Task Team Meeting - Rolling Agenda and Minutes 2018

From CCMDB Wiki
Revision as of 16:12, 2018 May 18 by TOstryzniuk (talk | contribs)
Jump to navigation Jump to search

List of items to bring to task meeting

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

{{DiscussTask | explanation}}



ICU Database Task Group Meeting – May 17, 2018

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

1. Continued discussion about items we are counting the number of certain things done: certain radiology tests, blood products, ABG, VBG and echos. Allan reported that the Database Steering Committee endorsed the idea of shifting all these counts over to CCI, in order to also acquire timing and distribution of them in addition to the total counts. In response to continued concern about the extra burden of work from shifting these counts to CCI, we decided that Trish will work with some of the data collectors to obtain a better estimate of the extra work it would entail.

2. There was a long, follow-up discussion about if/how to code “bed borrowing”.

  • Allan made the following suggestion of “machinery” to handle bed borrowing:
    • There are 2 sorts of bed borrowing when a TypeA patient should be in UnitA but is boarded in UnitB
      • (1) Patient is in Unit B but still under the care of TeamA -- this is coded in the database as a UnitA patient
      • (2) Patient is in Unit B and under the care of TeamB -- this is coded in the database as a UnitB patient
    • Here’s what they’re borrowing:
      • For ‘1’ the patient is just borrowing a bed
      • For ‘2’ the patient is borrowing a bed AND the service, i.e. TeamB
    • To capture all of this, we need to create 2 PAIRS of parameters -- that can be coded at a given time each and every day (e.g. noon). The pairs are a Flag + a FlaggedUnitcode
      • (a) Flag1 indicates that on the given day the patient is borrowing a bed -- the Unit code indicates WHERE that bed is.
      • (b) Flag2 indicates that on the given day the patient is borrowing a bed & service -- the Unit code indicates WHERE that patient would otherwise be located
    • Example: A=ACCU B=ICMS
      • In situation#1, set Flag1=1 and FlaggedUnitCode=ICMS
      • In situation#2, set Flag2=1 and FlaggedUnitCode=ACCU
  • Further discussion highlighted that the same information will likely be possible with identifying one of the 2 parameter pairs indicated above, but ALSO including a date/time --- so that instead of recording the bed borrowing situation each day, when either of the 2 bed borrowing scenarios occurs, we code the date/time of it BEGINNING and date/time of it ENDING. Allan will further ponder this alternative and we will discuss more at our next Task Group meeting.

3. Discussion about keeping track of amounts of blood products transfused. We made the following decisions, after Allan spoke to Ryan Zarychanski:

  • We will NOT track transfusion of WBC, whole blood, or stem cells
  • For PRBC, FFP and Platelets: We will quantify the quantity transfused just as is done by Canadian Blood Services -- i.e. as “units” given, where each unit given is accompanied by a unique STICKER which is placed on the Blood Products Administration Record in hospital charts. Consequences of this are that, going forward:
    • What the blood bank calls a single adult transfusion unit of platelets is what in the past has been called a “five pack”.
    • What the blood bank calls a single unit of FFP is what we will track for that product, NOT the number of liters of the product.
  • For Albumin we will only track whether ANY albumin, of any concentration and in any volume, was given on a given day.

4. New collector issue: If/how/when -- for acute respiratory failure patients who stay in the hospital a long time and appear to transition to chronic respiratory failure -- to code that. Options include Chronic dependence on mechanical ventilator and Respiratory failure (insufficiency), chronic. Allan will ponder this further.

5. New collector issue: To clarify the distinction between Cardiac pressure measurements via a catheter in cath lab - incl right heart cath done for pressures and C.O. and CCI coding for a cardiac catheterization. Allan will ponder this further.

6. Items left over from prior weeks, for Allan and/or Tina to do:

  • To the current list of procedures performed in the unit that will be coded only the first time they’re done, we will add: chest tube, ECMO, Continuous hemodialysis, cardioversion, IABP, temporary pacemaker placement
  • Working on a new PIA
  • Check the editing done on the VAP Wiki page.


