii) have a new rule that in coding a Cardiac arrest, the INITIAL observed rhythm should be coded and linked to it. HOWEVER, after the meeting, Allan discovered that there are no specific ICD10 codes for asystole or PEA, i.e. they’re included under cardiac arrest. Thus, at our next meeting we will discuss whether we want to invent codes for them.
  • (c) Questions in relation to renal transplant admissions:
    • As it is a routine part of the procedure, the ureteral stent placed should NOT be separately coded.
    • For a patient who previously had hyperparathyroidism (secondary or tertiary as a result of chronic renal disease), but had removal of most of the parathyroids and now that hyperpara is gone, we will NOT code hyperparathyroidism. Regarding hyperparathyroidism, our current code E21.3 should include tertiary -- Allan will make this change in the wiki -- DONE.

7. Leftover item: Blood Product Data - Regarding seeking to obtain all CBS data to merge in an ongoing way with our databases. Allan reports he’s continuing to try and contact Anthony Loewen (anthony.loewen@blood.ca) at CBS to pursue this.

Next Meeting -- August 22, 2019 at 11 am

ICU Database Task Group Meeting – July 16, 2019

  • Present: Allan, Barret, Con, Joanna, Julie, Trish
  • Absent: Laura, Tina
  • Minutes prepared by: AG
  • Action items in BOLD

1. Followup -- Iatrogenic, puncture or laceration, related to a procedure or surgery NOS -- it currently requires a mechanism of injury code. We will change all the iatrogenic codes to not require a mechanism.

  • This had already been done.

2. Regarding H. pylori. After discussion it was agreed that this should be in the category of infection with implied pathogen (or equally, pathogen with implied infection).

  • This had already been done.

3. Followup -- how to code cognitive impairment that was not present at admission but developed after admission. Tina created a Template, called Template:ICD10 Guideline Altered mental status, and Allan has populated it.

  • Now Tina will link it to the appropriate pages.

4. Followup -- We will add to the list of dispositions a group of codes such as: “Vic-transitonal care”.

  • This had already been done.

5. Followup of bed borrowing and EMIP, i.e. see: Boarding Loc.

  • Julie and Tina will look at the “machinery” listed in the minutes from 6/26/2019 to ensure what is written there makes sense.
  • Plan is to extend this machinery to Medicine and in particular plan a 1 month test where we do it for both moves BEFORE AND AFTER arrival at the expected service location.

6. Followup: What code to use for large bowel ulcers -- answer is that the current code K63.3, now entitled Small intestine, ulcer actually should be renamed Intestinal ulcer.

  • This had already been done.

7. NEW ITEMS:

  • (a) How to code fetal intrauterine death”
  • (b) How to identify the onset date of a VAP? Answer is that there already, in the VAP article on Wiki, a part that addresses this.
    • But in direct answer to Laura’s question, we should use as the data of onset, for a person who as of NOW has all 4 criteria, the point at which the first of criteria #2, #3 or #4 were met.
  • (c) Should we have “presence of an indwelling Foley” as an ICD10 comorbid code.
  • (d) If during endoscopy, multiple different anatomic sites are biopsied, should a CCI entry be made for each site biopsied.
    • Answer: Yes, e.g. if during EGD they biopsy esophagus, stomach and pylorus, then there should be a CCI code for each of those sites. However, just one code for each site, regardless of how many biopsies were done of that site.
    • confirmed that this info is on the wiki pages
  • (e) Julie has noted that up to 15 different ICD10 codes are being coded as: Combined ICD10 codes.
    • Plan: To assess whether this is appropriate, Julie and Allan will meet to discuss.
      • For clarification, we identified 4 situations in which linking is completely reasonable:
        • (1) link bugs with an infection,
        • (2) link trauma with its mechanism,
        • (3) link codes to “create” an entity for which no separate ICD10 code exists, such as retroperitoneal hemorrhage,
        • (4) to connect cause with effect(s), e.g. a trauma combined with all the separate fractured bones.
  • (f) How to code local invasion by a growing tumor.
  • (g) Whether to code removal of a urethral stent?
    • Answer: No, because we previously determined that coding removals of things will require the coders to constantly check back for removal of things. However, e.g. a cystoscopy that was done to remove it SHOULD be coded.
    • Made sure this is reflected in List of CCI procedures we don't code
  • (h) Whether to code (as Past Medical History) allergies to drugs.
  • (i) Whether to code a failed procedure, i.e. one where the procedure was begun but they were not able to complete it. Examples include surgical procedures, as well as failing in CVC insertion due to not being able to thread the wire.
    • Answer: Yes
    • Aborted Procedure provides more details; I added a redirect from "failed procedure" so this can be found more easily. Ttenbergen 11:19, 2019 July 25 (CDT)
  • (j) Whether to remove the ICD10 codes for care of various artificial orifices, e.g. colostomy.

8. Leftover item: Blood Product Data regarding seeking to obtain all CBS data to merge in an ongoing way with our databases.

  • Allan reports he’s continuing to try and contact Anthony Loewen (anthony.loewen@blood.ca) at CBS to pursue this.

Next Meeting -- July 30, 2019 at 11:15 pm

ICU Database Task Group Meeting – July 3, 2019

  • Present: Allan, Barret, Con, Julie, Tina
  • Absent: Joanna, Trish, Laura
  • Minutes prepared by: AG
  • Action items in BOLD

1. Followup: Continued discussion about what to code when what is written is just “depression”.

2. Followup -- Iatrogenic, puncture or laceration, related to a procedure or surgery NOS -- it currently requires a mechanism of injury code. We will change all the iatrogenic codes to not require a mechanism. Tina has taken care of this.

3. Regarding H. pylori. After discussion it was agreed that this should be in the category of infection with implied pathogen (or equally, pathogen with implied infection). Tina has taken care of this.

4. Followup -- how to code cognitive impairment that was not present at admission but developed after admission.

5. Followup -- We will add to the list of dispositions a group of codes such as: “Vic-transitonal care”. Tina has taken care of this.

6. Followup about coding a soft tissue hematoma. The Wiki article for Hemorrhage or Bleeding in ICD10 When There is No Specific Code now explains to code this as a combination of Hemorrhage, NOS and Soft tissue, disorder NOS.

7. Followup --- coding of Previous Location, Previous Service, and Prior Inpatient Location for people who are admitted to one of our locations from dialysis. There are 3 such situations and here are the decisions for them:

  • Came to dialysis from being an outpatient: Previous Location=dialysis; Previous Service=Nephrology; Prior Inpatient Location=n.a.
  • Came to dialysis from ED: Previous Location=dialysis; Previous Service=ED; Prior Inpatient Location=n.a.
  • Came to dialysis from a prior, different, inpatient location: Previous Location=dialysis; Previous Service=service of that prior inpatient location; Prior Inpatient Location=that prior inpatient location

8. Followup of bed borrowing and EMIP, i.e. “boarding”.

  • Julie and Tina will look at the “machinery” listed in the minutes from 6/26/2019 to ensure what is written there makes sense.
  • Plan is to extend this machinery to Medicine and in particular plan a 1 month test where we do it for both moves BEFORE AND AFTER arrival at the expected service location.
  • see Boarding Loc

9. New items:

  • (a) What bug to use for Viral hepatitis, chronic, NOS when the specific virus is not known -- answer is Virus, NOS
  • (b) What code to use for large bowel ulcers -- answer is that the current code K63.3, previously entitled "Intestinal ulcer" actually should be renamed Intestinal ulcer. Tina has made that change. Allan has edited the text in K63.3 to make this clear.
  • (c) What to do for a urine culture that was “positive” but not noted by the team in the chart, and thus not specifically treated. After discussion we agreed that in this case, we should NOT call this a UTI.
  • (d) If/how to code “mass effect” in the brain. Answer is that the existing code G93.5, which currently is called Brain compression, including herniation, actually is officially called “Brain compression, including herniation”. Tina has changed it’s name to that. Allan has added info to that Wiki page to make clear what it includes.

10. Leftover item: Blood Product Data - Regarding seeking to obtain all CBS data to merge in an ongoing way with our databases. Allan reports he’s continuing to try and contact Anthony Loewen (anthony.loewen@blood.ca) at CBS to pursue this.