Next Task Group Meeting: June 11, 2018 at 11:00 am

ICU Database Task Group Meeting – May 9, 2018

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

1. Continued discussion about items we are counting the number of certain things done: certain radiology tests, blood products, ABG, VBG and echos. Allan reported that the Database Steering Committee endorsed the idea of shifting all these counts over to CCI, in order to also acquire timing and distribution of them in addition to the total counts. In response to continued concern about the extra burden of work from shifting these counts to CCI, we decided that Trish will work with some of the data collectors to obtain a better estimate of the extra work it would entail.

2. There was a long discussion about if/how to code “bed borrowing”.

  • This takes two separate forms: (i) patient of type A who “should” be in unitA, but there’s no capacity there, so pt is put in UnitB but continues under the care of TeamA, (ii) patient of type A who “should” be in unitA, but there’s no capacity there, so pt is put in UnitB and is under the care of TeamB. A starting point for this is that in both cases, the patient is unit the patient is assigned to in the database is according to the Medical Team taking care of them, not their physical location.
  • Although this has arisen mainly as regards A=cardiology, this can happen more generally, and we would like (if possible), a generic solution.
    • This has been discussed a lot in the past, but abandoned because of the complexity involved with switches of service and switches of physical location.
  • Allan will work on this problem and report back to the group.

3. In response to issues raised by Laura about therapeutic procedure coding, we decided that to the current list of procedures performed in the unit that will be coded only the first time they’re done, we will add: chest tube, ECMO, Continuous hemodialysis, cardioversion, IABP, temporary pacemaker placement'

4. Items left over from prior weeks, for Allan to do:

  • Working on a new PIA
  • Check the editing done on the VAP Wiki page.
  • Add TPN to CCI picklist

5. We further discussed the Medical/Surgical/Cardiac coding of reason for admission. We agreed that the exact purpose of use of this delineation would dictate the algorithm used for assigning it. Since we are unsure if/how this is being used, before going further we will discuss at the Steering Committee.

Next Task Group Meeting: May 17, 2018 at 10:00 pm


ICU Database Task Group Meeting – April 25, 2018

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

1. Multifactorial discussion about Julie’s reporting of top diagnoses.

  • After discussion we agreed that when the #1 priority Admit Diagnosis is a combined code (i.e. 2 or more linked codes), we will have the data collectors check the box Dx Primary of the combined codes is Primary Admit Diagnosis.
  • For reporting to Medicine, Julie will choose that code to report. For reporting to ICU, she will do as she always does, including ALL admit diagnoses in the algorithm to identify the APACHE II diagnosis category that is “worst”, i.e. associated with the highest hospital mortality.

2. Translation of Dx Data from CCMDB schema to ICD10/CCI - Allan reported that when queried about whether to “map” the old coding schema to the new one, Paunovic, Hajadiacos, Zarychanski, and Walker-Tweed all agreed that rather than mapping (which will require users using data that spans the transition to be familiar with the limits of the mapping), we will code pre-change data using the old schema and post-change data using the new schema.

3. Converting Lab Counts to CCI - Discussion about items we are counting the number of. These are presently: certain radiology tests, blood products, ABG, VBG and echos. The question is whether or not to change all of these simple counts to CCI coding. The issue is to continue “pure counting” where we just record the number over the entire admission vs. using CCI to code the timing of them all.

  • A preliminary report was that Laura (absent today) did a comparison of “pure counting” vs. using CCI to code ABGs -- and that the latter took MUCH more time. However, details weren’t clear and before we make a decision we need to obtain them from Laura. Thus, we’ll discuss more next time.

4. Outreach efforts - After discussion we agreed that Allan will offer at the next Dept of Medicine Executive Committee that he and Julie will offer to provide a 15 minute summary to each section in Internal Medicine, at their section meetings, of the 2 databases.

5. Discussion of OIT Symposium poster. We agreed that the best use of this outreach would be to include a few examples of the data available, and specifically how it can be used for QI.