Next Meeting -- July 18, 2019 at 2 pm

ICU Database Task Group Meeting – June 26, 2019

  • Present: Allan, Barret, Con, Joanna, Julie, Tina, Trish,
  • Absent: none
  • Minutes prepared by: AG
  • Action items in BOLD

1. Today we only went over new collector items, as follows:

  • (a) Regarding depression: As most commonly the only info about this in the notes is “depression”, without further details, there may be a benefit for some sort of depression NOS code. Allan will look into this ---> ANSWER: There are 4 “depression” codes as follows
  • (b) Regarding Iatrogenic, puncture or laceration, related to a procedure or surgery NOS -- it currently requires a mechanism of injury code. Response is that we intend to change all the iatrogenic codes to not require a mechanism. Tina has fixed this.
  • (c) In a ward-wide C-diff outbreak, EVERYBODY was put on Isolation, infectious, most prophylactically. After discussion we agreed that the rule would be that such prophylactic isolation will NOT be coded. However, if someone who begins on prophylactic isolation then develops the infection, that as of that day he/she WILL have the isolation coded. Allan will make this addition in the Wiki article on isolation --> DONE.
  • (d) Regarding Helicobacter pylori (H. pylori). After discussion it was agreed that this should be in the category of infection with implied pathogen (or equally, pathogen with implied infection). Tina has taken care of this.
  • (e) Question about a code to cover cognitive impairment that was not present at admission but developed after admission. We will need to discuss this further at the next Task meeting as there is a large range of causes/disorders that can fit this description. By far the most common sets of these are caused by so-called “toxic-metabolic” causes of confusion, delerium, or the even less well defined item of “altered mental status” -- which encompass many common things in the ICU such as sepsis, medications, abnormal blood chemistry, withdrawal effects from drugs or alcohol, etc. We already have codes that could cover all of this, and need to figure out exactly what is being sought in this question.
  • (f) Discharge to the increasing number of "Transitional Care" beds and sites -- e.g. Vic, etc. It was decided that we will add to the list of dispositions a group of codes such as: “Vic-transitonal care”. Tina has taken care of this.
  • (g) Question about a code for a nonspecific hematoma. Allan will look into this --> ANSWER: Assuming what is being sought here is a soft tissue hematoma, there is no specific ICD10 code for that. The closest is our existing code Soft tissue, disorder NOS. We will discuss this more at the next Task meeting.
  • (h) Question about how to code a disseminated infection. Allan will look into this --> ANSWER: There are specific subcodes for specific disseminated infections (mainly TB and fungal infections) but more generally for this use the existing code Infectious disease NOS, for which Allan has added language to indicate its use in this circumstance.
  • (i) Question about coding of sex vs. gender. After discussion that included recognition of the fact that our main interest here is biologic and not sociologic, and in light of the increasingly fluid nature of gender definitions, that our goal will continue to be seeking to code birth sex, doing the best we can with the data we have to discern it.
  • (j) We began, but did not have time to complete discussion of what to code for “Sending Location” and “Prior Service” for patients who come to a ward or ICU from Dialysis, after having before that been either: (i) outpatients, (ii) in ED, or (iii) in a different inpatient location. We will continue this discuss at the next Task meeting.

2. Followup: Continued discussion about dealing with the related issues (in both Medicine and ICU) of bed borrowing and EMIP, i.e. “boarding”. Essentially we decided to adapt the “machinery” currently being used in ICU, as follows:

  • Collect the following times for all patients in ICU and Medicine:
    • Service accept date/time -- defined as when the destination team takes over care. For Medicine this is most easily identified as when the Medicine service admitting orders are timed.
    • Initial arrival date/time -- defined as when the patient first arrives at ANY non-ED destination bed. So, this will include ICU bed, any ward bed. For example, if a patient accepted on the MICU service goes and boards in SICU, this time will be the arrival at the SICU. Similarly if a Medicine patient whose expected final destination is Ward1 is first put in a contingency bed, this is the arrival in the contingency bed.
    • Temp date/time -- this field can have multiple entries, and each one is a date/time and a location. For the patient who initiallly does go to their expected final bed, this entry simply says “No bed borrow”. But for a patient who initially goes to a borrowed bed, the first entry here duplicates the Arrival date/time as above, with a location of where they actually went (in the above example of MICU patient in an SICU bed, that location would be SICU). Subsequent moves (including a move from a contingency bed to the final expected service location) has an entry here too. In ICU this currently mechanism is used for ALL moves both before and after arrival in the expected service location. We are proposing extending this machinery to Medicine and in particular plan a 1 month test where we do it for both moves BEFORE AND AFTER arrival at the expected service location.
      • Such temp locations for Medicine will be simple, i.e. for each hospital there’ll be a single contingency bed item, e.g. “Contingency bed-Grace”.
      • Temp locations for ICU will be PACU, and all the other ICUs in the given hospital.

3. Followup. Decision from last time regarding coding repeated clinical events. Specifically that for such repeated events in 3 categories (arrythmias, signs/symptoms, and abnormal lab tests) they should only be listed in a given database record ONCE.

    • e.g. A person has a self-limited episode of A-fib. It goes away and then recurs. This should only be listed the first time.
    • e.g. A person has a self-limited episode of A-fib. It goes away but then he has an episode of V-tach. As this is a different diagnosis, both of these should be listed, but only once each.
    • e.g. Patient comes in with hypokalemia. It’s treated and remits, but the next day it recurs. Only list it the first time.

Tina has set up Template:ICD10 Guideline repeated events with this info and applied it to all the dxs in those three categories.

4. Leftover item: Blood Product Data - Regarding seeking to obtain all CBS data to merge in an ongoing way with our databases. Allan reports he’s continuing to try and contact Anthony Loewen (anthony.loewen@blood.ca) at CBS to pursue this.

Next Meeting -- July 3, 2019 at 1:30 pm

ICU Database Task Group Meeting – June 20, 2019

  • Present: Allan, Valerie, Julie, Michelle, Tina, Trish
  • Absent: Laura, Con, Joanna
  • Minutes prepared by: AG
  • Action items in BOLD

1. Ongoing item. Continued discussion about dealing with the related issues (in both Medicine and ICU) of bed borrowing and EMIP, i.e. “boarding”. Essentially we decided to adapt the “machinery” currently being used in ICU, as follows:

  • Collect the following times for all patients in ICU and Medicine:
    • Service accept date/time -- defined as when the destination team takes over care. For Medicine this is most easily identified as when the Medicine service admitting orders are timed.
    • Initial arrival date/time -- defined as when the patient first arrives at ANY non-ED destination bed. So, this will include ICU bed, any ward bed. For example, if a patient accepted on the MICU service goes and boards in SICU, this time will be the arrival at the SICU. Similarly if a Medicine patient whose expected final destination is Ward1 is first put in a contingency bed, this is the arrival in the contingency bed.
    • Temp date/time -- this field can have multiple entries, and each one is a date/time and a location. For the patient who initiallly does go to their expected final bed, this entry simply says “No bed borrow”. But for a patient who initially goes to a borrowed bed, the first entry here duplicates the Arrival date/time as above, with a location of where they actually went (in the above example of MICU patient in an SICU bed, that location would be SICU). Subsequent moves (including a move from a contingency bed to the final expected service location) has an entry here too. In ICU this currently mechanism is used for ALL moves both before and after arrival in the expected service location. We are proposing extending this machinery to Medicine and in particular plan a 1 month test where we do it for both moves BEFORE AND AFTER arrival at the expected service location.
      • Such temp locations for Medicine will be simple, i.e. for each hospital there’ll be a single contingency bed item, e.g. “Contingency bed-Grace”.
      • Temp locations for ICU will be PACU, and all the other ICUs in the given hospital.

2. Followup. Decision from last time regarding coding repeated clinical events. Specifically that for such repeated events in 3 categories (arrythmias, signs/symptoms, and abnormal lab tests) they should only be listed in a given database record ONCE.

  • e.g. A person has a self-limited episode of A-fib. It goes away and then recurs. This should only be listed the first time.
  • e.g. A person has a self-limited episode of A-fib. It goes away but then he has an episode of V-tach. As this is a different diagnosis, both of these should be listed, but only once each.
  • e.g. Patient comes in with hypokalemia. It’s treated and remits, but the next day it recurs. Only list it the first time.

Tina has set up Template:ICD10 Guideline repeated events with this info and applied it to all the dxs in those three categories.

3. Followup. Regarding potential confusion between when to code Urinary tract infection, NOS versus Bladder, cystitis, acute infectious. Final decision was to code whichever the clinical teams calls it. IF they write both, then use the more specific one, i.e. cystitis.

4. Followup. After further discussion, it was decided to eliminate the ICD10 code Physical rehabilitation care. Tina has taken care of this.

5. Leftover item: Blood Product Data - Regarding seeking to obtain all CBS data to merge in an ongoing way with our databases. Allan is continuing to try and contact Anthony Loewen (anthony.loewen@blood.ca) at CBS to pursue this.

Next Meeting -- June 26 at 1130 am.

ICU Database Task Group Meeting – June 12, 2019

  • Present: Allan, Joanna, Valerie, Julie, Tina, Trish
  • Absent: Laura, Con
  • Minutes prepared by: AG
  • Action items in BOLD

1. At this meeting we continued to discuss decisions about dealing with the related issues (in both Medicine and ICU) of bed borrowing and EMIP, i.e. “boarding” -- specifically PRIOR to arriving in the primary service location. Essentially we will adapt the “machinery” currently being used, as follows:

  • Collect the following times for all patients in ICU and Medicine:
    • Service accept date/time -- defined as when the destination team takes over care. For Medicine this is most easily identified as when the Medicine service admitting orders are timed.
    • Arrival date/time -- defined as when the patient first arrives at ANY non-ED destination bed. So, this will include ICU bed, any ward bed. For example, if a patient accepted on the MICU service goes and boards in SICU, this time will be the arrival at the SICU. Similarly if a Medicine patient whose expected final destination is Ward1 is first put in a contingency bed, this is the arrival in the contingency bed.
    • Temp date/time -- this field can have multiple entries, and each one is a date/time and a location. For the patient who initiallly does go to their expected final bed, this entry simply says “No bed borrow”. But for a patient who initially goes to a borrowed bed, the first entry here duplicates the Arrival date/time as above, with a location of where they actually went (in the above example of MICU patient in an SICU bed, that location would be SICU). Subsequent moves (including a move from a contingency bed to the final expected service location) has an entry here too. In ICU this currently mechanism is used for ALL moves both before and after arrival in the expected service location. We are proposing extending this machinery to Medicine.
      • Such temp locations for Medicine will be simple, i.e. for each hospital there’ll be a single contingency bed item, e.g. “Contingency bed-Grace”.
      • Temp locations for ICU will be PACU, and all the other ICUs in the given hospital.

2. Leftover item. As we are now collecting lab data in ICUs and wards, and we don’t want to double count them, we agreed that when patients transfer from one collecting site to another that we need coordination as regards time leaving/arriving. After prior discussion, it was decided to use the times from EPR in which people were moved to avoid this. Val or Joanna will put this into the Wiki.

3. Followup. There’s a need for a code for nonspecific obstruction of the upper airways. We will use J39.8 and call it Upper airway obstruction, NOS. The wiki page will include a notation that other, more specific options may include:

Tina has added this code.

4. Followup. Decision from last time regarding coding repeated clinical events. Specifically that for such repeated events in 3 categories (arrythmias, signs/symptoms, and abnormal lab tests) they should only be listed in a given database record ONCE.

    • e.g. A person has a self-limited episode of A-fib. It goes away and then recurs. This should only be listed the first time.
    • e.g. A person has a self-limited episode of A-fib. It goes away but then he has an episode of V-tach. As this is a different diagnosis, both of these should be listed, but only once each.
    • e.g. Patient comes in with hypokalemia. It’s treated and remits, but the next day it recurs. Only list it the first time.

Need to followup and ensure that these ideas are put somehow into the Wiki.

5. New item. Agreed that should use the ICD10 code for Palliative care if the patient is made ACP-C.

6. New item. How to code an infected PD catheter. Answer is combination of Iatrogenic, infection, internal prosthetic device or implant or graft NOS and Peritonitis, infectious.

7. New item. There is confusion between when to code Urinary tract infection, NOS versus Bladder, cystitis, acute infectious.

8. New item. When to code Physical rehabilitation care, specifically whether it covers: (i) simple bedside physio and/or OT, (ii) cardiac rehab (as this is ordered in every post-MI patient) Thoughts on this from Allan, for discussion at next meeting: What we are seeking here is “real rehab, as guided by a Rehab physician (i.e. a physiatrist). This almost always occurs only in specific rehab floors (as exists at HSC). However, it is also possible for such a person to consult on a ward patient (rarely ever in ICU). The question we must address is whether it’s easily possible to distinguish such rehab -- because if not we should probably just remove this code.

9. Leftover item: Blood Product Data - Regarding seeking to obtain all CBS data to merge in an ongoing way with our databases. Allan reports he’s continuing to try and contact Anthony Loewen (anthony.loewen@blood.ca) at CBS to pursue this.

Next Meeting -- June 20 at 11 am.

ICU Database Task Group Meeting – June 6, 2019

  • Present: Allan, Joanna, Julie, Tina, Trish, Val
  • Absent: Laura, Con
  • Minutes prepared by: AG
  • Action items in BOLD

1. There was more discussion of options for dealing with the related issues (in both Medicine and ICU) of bed borrowing and EMIP, i.e. “boarding” -- specifically PRIOR to arriving in the primary service location. Allan related that he spoke to Nick Hajidiacos and that the Medicine program would like BOTH the total time spent boarding, and the time spent boarding in ED. Thus, we agreed to:

  • For both Medicine and ICU we will collect the following times:
    • (a) Accepted by service -- may call it something like “Service Accept Date/Time”
    • (b) Arrived in primary service location, where if they never did arrive in that location this will be set to the disposition time -- may call it something like “Primary Service Location Arrival Date/Time”
    • (c) Arrived in a “contigency” bed where this is defined as a non-ED bed on which the patient is boarding in the hope/expectation that he/she will eventually obtain a bed in the Primary Service Location. We’ll call this something like “Boarding Bed Arrival Date/Time”
      • For ICUs the list of contingency beds will include: PACU, ED, and all other ICUs
      • For Medicine the list of contingency beds will be, e.g. Contingency bed/HSC, etc. That is to say we will include all contingency beds in a given hospital as “one thing”.
  • For Medicine, the primary location to which a person will be assigned before they get a bed in that primary location (i.e. while boarding in ED or in a contingency bed) will be the expected primary ward.
  • Thus, for Medicine, for patients who never got to their Primary Service Location, i.e. what used to be called EMIP, will now be the time interval between Service Accept Date/Time and Boarding Bed Arrival Date/Time, and the ward assigned will be the expected primary ward (based on the service to which he/she was accepted).
  • For ICU, as we are already collecting data that indicates all moves, we will continue to do that.
  • For Medicine, but not ICU, to indicate any time spent in a contingency bed (i.e. different from the Primary Service Location) will be simply indicated with a check box -- i.e. we are not going to keep track of post-arrival boarding/moves.
    • more discussion to follow. Julie will bring whiteboard of main office suggested process.
    • Allan was given (took) the whiteboard with the process main office suggested.
    • wiki page for this will be Boarding Loc

2. Followup. In response to prior item of no ICD10 code for generic autoimmune disease, in fact we already had a code for this, i.e D89.9 = Disorder of the immune system, NOS.

  • Allan has added to this Wiki page a note that it includes nonspecific autoimmune diseases.

3. Followup. It was previously agreed that we will change the rule for recording Radiation Therapy under CCI to be recorded only the first date it occurs during an admission. Tina has changed this on wiki and in ccmdb.

4. Followup. The number of chest X-rays, ABGs and transfusions have dramatically fallen since switching to ICD10/CCI.

  • For transfusions Joanna pointed out that this could be real for FFP and platelets, as we changed the rule of what constitutes one unit.
  • For Xrays and ABGs, it’s likely that this is a problem with the coders not following the current rules with the current laptop “machinery” to capture all of them.
    • To deal with this, Trish will re-emphasize to all coders the expectations for recording ALL of these.
    • 2019-May-27: an email was sent out to collectors in regards to this matter.

Collectors were advised when doing CCI's, to keep eye on FAR left of row in after selecting CCI item. This tell you how or if it should be counted. Also sent a link to review CCI Collection Mode Wiki article. Trish Ostryzniuk 11:33, 2019 June 13 (CDT)

5. Followup. Regarding our new categorization of death (Death to morgue, Death with organ donor to OR, Death with brain death to another ICU). Contrary to what is written in the prior task group minutes, since brain dead organ donors are rare, in calculating ICU-specific death rates, we will not remove people admitted with brain death as an ADMISSION diagnosis from the numerator and denominator (which would be the rigorously correct thing to do).

6. Leftover item. As we are now collecting lab data in ICUs and wards, and we don’t want to double count them, we agreed that when patients transfer from one collecting site to another that we need coordination as regards time leaving/arriving.

  • We decided that Tina, Trish, Con and Joanna (and possibly other collectors) will discuss a preferred approach to this, which we’ll discuss at next Task meeting.

7. New item. Due to it’s cost, the program wants us to collect CAR-T (Chimeric Antigen Receptor T-Cell) treatment data. This is technically a procedure so will need a CCI code, but like prior chemotherapy we will want to capture it as a past history diagnosis too.