6. Further discussion of the pathogen list. We decided that:

7. Items left over from prior weeks, for Allan to do:

  • Working on a new PIA
  • Check the editing done on the VAP Wiki page.
  • Add TPN to CCI picklist

8. Registry Patient Type - We further discussed the Medical/Surgical/Cardiac coding of reason for admission. We agreed that the exact purpose of use of this delineation would dictate the algorithm used for assigning it. Since we are unsure if/how this is being used, before going further we will discuss at the Steering Committee.

Next Task Group Meeting: May 9, 2018 at 1:00 pm

ICU Database Task Group Meeting – April 4, 2018

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

1. Regarding Chapter 1 of CCI. After discussion, we agreed that while we still do NOT want to extend our CCI coding to the 3rd level (i.e. how the thing was done), we do have a need to dig deeper into the 2nd level category of “Pharmacotherapy”. Specifically, we at least want to have a subcategory of “Pharmacotherapy, thrombolytic agent”. And furthermore, there are >50 other existing subcategories of Pharmacotherapy that we may want to use in the future. Thus, for now we agreed we’ll change the name of Pharmacotherapy to “Pharmacotherapy NOS” and add to the 2nd level Pharmacotherapy, thrombolytic agent. ( Has been added. Ttenbergen 12:48, 2018 April 11 (CDT))

2. Regarding the ICD10 item U22.9 Infectious organism, unknown we recognized that it includes two situations, that previously were coded separately: (a) cultures sent but were negative, and (b) no cultures sent. Allan will modify the wiki page to reflect this explicitly.

3. Regarding picklists for ICD10. We agreed that we will NOT have such picklists. But given the particular issues with Hemorrhage or Bleeding in ICD10 When There is No Specific Code such as retroperitoneal hemorrhage, Allan will make a wiki page that discusses how to use combined codes to make these entities.

4. Agreed that we DO want the CCI Picklist. Next time we’ll discuss it.

5. Discussion about how to organize ICD10 codes for Julie’s regular reports (Reporting from ICD10/CCI). The issue is that she reports the distribution of the main admission diagnosis, but that going forward some entities cannot be captured with a single ICD10 code. For example, retroperitoneal hemorrhage is coded as a combination of Hemorrhage, NOS and Retroperitoneal area, diagnostic imaging, abnormal. This leads to uncertainty about which of a combination entity to report.

  • We recognized 4 main options:
    • (i) Allow the data collectors to make a judgement about which of the combined ICD10 codes is “worst” and use that as the main diagnosis, or (ii) allow
    • (ii) Allow multiple ICD10 codes to “jointly” be identified as the “main diagnosis” -- so that the denominator for reporting will no longer be patients, but will be the number of main diagnoses. Furthermore, the delineation of combined ICD10 codes mutually coded as the main diagnosis will not “remain together” in the reporting.
    • (iii) Allow multiple ICD10 codes to “jointly” be identified as the “main diagnosis” -- but here report them as the combination. The problem here is that there will be so many combinations that the reporting may become untenable.
    • (iv) Report main diagnoses by ICD10 blocks not individual ICD10 diagnoses. This solution still will need to deal, however, with combination diagnosis reporting.
  • It was agreed that the users of these routine reports must help us understand what they need/want. To this end, Allan will discuss this with Bojan (and possibly the group of ICU medical directors), and we’ll bring it to the Database Steering Committee.

6. Discussion about items we are counting the number of (see Converting Lab Counts to CCI). These are presently: certain radiology tests, blood products, ABG, VBG and echos. The question is whether or not to change all of these simple counts to CCI coding.

  • It will be more work, but by including dates it will dramatically improve the ability of the database to be used for assessing the course of care.
  • We decided that Allan will discuss this with Bojan, and we’ll bring it to the Database Steering Committee.

7. Allan has mapped the Nephro as “reasons for CRRT” to their ICD10 equivalents. He sent it out and next time we’ll discuss it, including how and where (HSC only or also St. B) to implement it.