8. New item. How to code assault with the assaulter’s own body (e.g. fists, feet). The answer is we’ll add a new item: Y04, Mechanism of injury: assault with bodily force. (Tina added this to CCMDB and wiki)

9. New item. We agreed to remove R06.7, Sneezing -- from the list of signs/symptoms (Tina has retired this code in CCMDB and wiki)

10. New item. There’s a need for a code for nonspecific obstruction of the upper airways. We will use J39.8 and call it “Upper airway obstruction, NOS”. That entry must include a notation that other, more specific options may include:

(Tina has added this)

11. New item. How to code neutropenic enterocolitis. Answer:

12. New item. Whether in a patient whose primary admit cause is Respiratory failure (insufficiency) NOS, acute linked to a presumed etiology (e.g. COPD exacerbation, or pneumonia) whether the ARF or the cause should get the check box for being THE SINGLE primary admission cause.

  • After discussion, we agreed there are pros and cons of each and that either is OK.

13. New item. It’s been brought to the group’s attention that the coders are sometimes coding every level reached with an endoscopy. Inspection, Exploration (endoscopic). This is contrary to the guideline to only code the furthest area reached.

  • Trish will re-emphasize to all coders the expectations here..
  • 2019-Jun-10: Wiki article updated.
  • 2019-June-13: email sent to staff with link to article to review update collection instruction details:Inspection, Exploration (endoscopic) .

14. New item. Whether a new diagnosis (reason) of the reason for respiratory failure is needed each time a patient is reintubated. The answer is it depends on whether the repeat intubation is for a new or old diagnosis.

  • For example if the patient was admitted and intubated for a COPD exacerbation, gets better, gets extubated and stays extubated for a few days, but then develops a new pneumonia and gets reintubated -- the pneumonia would be a second, acquired, diagnosis -- separate from the original COPD exacerbation.
  • But if instead the patient has a pneumonia as the original cause, seems to improve, gets extubated because he passed a SBT, but fails and gets reintubated and it’s believed the ongoing cause of the respiratory failure and reintubation is just the continuation of the original pneumonia, then no new diagnosis needed

15. New item. Whether to code repeated clinical events -- e.g. multiple recurrent episodes of an Arrhythmia, or an abnormal sign or symptom (e.g. a lab result).

  • We realized that there can be many of these examples and that we’ll have to deal with them as they arise. But we decided that for now we will address the following 3 categories: arrythmias, signs/symptoms, and abnormal lab tests. For ALL of these, they should only be listed in a given database record ONCE.
    • e.g. A person has a self-limited episode of A-fib. It goes away and then recurs. ***This should only be listed the first time.
    • e.g. A person has a self-limited episode of A-fib. It goes away but then he has an episode of V-tach.
      • As this is a different diagnosis, both of these should be listed, but only once each.
    • e.g. Patient comes in with hypokalemia. It’s treated and remits, but the next day it recurs.
      • Only list it the first time.

16. Leftover item: Blood Product Data - Regarding seeking to obtain all CBS data to merge in an ongoing way with our databases.

  • Allan reports he’s continuing to try and contact Anthony Loewen (anthony.loewen@blood.ca) at CBS to pursue this.

Next Meeting -- June 12 at 11 am.

ICU Database Task Group Meeting – May 23, 2019

  • Present: Allan, Con Joanna, Julie, Laura, Michelle, Tina, Trish
  • Absent: none
  • Minutes prepared by: AG
  • Action items in BOLD

1. There was more discussion of options for dealing with the related issues (in both Medicine and ICU) of Bed borrowing and EMIP, i.e. “boarding”. A sticking point is the important question of whether or not the Medicine program needs/wants to separate out time spent waiting for a ward bed in ED, from the total time spent waiting for a ward bed (ED plus so-called contingency beds). Allan will contact Nick Hajiadacos and ask ---> Answer: They want both the total time prior to arriving in their destination bed AND the amount of that time spent specifically waiting in ED.

2. Julie indicated that she would like to routinely obtain, for the Minimal dataset, the Accept DtTms and Arrive DtTms. This information is in the ADT system to which the collectors have access. Trish & Tina will work to make this happen.

3. New item. There is no ICD10 code for generic autoimmune disease. This is relevant because, apparently, it occurs that a diagnosis is made of an autoimmune disease without a specific name or a specific organ. Allan will look this up  Answer is that we already had a code for this, i.e D89.9 = Disorder of the immune system, NOS. Allan has added to this Wiki page a note that it includes nonspecific autoimmune diseases. We could also consider modifying it’s name to something like: Disorder of the immune system, NOS (including autoimmune disease).

4. New item. It was agreed that we will change the rule for recording Radiation Therapy under CCI to be recorded only the first date it occurs during an admission. Tina will make that change.

5. New item. Apparently, the number of chest X-rays recorded has dramatically fallen since switching to ICD10/CCI. As this is not likely due to an actual fall in CXRs, it’s almost certainly due to failure of coders to code them. Same has occurred with transfusions and ABGs. To deal with this, Trish will re-emphasize to all coders the expectations for recording ALL of these.

6. New item. This relates to our new categorization of death (Death to morgue, Death with organ donor to OR, Death with brain death to another ICU). Specifically, we had previously decided that for brain death cases who become donors, to count the end of the ICU admission as the time of brain death. But it is now pointed out that this fails to consider the workload after brain death for nurses. Thus we agreed to alter that rule such that we will instead count the end of the ICU admission as when that brain dead person actually leaves ICU (for morgue or another ICU or OR) ----> but that Julie will alter her code so that people admitted with brain death as an ADMISSION diagnosis will not be counted in the numerator or denominator for death rates.

7. Leftover item. As we are now collecting lab data in ICUs and wards, and we don’t want to double count them, we agreed that when patients transfer from one collecting site to another that we need coordination as regards time leaving/arriving. We decided that the Tina, Trish, Con and Joanna (and possibly other collectors) will discuss a preferred approach to this, which we’ll discuss at next Task meeting.

8. Leftover item: Blood Product Data - Regarding seeking to obtain all CBS data to merge in an ongoing way with our databases. Allan reports he’s continuing to try and contact Anthony Loewen (anthony.loewen@blood.ca) at CBS to pursue this.

Next Meeting -- June 6 at 11 am.

ICU Database Task Group Meeting – May 7, 2019

  • Present: Allan, Joanna, Julie, Laura, Michelle, Tina, Trish, Val
  • Absent: Con
  • Minutes prepared by: AG
  • Action items in BOLD

1. Most of this meeting was taken up with discussing options for dealing with the related issues (in both Medicine and ICU) of bed borrowing and EMIP, i.e. “boarding”.

  • An important difference between Medicine and ICU is that for the former we do not always know where the eventual service location will be before they arrive, while for the latter we do. This is because there are Medicine teams that can admit to one of several final locations, and what we keep track of is the final ward/location, not the team. This fact will necessitate some difference between how we handle boarding.
  • After much discussion we decided that Allan will ponder this further and we’ll discuss more at the next Task meeting. Allan’s suggestion for next meeting is that we put in place the following data “machinery”:
    • ICU:
      • Record the accept time, time of arrival at the final ICU location, and the disposition time. The interval between accept and arrival times is the “Boarding Time”. Interval between the accept and dispo times is the “Service LOS”, and is used to calculate the “Service Census”. Interval between the arrival and dispo times is the “Bed Occupancy LOS” and is used to calculate the “Bed Census”. For patients who never get to the final ICU location, the time of arrival is set at the dispo time, ensuring that Bed Occupancy LOS is zero. One example of bed borrowing is the MICU patient who is boarded in SICU and taken care of during that bed borrowing time by MICU team.
      • The current check box used to identify boarding during any part of care will have it’s purpose altered to indicate a patient who AFTER arrival in their final ICU location are sent elsewhere to board for any period(s) of time, while still cared for by the original ICU team.
      • Make a new field, called something like “Boarding Location” that for ICU has options: ED, PACU, Ward, Other. When a given patient is boarded in multiple locations prior to getting to the eventual ICU, we record the FIRST such location only.
    • Medicine:
      • Record the accept time, time of arrival at the final ward location, and the disposition time. For patients who never get to the final ward location, the time of arrival is set at the dispo time. The interval between accept and arrival times is the “Boarding Time”. The interval between the arrival and dispo times is the “Bed Occupancy LOS” and is used to calculate the “Bed Census”.
      • Now we diverge between those who DO vs. DO NOT get to a final ward location.
        • For those who DO get to a final ward location, we do just as we did for ICU: Interval between the accept and dispo times is the “Service LOS”, and is used to calculate the “Service Census”. Boarding Locations will have options: ED, and a list of wards that requires further discussion. Again, when a given patient is boarded in multiple locations prior to getting to the eventual ward site, we record the FIRST such location only.
        • For those who NEVER got to a final ward location: Interval between accep and dispo times is either “EMIP Time” or another title we choose. Since NOT all these will be boarded in ED, using EMIP time makes a link with current nomenclature, but is not fully accurate.

2. Julie indicated that she would like to routinely obtain, for the Minimal Dataset, the accept and admit times. This information is in the ADT system to which the collectors have access. Trish & Tina will work to make this happen.

3. There was a discussion about the CCI guidelines for which procedures to code that occurred PRIOR to actual arrival/admission.

  • For example, many collectors are coding transfusions and central lines done prior to admit. It appears the lack of clarity is around this criterion in the Admit Procedure article: “was directly related to an Admit Diagnosis
  • We decided that this criterion will be augmented to indicate to INCLUDE therapeutic procedures but only include diagnostic procedures if they led to a complication that was related to admission. Allan has made this change.

4. Leftover item. As we are now collecting lab data in ICUs and wards, and we don’t want to double count them, we agreed that when patients transfer from one collecting site to another that we need coordination as regards time leaving/arriving. We decided that the Tina, Trish, Con and Joanna (and possibly other collectors) will discuss a preferred approach to this, which we’ll discuss at next Task meeting.

5. Leftover time: Blood Product Data - Regarding seeking to obtain all CBS data to merge in an ongoing way with our databases. Allan reports he’s continuing to try and contact Anthony Loewen (anthony.loewen@blood.ca) at CBS to pursue this.

Next Meeting -- May 23 at 11 am.

ICU Database Task Group Meeting – April 23, 2019

  • Present: Allan, Con, Joanna, Julie, Tina, Trish
  • Absent: Laura
  • Minutes prepared by: AG
  • Action items in BOLD

1. Blood Product Data - Regarding seeking to obtain all CBS data to merge in an ongoing way with our databases. Allan reports he’s continuing to work with Anthony Loewen (anthony.loewen@blood.ca) at CBS to pursue this.

2. Continued discussion of coding deaths in the setting of Brain death, ( Deceased patients article) and organ donor. Summary of decisions:

  • Redefine Length-of-Stay. From now on, for patients who experience Brain death, LOS will include time from admission to Brain death. Time in ICU spent after Brain death being evaluated/optimized for being an organ donor will not be included in LOS. But, of course, that time will be included in calculations of Bed occupancy. Julie will have to revise her SAS code to make this happen.

Julie understood that this change was agreed to for dispo table?

  • Revise list of dispostions to now include THREE types of Dispo_field#Deceased_patients:
    • Death with transfer to morgue
    • Death with transfer to OR for organ donation
    • Death with transfer to another ICU
  • For a patient who is not yet braindead in location A and then transfers to location B with expectation of near-future Brain death or Donation after Cardiac Death (DCD)
    • In location A the dispo will be transfer to location B. The record in location B will have Admit Diagnosis of whatever is present. IF that person develops Brain death in location B, then that will be coded as an acquired diagnosis.
    • If the patient develops Brain death in location B, then the LOS in location B will only be the time from Arrive DtTm to Date and time of Brain death (The Date and time is recorded for ALL acquired DX's). For calculating this patient’s total LOS, it will be the entire time in location A + the time in location B until Brain death. Julie to change her algorithm accordingly.
  • For a patient who develops Brain death in location A, and then transfers to location B for evaluation for organ donation
    • In location A the patient will have an Acquired Diagnosis / Complication of Brain death, and their dispo will be Death with transfer to another ICU. The LOS in location A will be the date and time from arrival to the date and time of Brain death. Any time spent in location A after Brain death will be counted towards Bed occupancy but not LOS.
    • In location B the patient will have Admit Diagnosis of Brain death, and their dispo will be Death with transfer to OR for organ donation IF they go for donation. If they do not go for donation, the disop is Death with transfer to morgue. While all of this time will be included in Bed occupancy, none of it will be included in LOS (since the person was not alive).
    • For this situation, Julie will need to alter her coding for linking records into an episode of care, such that no linking will be done:
    • Regarding death rates in these records
      • It’s necessary to avoid double counting the death in location A and location B. To ensure this, Julie will alter her code such that (as directly above) there’s no linking of records with Brain deathand she will not include in the calculation any record where Brain death was an admit diagnosis.
  • For a patient who is in one location (e.g. MICU), where he develops Brain death, and then stays in that location to be evaluated/optimized for being an organ donor
    • Brain death will be an acquired diagnosis. LOS will only be the time from admit to Brain death. Time after Brain death will be counted towards bed occupancy but not LOS. Julie will alter her code to reflect this.
    • The dispo will be Death with transfer to OR for organ donation IF they go for donation and Death with transfer to morgue if not.
  • Regarding death rates in these circumstances

3. New item. As we are now collecting lab data in ICUs and wards, and we don’t want to double count them, we agreed that when patients transfer from one collecting site to another that we need coordination as regards time leaving/arriving. We decided that the Tina, Trish, Con and Joanna (and possibly other collectors) will discuss a preferred approach to this, which we’ll discuss at next Task meeting.

4. Bed borrow - We did not get to this item (left over from April 9, 2019 minutes) today. Will do next meeting.

Next Meeting -- May 7 at 11 am.

ICU Database Task Group Meeting – April 9, 2019

  • Present: Allan, Con, Julie, Michelle Tina, Trish
  • Absent: Laura
  • Minutes prepared by: AG
  • Action items in BOLD

1. Blood Product Data - Regarding seeking to obtain all CBS data to merge in an ongoing way with our databases. Allan reports he’s continuing to work with Anthony Loewen (anthony.loewen@blood.ca) at CBS to pursue this.

2. Allan reports that he made modifications to the 10 pages for electrolyte disturbances (high and low: Na, K, Ca, Mg, PO4). We agreed that we'll change the name of all of them to be, for example "Hypokalemia, severe or symptomatic" in order to clarify that we're NOT identifying ALL cases where the serum level is above or below the reference range. (Done - Tina)

3. Deceased patients - Continued discussion about coding the interrelated items of: death/disposition, brain death, and organ donor status. After discussion we agreed to solve this problem as follows:

  • First, it's only a problem when an already brain dead person is transferred elsewhere for reasons related to organ harvesting.
  • For those who are not brain dead when transferred with plans for DCD (donation after cardiac death) and organ harvesting, the Admit Diagnosis to the Transfer to location will just be whatever acute issues are involved.
  • For those who are declared brain dead at the TRANSFER FROM location, their diagnoses in record will include Brain death -- their Admit Diagnosis in the TRANSFER TO location will be Brain death and Organ donor (organ/tissue donation by the donor)
    • It's only this specific situation which is tricky, and our solution is that for records which contain the diagnosis of Brain death, Julie will change her algorithm for linking successive records such that no linking will be done:
      • in the forward time direction for a record where that diagnosis was NOT an admission diagnosis
      • in the backward time direction for a record where that diagnosis was either an admission diagnosis or a comorbid diagnosis

2) For both Medicine and ICU database, we check the continuity of transfers by checking the admission and discharge dates, the previous locations and discharge locations (or disposition). Errors on any of these four fields will affect either LOS, Inter-facility transfers, Readmission, mortality (Mortality and readmission report?), transfer delays (Transfer Delay?), occupancy (Bed occupancy?) which are included in the regular CC and Med reports. For ICU patient, this rule will only affect the SAS linking check program which can be modified so it will not show up as an error. How about in the ACCESS query of populate linking (Populate linking pairs) error (Pre-linking checks), this has to be changed too? In terms of diagnosis requests specific to brain dead, the counts will the include only those on acquired and not double count by including those on admit. --JMojica 10:13, 2019 April 11 (CDT)

4. Bed borrow - We began a very long discussion about the complex, interrelated issues -- RELATED TO TIME UNDER CARE PRIOR TO ARRIVAL IN THE SERVICE LOCATION -- of: bed borrowing, bed parking, and EMIP.