8. Items left over from prior weeks, for Allan to do:

  • Working on a new PIA
  • Check the editing done on the VAP Wiki page.
  • Add TPN to CCI picklist

9. We further discussed the Medical/Surgical/Cardiac Registry Patient Type coding of reason for admission. We agreed that the exact purpose of use of this delineation would dictate the algorithm used for assigning it. Since we are unsure if/how this is being used, before going further we will discuss this with the Steering Committee.

Next Task Group Meeting: April 11, 2018 at 2:00 pm

ICU Database Task Group Meeting – March 14, 2018

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

1. Discussion of wording for GI bleeding. After discussion we agreed:

  • We will alter K92.1 to be simply the symptom of “Melena or Hematochezia”, without any mention of site of bleeding. Likewise we’ll alter K92.0 to be simply the symptom of “Hematemeis” without mention of the site of bleeding. Allan has altered the Wiki pages, but Tina will make the necessary name changes.
  • Along with this, the instructions for coding GI bleeding will be:
    • If the site/cause is known, code that. It is OK, but not required in that case to code the symptom (Hematemsis vs. Melena or Hematochezia).
    • If the site/cause is unknown, then use code K92.2 (Gastrointestinal hemorrhage (GI bleed), not specified if lower or upper), and again it’s OK but not required to code the symptom.

2. Regarding query about Status asthmaticus. In the Canadian version, this is J45.91, not J46, which doesn’t actually exist in the Canadian version.

3. Followup from last meeting’s discussion of organisms (buglist).

  • New invented codes for bugs we want to code but don’t exist in ICD-10
    • U22.11 Stenotrophomonas species
    • U22.12 Citrobacter species
    • U22.13 Enterobacter species
    • U22.14 Acinetobacter species
  • New code invented for when the bug causing infection is unknown
      • U22.9 Infectious organism, unknown
  • We agreed to change the name of the bug code Klebsiella pneumoniae to Klebsiella species, to include all types of that genus (tho about 90% are pneumoniae)

4. There are a number of ICD10/CCI items for which Tina and Allan will meet separately from the Task Group to go over. Tina will arrange a time.

5. Allan has created a picklist of the ICD10 versions of the 100 or so most common ICU diagnoses. He will send it out and we’ll talk next time about which to include, and how to do so.

6. Allan has mapped the Nephro as “reasons for CRRT” to their ICD10 equivalents. He will send it out and we’ll talk next time about how and where (HSC only or also St. B) to implement it.

7. Items left over from prior weeks, for Allan to do:

  • Working on a new PIA
  • Check the editing done on the VAP Wiki page.

8. Julie identified that there has been a request to track use of TPN, not only in ICUs (where it’s part of TISS), but also on wards. We identified that we can easily add this as a therapeutic procedure (CCI) as its CCI code is 1.LZ.35.HH-C6

9. At Julie’s request, we agreed to have separate “location” listings for the Invasive Radiology at HSC vs. St. B

10. We further discussed the Medical/Surgical/Cardiac coding of reason for admission. We agreed that the exact purpose of use of this delineation would dictate the algorithm used for assigning it. Since we are unsure if/how this is being used, before going further we will discuss it at the next Steering Committee meeting.

  • Next Task Group Meeting: April 4, 2018 at 11:00 am

ICU Database Task Group Meeting – March 6, 2018

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

1. Issues related to organisms (buglist).

  • First, to deal with the fact that various users need/want to be able to code some bacteria that don’t have specific ICD10 codes, we agreed to add the ones requested (Stenotrophomonas, Citrobacter, Enterobacter, Acinetobacter) but that we will need to “invent” some new ICD10 codes. To distinguish these, we’ll use the single unused chapter letter code prefix (U) for any/all ICD10 codes we need to invent. Allan will create codes for these 4 bugs.
  • There was a discussion about how to differentially handle the situation where: (a) bug is known but it’s not on our list of specific bugs, vs. (b) the bug is unknown.
    • It was decided that when bugs are known but not specified on our bug list, we’ll use one or another of the existing wastebasket codes in ICD10:
      • Bacteria, NOS
      • Virus, NOS
      • Fungus or yeast, NOS
      • Infectious disease NOS OR for buglist organism NOS
    • But when the bug causing infection is unknown, we’ll “invent” a new ICD10 code for that situation, which will be called “Infectious organism, unknown”