  • First, we recognized that similar issues are dealt with differently in Medicine and ICU, and that different names are given to what are equivalent phenomena
  • We have tentatively agreed (more discussion needed at next Task Meeting) on this schema for dealing with any situation that occurs between service acceptance and physical arrival on the ward/unit where time is spent NOT on the Service Location:
    • We will only use the word "occupancy" to refer to the number of patients physically occupying beds at the Service Location
    • We will refer to ALL situations (for Medicine and ICU) where a patient is under care of a service but not in that service's Service Location as BED BORROWING. So, we'll do away with the terms Bed Parking, and EMIP.
    • We will use the term "Service Number" to refer to the number of patients who are being cared for by the service team, whether those people are in the Service Location or in a borrowed bed
    • We realized that what was called EMIP (and which engendered consideration of ECIP) is really just a subcategory of bed borrowing where the patient never actually did get moved to the Service Location
    • We will continue to record the ACCEPT TIME and the ARRIVAL TIME regardless of bed borrowing --- BUT when a patient is put in a borrowed bed after Accept Time but never does get moved to the Service Location, we will assign the ARRIVAL TIME in that case to be identical to the DISPO TIME
      • This ensures that the Service Number is correct, and furthermore that such a person does not contribute to calculation of the occupancy of the Service Location
    • We will make a dropdown of the BED BORROW LOCATION -- with specific options including EDs, Wards, PACUs, ICUs, and an option for "MULTIPLE" which will be used when during the duration of pre-arrival bed borrowing the person is moved from one borrowed bed to another borrowed bed.
  • Issues related to this pre-arrival bed borrowing which still need to be addressed:
    • Whether Medicine needs to know just the total time in pre-arrival bed borrowing, or it's subsets when it occurs in multiple places -- Allan to ask the GIM Section Heads and Maryanne Lynch -- email sent
    • Whether EDIS is capable of providing the time specifically spent as a borrowed bed in ED
    • In these bed borrow situation, how to assign the PRIOR LOCATION and PRIOR INPATIENT LOCATION -- Allan to contemplate this, and ensure that Randy Martens data needs remain addressed after any change
      • Allan's proposal for this: Simply keep the definitions of those items exactly as they are now. This works just fine for Bed Borrowing with ONE exception.
        • When a patient never gets to their Service Location (i.e. dies or goes home or elsewhere before that occurs): In this case the PRIOR INPATIENT LOCATION still works fine. But the PRIOR LOCATION does not, since if the patient never gets to their Service Location, there is no prior physical location -- so the solution is to just add an entry to the dropdown list of possible physical locations that covers this situation; it could be called something like "NA-Never made it to Service Location".

5. A related topic to #4 is bed borrowing AFTER the Arrival Time on the Service Location. After discussion, and given that: (i) patients can and do (esp on Medicine) frequently get moved between bed borrow locations, and (ii) attempts in the past to keep track of all that movement failed miserably --> we agreed that what we will do, for both Medicine and ICU, is have a flag identifying that after the Arrival Time some unspecified period was spent off the Service Location, i.e. in a borrowed bed. (Depending on how we change the other definitions we might be able to re-purpose Off ward field for this; it's currently used for borrows at any time, but since pre-arrival borrows would now be marked differently we could change the definition for this. )

ICU Database Task Group Meeting – March 28, 2019

  • Present: Allan, Con, Joanna, Julie, Tina
  • Absent: Laura, Trish
  • Minutes prepared by: AG
  • Action items in BOLD

1. Blood Product Data - Regarding seeking to obtain all CBS data to merge in an ongoing way with our databases. Allan reports he’s continuing to work with Anthony Loewen (anthony.loewen@blood.ca) at CBS to pursue this.

2. Eliminating distinction between different ward types - Following up on providing the Medicine Database with information about whether ward patients are CTU or NTU. Allan reported that he has now also heard back from MaryAnne Lynch, who also indicated that she sees no reason to collect that information.

3. Followup of coding of STDs, particularly Syphilis. The changes agreed upon at the March 6, 2019 task meeting were implemented, except that Allan is still to populate Template:ICD10 Guideline STD to explain the general coding in ICD10 -- DONE.

4. Followup on the discussion at March 6, 2019 task meeting about defining cutoffs to identify electrolyte disturbances. The old codes for this can be found at Category:Metabolic_(old) Allan still to work on this.

5. IICU consult - Followup on March 6, 2019 task group discussion about recording only the initial date of referral made from an ICU to IICU. The question arose of what to do when patient is in ICU#1 when referral is made and then is transferred to ICU#2 (before getting to IICU). After discussion we agreed that we will NOT also code the IICU referral in ICU#2.

6. New question: How to code an anastamotic leak as a diagnosis, and how to code the repair as a procedure. The answer is that it is included in T81.3, i.e. Iatrogenic, dehiscence or rupture or disruption, surgical wound NOS. This article has now been augmented to reflect this inclusion.

7. New question: How to code pneumomediastinum or pneumoperitoneum.

8. New question: How to code Helicobacter pylori infection. An important issue here is that this bacterium is a very common colonizer and although it is the major cause of duodenal ulcers and gastric ulcers, it is present without causing any disorder in about 50% of all people. Additionally, the ONLY disease it is currently know to cause is peptic ulcers (including gastritis and duodenitis). For these reasons we will NOT add a specific bug code for it, and when such a disorder is believed due to this bug, one can combine the ulcer code with Bacteria, NOS.

9. New issue: Query NDC_Dxs_vs_TISS_Dialysis checks that in ICU, when a diagnosis is listed that “might” require dialysis, that the TISS is checked to ensure that a dialysis code is present. The problem is that with our new KDIGO definition of AKI/ARF, the vast majority of acute renal injuries are so minor that they do NOT end up getting dialysis. Thus, after discussion it was agreed that we will remove this crosscheck. (Query NDC_Dxs_vs_TISS_Dialysis has been updated to remove this component)

10. New question: What should be the primary (#1) diagnosis for someone transferred to D5 Medicine at HSC who is mainly there awaiting transfer to another facility (e.g. LTC home)? After discussion it was decided that if the patient has no active acute issues and is really only waiting for the transfer, then use one of the “ Awaiting/delayed transfer” codes as #1, but if there is still active treatment ongoing for an acute medical issue (e.g. finishing antibiotics for an infection) then code that issue as #1 and code the “Awaiting/delayed transfer” further down the priority list.

11. New question: How to code aspiration of a joint. Answer is:

12. Deceased patients - We had a long discussion about coding of the interrelated items of: death/disposition, brain death, and organ donor status.

  • This turns out to be complicated.
  • We currently have 2 death-related dispo codes, i.e. with and without organ donation. Organ donors may be alive (and remain alive), dead (i.e. braindead), or alive but they go for Donation after Cardiac Death (DCD) in the OR which is where they actually die.
  • A relevant issue here is the desire, when patient transfers from Location#1 to Location#2 (both of which we collect for) that the “TO” code for #1 matches up with the “FROM” code for #2. This is especially problematic when the person is braindead in #1, because current checks don’t allow a dead person to be transferred to another of our collected locations.
  • We discussed some approaches, including:
    • Subdivide the dispo death with donation code into: (i) death with donation went to OR and (ii) a cluster of codes that are death with donation went to specific other ICU. This would allow us to continue using the Disposition field to accurately identify unit-specific mortality rates.
    • Instead, maintain the 2-level dispo death coding, and use the ICD10 diagnosis code for braindeath to ensure accurate mortality rate calculations. This has the disadvantage that dispo codes are no longer accurate for death rates, and though Julie will know this, other users will almost certainly be confused by it.
  • After discussion, we agreed that we all need to ponder this issue, and continue discussion of it at the next task group meeting.

13. see: Tracheostomy care

Next Task Group Meeting: April 9 at 1pm

ICU Database Task Group Meeting – March 6, 2019

  • Present: Allan, Con, Joanna, Julie, Tina, Trish
  • Absent: Laura, Michelle
  • Minutes prepared by: AG
  • Action items in BOLD

1. Blood Product Data - Regarding seeking to obtain all CBS data to merge in an ongoing way with our databases. Their database is called Traceline. Allan will contact Anthony Loewen (anthony.loewen@blood.ca) at CBS to pursue this.

2. There was further discussion about consternation in the ICUs about our switch to the current CDC criteria for VAP. Specifically that the SICU Quality Circle is concerned about the apparent rise in VAP rates that occurred. Allan will send an explanatory email to the SICU leadership about this -- DONE.

3. Eliminating distinction between different ward types - Following up on providing the Medicine Database with information about whether ward patients are CTU or NTU. Allan reported that he heard back from the co-section heads (Griffin, VanAmeyde) who indicated that they appreciate the difficulty of collecting this data, and that the assignment has a large arbitrary element. They stated that it would be fine with them if we just recorded whether -- at the outset of any given Medicine database record -- the patient was CTU or NTU. However, Tina brought up that this may be of almost no actual value. As neither of the co-section heads place much value in this datum, Allan today left a voicemail with Maryanne Lynch to try and chat about it before we make a decision.

4. Followup on Bojan’s wish for IICU consult information from the ICUs. Allan reported that Bojan said that the SOLE piece of information he wants is the date the consult sult is first made to IICU. Tina has updated IICU consult and CCMDB.mdb s_tmp table to stop collection of consult decision date.

5. Regarding coding of STDs, particularly Syphilis. After discussion we agreed to switch the single existing code for syphilis, "Syphilis (due to Treponema pallidum)", from an “infection with implied pathogen”, to just being a pathogen (and changing the code's name to Treponema pallidum (Syphilis) in the process). Tina changed Treponema pallidum (Syphilis) to reflect this. That way, along with other STD-causing pathogens (e.g. Neisseria gonorrhea (gonococcus), Chlamydia trachomatis (bug responsible for regular sexually transmitted chlamydia)), we can code any type of infection. If it’s a simple cutaneous STD, you’d attach the appropriate STD pathogen to the generic code Sexually transmitted (venereal) infections, NOS. But for CNS syphilis we then have more flexibility in that we can use Encephalitis, meningoencephalitis, myelitis, encephalomyelitis, bacterial with the bug being Treponema pallidum.

6. A discussion occurred about coding of electrolyte disturbances, e.g. Hyperkalemia, severe or symptomatic. We agreed that we do NOT really want to record every time an electrolyte level is outside of the “normal range”. After discussion, we agreed that what we should do is identify thresholds for “severe” numerical disturbances, as existed before. Trish will send Allan the previously used cutoffs, and he will then come up with cutoffs to use with our ICD10 codes --- for which we will alter the names of these entities, for our purposes, to (example) “Hyperkalemia, severe”. - The old codes for this can be found at Category:Metabolic_(old)

7. After discussion we again agreed that the construct of a “failed discharge” isn’t a proper diagnosis. Instead the reason for readmission should be coded in that readmission, and if it’s due to psychosocial (rather than medical) causes, then one can use as the admission diagnosis Problem related to unspecified psychosocial circumstances.

8. In response to a question of if/how to code a KUB x-ray of the pelvis, we agreed it would be included as AXR (abdominal plain X-ray).

9. For clarity, we agreed that we will alter the names of "(T) Abdominal or Pelvic Cavity, NOS" to (T) Abdominal, Pelvic or Peritoneal Cavity, NOS and of "(D) Abdominal or Pelvic Cavity, NOS" to (D) Abdominal, Pelvic or Peritoneal Cavity, NOS. Tina has changed this on wiki and in CCMDB.mdb.

10. A question arose of how to code, in CCI, a lavage, e.g. gastric or peritoneal. Allan sought this out and it turns out that the “what was done” code used depends on the purpose of lavage. For example, if it is to warm up or cool down the body, the “what was done” codes correctly used are ones we decided NOT to include (Hypothermy; Hyperthermy). Similarly if it’s to instill medication, then the “what was done” code is another one we aren’t using (Local pharmacotherapy). THUS, it appears that for therapeutic lavages we are NOT coding those. On the other hand, for a DIAGNOSTIC lavage (as might be done for diagnosis of intraperitoneal bleeding or infection) the “what was done” is considered to be a “biopsy”, which we DO collect. Tina has added this info to List of CCI procedures we don't code, Biopsy (endoscopic), Biopsy (non-endoscopic).

11. A question arose of if/how to code the clinical finding of “cognitive impairment” without a clear/known cause. It turns out we already have a code for this Somnolence, stupor or obtundation, as the text of this Wiki page spells out that “This code includes the vague diagnoses of: altered mental status AND decreased LOC”. Added “cognitive impairment” to Somnolence, stupor or obtundation.

Next Task Group Meeting: March 28, 2019 at 11am

ICU Database Task Group Meeting – February 25, 2019

  • Present: Allan, Julie, Tina, Trish
  • Absent: Con, Joanna, Laura, Michelle
  • Minutes prepared by: AG
  • Action items in BOLD

1. Blood Product Data - Regarding seeking to obtain all CBS data to merge in an ongoing way with our databases. Their database is called Traceline. Allan will contact Anthony Loewen (anthony.loewen@blood.ca) at CBS to pursue this.

2. Regarding quarterly reporting of primary admit diagnoses. We have previously decided to report these by ICD10 chapter. However, we do want to pull out some particularly important diagnoses, e.g. sepsis-related diagnoses. For this purpose, Julie will extract Sepsis (SIRS due to infection, without acute organ failure) (R65.0) and Severe sepsis (R65.1) from the “R” chapter, and report those separately, noting for the remainder of “R” that it excludes those codes. Tina has updated the wiki pages to reflect this.

3. Regarding PHIN/pseudPHINs for people with multiple such identifiers. This can occur in 3 main ways: (a) moving from OOP to MB, (b) moving from MB to OOP, and (c) a single individual having multiple PHINs. For the purpose of identifying “same person”, and uniformity, we agreed that we will capture the current/most recent identifier (PHIN or pseudoPHIN) and record prior identifiers in the Alias ID collection.

4. An incident was noted where (at Grace) the patient qualified by CDC criteria as having a VAP, but the attending physician disagreed. Allan indicated that for diagnoses with specific criteria listed on the wiki (especially if those criteria are from CDC) that they should code them regardless of the physician’s concern. To try and deal with that concern, the collector should share the wiki page definition, and refer the physician to Allan. Allan sent out an email 2/25/2019 to all ICU attendings about this.

5. Eliminating distinction between different ward types - Regarding coding of different types of Medicine wards. Apparently the Medicine Program wants this information, but it is very challenging because patients not only switch rooms and wards, but can switch back and forth between CTU and NTU services. Probably the best we could do here is record CTU vs. NTU at the very start of admission. Allan sent an email about this to Maryanne Lynch, Ken VanAmeyde and Paddy Griffin -- the reply from Drs. Griffin and VanAmeyde was that they are completely OK with that plan.

6. IICU consult - Regarding the current temp project of recording IICU consult date and IICU accept date. The apparent purpose of this is for Bojan to be able to quantify the number of patient-days spent in ICU waiting for an IICU bed. However, we feel that simply summing the total intervals between IICU acceptance and (either) death or going to IICU or going elsewhere probably overestimates the actual time spent on the IICU waiting list. The reason is that these waits are often prolonged, and we have no way of recording if/when a patient is REMOVED from that waiting list. Allan opined that having the IICU attendings maintain a log of entry requests would be much more accurate and easier. Specifically, each entry could include: Patient idnetifiers, Date of initial IICU request, Sending ICU, Date of patient acceptance/refusal, Date patient removed from the IICU waiting list (for any reason). Allan discussed this with Bojan --- who said that the only information he needs is the date when the consult to IICU was first requested.

Next Task Group Meeting: March 6, 2019 at 11am

ICU Database Task Group Meeting – February 6, 2019

  • Present: Allan, Con, Joanna, Michelle, Tina, Trish
  • Absent: Julie, Laura
  • Minutes prepared by: AG
  • Action items in BOLD

1. Further consideration of reinstating CCI code Pharmacotherapy, antineoplastic agent, whole body for chemotherapy: We decided not to do so. The rationale is that what is considered chemotherapy is not quite confusing since: (a) many cytotoxic drugs are given for indications other than cancer, and (b) increasingly the drugs used for cancer are not cytotoxic but are biologics. - Tina updated Pharmacotherapy, antineoplastic agent, whole body with this info

2. Further consideration of adding a specific ICD10 code for Intravenous drug abuse (IVDA) -- There is no ICD10 code for IVDA and at this time we decided not to add a custom code unless some user really wants this information. - Tina has added this to List of ICD10 Diagnoses we don't code

3. Regarding coding of FFP and albumin transfusions:

  • Transfusion of plasma products: After discussion we decided that when stickers are not used, that the number of units will be determined by the current procedures of Canadian Blood Services that 1 unit is approximately 250 mL.
  • Transfusion of albumin: Because it comes in different concentrations, given Laura’s concerns, we will change CCI Collection Mode to CCI collect first - Tina has updated Transfusion of albumin and CCMDB.mdb with this.
  • Blood Product Data - There was discussion about seeking to obtain all CBS data to merge in an ongoing way with our databases. Their database is called Traceline. Allan will contact Anthony Loewen (anthony.loewen@blood.ca) at CBS to pursue this.