2. Continued discussion about CKD coding.

  • After discussion with some data collectors, it turns out that obtaining heights and weights on everybody is not feasible. Thus calculating Creatinine clearance will not be feasible either.
  • For some, but not all, CKD patients, the Nephrology consultant writes the Creatinine clearance in the chart -- these will be the only ones we will have available.
  • We decided that we’ll code CKD as follows:
    • When there is a Creatinine clearance listed, it will be used to specify between Stages 1, 2, 3, 4, or 5.
    • When no Creatinine clearance is listed, but the patient is a known dialysis patient, we will identify him/her as CKD, Stage 5
    • For all other CKD patients, i.e. those in whom we cannot easily identify the Stage, we’ll use the wastebasket code Chronic kidney disease, NOS (stage unspecified).

3. Regarding Sepsis coding, specifically whether to use the new Sepsis-3 definitions, and if so, when to start.

  • Laura has done some Sepsis-3 coding and reports it is doable.
  • After discussion, it was agreed that we WILL aim to move to Sepsis-3 identification of Sepsis and Septic shock, but that we will only implement this about 6 months after we introduce ICD10/CCI.

4. There are a number of ICD10/CCI items for which Tina and Allan will meet separately from the Task Group to go over. Tina will arrange a time.

5. Items left over from prior weeks, for Allan to do:

  • Make a picklist of the most common 20-50 ICD10 codes, for convenience
  • Identify the ICD10 code(s) that correspond to the diagnoses used by Nephro as “reasons for CRRT” for this purpose -- DONE, but still need to work out how to ensure that:
    • Whatever they code should find it’s way into our general list of diagnoses.
    • Towards this end, we agreed that that/those code(s) should be duplicated: (a) in the Temp file for the special purpose of identifying why they were on CRRT, and (b) in our general list of diagnoses.
  • Working on a new PIA
  • Check the editing done on the VAP Wiki page.

Next Task Group Meeting: March 14, 2018 at 11:00 am


ICU Database Task Group Meeting – February 12, 2018

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

1. Collector issue: How to code HAP, VAP and CAP in ICD10. Answers from Allan:

2. Collector issue: Metabolic derangements in ICD10.

  • We agreed that we will transfer the same thresholds used before for ICD10, but as these criteria are very strict (e.g. it’s currently not hyperkalemia unless K>6.5), we’ll add to the criteria by saying that a “hyper” or “hypo” type of metabolic derangement (specifically for Na, K, Ca, Mg, PO4, Glucose) will be defined going forward as EITHER: (a) value outside the bounds given OR, (b) was treated. Laura will make these changes to the various Wiki articles.

3. Collector issues: 3 related issues about organisms.

  • First is bugs not currently included in the “buglist” in ICD10.
    • There might be some relevant bugs not specifically code in ICD10. For example, Stenotrophomonas multophilia. We can add non-standard codes for these if we need to. To figure out what additional specific bugs we should add, we agreed:
      • Julie will send Allan a list of bug frequencies over the past 3 years, both overall and specific to VAP. These will inform specific bugs we should consider adding.
      • Allan will contact Rob Ariano regarding the VAP buglist he needs for pharmacy purposes in VAP -- done, and the bugs he'd like us to add are: Stenotrophomonas, Klebsiella, Serratia, Citrobacter, Enterobacter, Acinetobacter.
  • Second is whether to differentiate between “no cultures done” and “no positive cultures”
    • From an operational, database viewpoint, these are not different -- i.e. in both cases we have “bug not known”. As Julie told us that nobody has ever asked for a distinction between these two, we agreed that we will group these together.
  • Third is whether and how to differentiate between “bug not known” and “bug known but not in our list of specific bugs”.
    • This latter is necessary because we’ll never have specific codes for all of the thousands of known pathogens.
    • But each of the subcategories of bugs (bacteria, fungi, viruses, mycobacteria) has a “wastebasket” of NOS. THUS, when we have a bacteria, fungi, viruses, or mycobacteria that’s been identified but isn’t on our buglist, we use the appropriate NOS code (e.g. Bacteria, NOS). But when we have “bug not known” we’ll instead use Infectious disease NOS

4. Collector issue: regarding calculation of Creatinine clearance for grading CKD:

  • This requires a calculation involving: serum creatinine, age, Height and weight and sex.
    • Creatinine clearance = 1.257 * Cr^(-1.154) * Age^(-0.203) *(Weight in kg)^0.425 * (Height in cm)^0.725 * sex factor (1 for males, 0.742 for females)
  • This discussion of need for weight and height led to a more general discussion of whether we should start collecting those. Issues included:
    • They’re also needed to calculate BMI
    • They’re much less generally available on wards than ICU
  • We decided that for now we will not start collecting W and H routinely for all patients, but that Tina will create a calculator where the collector inputs serum creatinine, age, weight, height and sex, and it calculates Creatinine clearance for use in assessing the Stage of CKD (note that the creatinine that goes into that calculation should be the patients most recent chronic value).
    • polling of collectors if this information is available from chart, where in chart, if consistently available or not. See: Height and weight

6. Regarding the list of “reasons for CRRT” that is completed by the Nephro team.

  • After discussion, we agreed that whatever they code should find it’s way into our general list of diagnoses.
  • Towards this end, we agreed that that/those code(s) should be duplicated: (a) in the Temp file for the special purpose of identifying why they were on CRRT, and (b) in our general list of diagnoses. To operationalize this, Allan will identify the ICD10 code(s) that correspond to those diagnoses used by Nephro for this purpose.

7. Update on seeking data on PHIN validation data, hospitalization data, and mortality data from WRHA.

  • Allan is working on the new PIA.

8. Regarding Sepsis coding, specifically whether to use the new Sepsis-3 definitions, and if so, when to start.

  • For what was previously called Severe sepsis, the new guidelines merely call “sepsis”, which is identified as acute organ failure caused by proven or presumed infection. The official guidelines operationalize new organ failure as an increase in SOFA score by 2 or more points. Since SOFA scoring is nontrivial, there was concern about whether it is practical for the data collectors to try and do this. Laura will look at the SOFA scoring grid and we’ll discuss this more next time. We’ll also contemplate making the change, but delaying it.

9. Starting with CLI criteria, there was extensive discussion about the difficulties in adhering to criteria-based diagnoses.

  • Allan will check the editing done on this Wiki page.

Next Task Group Meeting: March 5, 2018 at 10:00 am

ICU Database Task Group Meeting – February 1, 2018

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

1. Most of this meeting was consumed discussing implementation of CCI coding for procedures. There was agreement with Allan’s proposal to simplify coding of Therapeutic Procedures by most omitting the field for “how an intervention was done”.

  • Everyone should take a look at the article on Procedure coding in ICD10/CCI’’’ (see CCI Collection instead).

2. Followup on influenza coding

  • Allan reported that Bojan indicated that the needs of the ICU Program can accept the planned alteration of our database definition of Influenza, as long as we also (as planned) also keep track separately of lab testing results and treatment.

3. Regarding the list of “reasons for CRRT” that is completed by the Nephro team.

  • After discussion, we agreed that whatever they code should find it’s way into our general list of diagnoses.
  • Towards this end, we agreed that that/those code(s) should be duplicated: (a) in the Temp file for the special purpose of identifying why they were on CRRT, and (b) in our general list of diagnoses. To operationalize this, Tina will send Allan the Wiki link to this list, and Allan will identify the ICD10 code(s) that correspond to those diagnoses used by Nephro for this purpose.

4. Update on seeking data on PHIN validation data, hospitalization data, and mortality data from WRHA.

  • Allan is working on the new PIA.