4. Regarding CAM coding. The current methods of collecting this data is in TISS and on the ICU flowsheet.

  • To even be evaluable for CAM assessment, a patient must be conscious. Thus, in reality, a given CAM assessment has 3 possible findings: (i) not evaluable, (ii) evaluable and CAM-positive, or (iii) evaluable and CAM-negative.
  • CAM status is currently recorded in ICUs in two places:
    • On the ICU flowsheet, as an indication of CAM-positive versus not being CAM-positive; as above the latter includes evaluable and CAM-negative or not evaluable. Of note, the ICU nurses evaluate for delerium multiple times per day.
    • On the TISS sheet, as a bubble for being CAM-positive versus not being CAM-positive. Here there is just a single yes/no bubble, so that NOT being CAM-positive can mean either that there were no evaluable assessments during that day or that there was at least one evaluable assessment and all of those were CAM-negative.
  • The users of this data (including Rakesh Arora in ICCS) are mainly counting the number of ICU days in which a patient is CAM-positive -- which of course only includes days on which there is at least one CAM-evaluable assessment.

- Allan has updated T9 - CAM positive (TISS Item) with this info

5. The data collectors made clear that it would be highly desirable for Wiki searching to be enhanced. Tina will ponder this and has started page Searching the wiki to document.

6. The idea was floated to (somehow) enhance the inclusion in Wiki pages of synonyms for diagnoses. We’ll come back to this sometime later. - See Searching the wiki

7. New question about how to code mets to the pancreas. Answer is Gastrointestinal system NOS, metastatic malignancy to it (also code primary site) and that page has been updated.

8. New question about whether it is OK to code bacterial Colonized with organism (not infected) as a comorbid. Answer is Yes.

9. After discussion relating to how diabetes might disappear, Allan has added Past history, transplanted pancreas or islet cells.

10. New question of situation whereby: Patient with ESRD is admitted for a kidney transplant -- gets the TP -- and kidney has trouble leading to acute renal failure/insufficiency in the transplanted organ. A number of questions arose with the following answers:

11. New question. There appear to be two conflicting ways that the Wiki indicates to code HAP. One listed in Iatrogenic, complication of medical or surgical care NOS and the other in Hospital-acquired pneumonia (HAP) in ICD10. Allan has fixed this with the correct approach being listed in the latter Wiki page.

Next Task Group Meeting: February 13, 2019 at 11am

ICU Database Task Group Meeting – January 24, 2019

  • Present: Allan, Con, Joanna, Julie, Tina, Trish
  • Absent: Laura
  • Minutes prepared by: AG
  • Action items in BOLD

1. There is still concern about the extra workload of ICD10/CCI. We will continue to monitor this and seek pithy suggestions for reducing the workload with minimal loss of content/value.

2. Consideration of adding Pharmacotherapy, antineoplastic agent, whole body back to the CCI list. This CCI picklist code would be 1.ZZ.35.HA-M0. At the Dec 7, 2018 task meeting we decided to eliminate it, though that item doesn’t explain why. We’ll reconsider at the next task.

3. Consideration of adding a specific ICD10 code for IVDA -- There is no ICD10 code for IVDA. The drug abuse codes go by the drug, not the route. If we decide we really need/want this, we can add a custom code. At the next meeting we’ll discuss this.

4. After discussion of whether we want to code CMV(+) status for organ transplants, we decided that we do not.

5. FFP does not have stickers that come with it from Blood Services. This led to a question of how to quantify FFP for coding Transfusion of plasma products.

  • Allan called the Winnipeg office of Canadian Blood Services and was told that for full units (approx. 250 mL) they do have stickers, but when they send half units that those do not have stickers. A solution appears to be to count the stickers, which should be there for whole units, but for half units, count them manually -- as 0.5 of a unit. We’ll discuss this more at the next Task meeting.

6. Question arose of how to code Factor V Leiden mutation. Allan will look into this ---> DONE, as the Wiki page indicates this is covered in Primary hypercoagulability (thrombophilia).

7. A complex question was raised about coding/counting CCI admit procedures that are done prior to admission, especially if done in a procedure suite on the way from one hospital ward or ICU to another hospital ward or ICU.

  • Our current criteria are listed in CCI Collection and that works fine when the patient comes to out ICU/ward from the ED or another location where we do not collect.
  • But, it’s complicated by the fact that is a patient goes from one to another of our collecting locations, that they might be counted in each place. Furthermore, a person being transferred from location A to location B may, in between, go to a procedure suite and get admit-type procedures.
  • We agreed to make 4 general rules for procedures:
    • (i) Transfer from collecting location A to collecting location B without any stop in between where procedures might occur -- all procedures done before leaving location A will be collected by location A only
    • (ii) Transfer from collecting location A to collecting location B WITH a stop in between where procedures occur -- all procedures done before leaving location A will be collected by location A only, while procedures done at the stop in between will be coded by location B only.
    • (iii) Transfer from noncollecting location A (which includes ED) to collecting location B without any stop in between where procedures might occur -- any qualifying admit procedures done before leaving location A will be collected by location B
    • (iv) Transfer from noncollecting location A (which included ED) to collecting location B WITH a stop in between where procedures occur -- all procedures done before leaving location A or during the stop in between will be coded by location B only.
  • Allan put the updated information on the wiki, and Tina moved it to CCI_Collection#Moved_patients from Admit Procedure since it applies to both admit and acquired.

8. Tina raised the issue of the possibility of the following true timing of events: First patient accepted for admission; Second patient deemed transfer ready to a lower level of care; Last is patient arrives. It’s an issue because the current cross-check Function Dispo Chronological() does not allow Transfer Ready DtTm to occur prior to Arrival D/T. After discussion (which unfortunately Tina was not present for), we agreed that the rule should be that Transfer Ready D/T can only be coded at or after Arrival D/T. The rationale has to do with the main desire for avoidable days to refer to actual bed occupancy days avoidable.

9. Discussion about coding Bacteremia. Although this is a finding and not an actual disease, because of it’s importance, we agreed that even though the general role is that coding findings/signs/symptoms is optional when the underlying cause is known, that for bacteremia we should ALWAYS code it when present. Furthermore, that at the discretion of the data collector, it can be linked to another presumed infection (e.g. Klebsiella pneumonia linked to Klebsiella bacteremia), but if it’s not completely clear that they’re related, to leave the bacteremia as “free standing”. Allan will modify the wiki page for Bacteremia, the sepsis template -- DONE.

Next Task Group Meeting: February 6, 2019 at 11am

ICU Database Task Group Meeting – January 9, 2019

  • Present: Allan, Con, Joanna, Julie, Tina, Trish
  • Absent: Laura
  • Minutes prepared by: AG
  • Action items in BOLD

1. Through discussion it became clear that there’s a need to modify the Kidney, renal tubular acidosis (RTA, all types) wiki page to clarify that by definition it is not an RTA if renal failure (acute or chronic) is present. Said another way, an RTA is a metabolic acidosis due to an inability of the renal tubules to excrete hydrogen ions in the presence of a normal creatinine clearance, as indicated usually by a normal creatinine. Allan will add this to the wiki article -- DONE.

2. There was substantial concern voiced by Con and Joanna about how long the new system is taking to code. At this point it’s as much as 4-fold longer than before. We discussed possible reasons, which include:

  • ICD10 coding, though this is possibly less burdensome than is CCI coding.
  • The biggest single issue raised was that among the 5 CCI Collection Modes:
    1. Collecting "CCI collect each" items
    2. Collecting "CCI collect count each" items
    3. Collecting "CCI collect count days" items
    4. Collecting "CCI collect count units" items
    5. Collecting "CCI collect first" items
    • We recognized that we probably could downgrade most of ‘1’ to be one of the others
    • And that for at least some of ‘2’, '3' and '4' we could downgrade to ‘5’
      • We decided today to do that for HD, PD, CRRT and ICP monitoring (Done - Tina)
      • Allan will take a look at the entire list, especially CCI Picklist, and consider further items that can be downgraded.
  • Other options for reducing workload for CCI include: (i) compressing the number of body parts, (ii) reducing and/or compressing the number of “what was done to the body part” items.
  • We’ll discuss all this at next Task meeting.

3. Julie raised the question of Charlson items -- specifically that previously most such items were allowed to be listed either as Admit Diagnosis or Comorbid Diagnosis. (See also Controlling Dx Type for ICD10 codes) The question is what do we want to do now about this. Allan will review both Charlson’s original description, and Quan’s administrative data implementation to see what THEY did regarding this --> DONE. The intention of this coding is to identify conditions that are present prior to admission. Thus, we should include admit and even acquired (post-admit) diagnoses for those Charlson items where it's pretty clear that the condition was almost certainly present prior to admission, even if that wasn't recognized -- and th is applies for 16 of the 17 Charlson conditions, i.e. all except "Myocardial infarction", and the only reason for that one being an exception is that there is an ICD10 code for Past history, myocardial infarction (old MI).

4. It was noted that the Template:ICD10 Guideline Como vs Admit is very confusing. Allan will work on it. (Template was added to Allan's list)