5. Regarding Sepsis coding. Allan reported that our coding will follow the new Sepsis-3 definitions.

  • This greatly simplifies identification of Septic shock, eliminates SIRS from all sepsis-related definitions, and eliminates what was previously called “sepsis” i.e. SIRS without acute organ failure.
  • For what was previously called “severe sepsis”, the new guidelines merely call “sepsis”, which is identified as acute organ failure caused by proven or presumed infection. The official guidelines operationalize new organ failure as an increase in SOFA score by 2 or more points. Since SOFA scoring is nontrivial, there was concern about whether it is practical for the data collectors to try and do this. Laura will look at the SOFA scoring grid and we’ll discuss this more next time.

6. Starting with CLI criteria, there was extensive discussion about the difficulties in adhering to criteria-based diagnoses. *Laura clarified that even more than CLI, VAP requires much checking back and forth at data over time. It was decided that we would like to create wiki-based tools that will make it easier for collectors to do these tasks. The tools would incorporate the diagnosis guidelines into some sort of checklist “machinery”. Trish will assign Laura and a few other coders to work on this.

Next Task Group Meeting: February 12, 2018 at 11:30 am


ICU Database Task Group Meeting – January 17, 2018

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

1. Followup on influenza coding

  • There is still considerable uncertainty about what we are doing and what we should be doing. The uncertainty relates to: (a) the fact that there can be three different lab tests done, one of which comes back quickly and on EMR, but the other two are done by Cadham and come back slowly and we do not have electronic access to them, (b) the needs of the WRHA appear to be different than our needs, (c) the fact that, apparently, clinicians sometimes misinterpret the meaning of the lab tests and incorrectly assume that a negative test is incontrovertable proof that influenza is not present.
  • It was agreed that we need to clarify this. Towards that end Allan will talk to Bojan about the needs of the ICU program in this regards; he will talk to the head of Cadham to see if we can develop a direct, electronic conduit to their influenza results.

2. Followup on CCI coding

  • It was agreed that for now we will suspend all activity by the data collectors in trying to do CCI coding.
  • For CCI chapters on: (i) Diagnostic radiology procedures, (ii) Non-radiology diagnostic procedures, (iii) Obstetrical procedures, and (iv) Miscellaneous Immunotherapy procedures -- simplification has been done. Everyone should take a look at the article on Procedure coding in ICD10/CCI’’’ (see CCI Collection instead).
  • Allan is going to work to simplify the CCI chapter on Therapeutic Procedures.

3. Regarding the list of “reasons for CRRT” that is completed by the Nephro team.

  • After discussion, we agreed that whatever they code should find it’s way into our general list of diagnoses.
  • Towards this end, we agreed that that/those code(s) should be duplicated: (a) in the Temp file for the special purpose of identifying why they were on CRRT, and (b) in our general list of diagnoses. To operationalize this, Tina will send Allan the Wiki link to this list, and Allan will identify the ICD10 code(s) that correspond to those diagnoses used by Nephro for this purpose.

4. Update on seeking data on PHIN validation data, hospitalization data, and mortality data from WRHA.

  • Allan is working on the new PIA.

5. Regarding Sepsis coding. Allan reported that our coding will follow the new Sepsis-3 definitions.

  • This greatly simplifies identification of Septic shock, eliminates SIRS from all sepsis-related definitions, and eliminates what was previously called “sepsis” i.e. SIRS without acute organ failure.
  • For what was previously called “severe sepsis”, the new guidelines merely call “sepsis”, which is identified as acute organ failure caused by proven or presumed infection. The official guidelines operationalize new organ failure as an increase in SOFA score by 2 or more points. Since SOFA scoring is nontrivial, there was concern about whether it is practical for the data collectors to try and do this. It was decided that Laura and some other ICU coders will look at the SOFA scoring grid and we’ll discuss this more next time.

6. Starting with CLI criteria, there was extensive discussion about the difficulties in adhering to criteria-based diagnoses.

  • Laura clarified that even more than CLI, VAP requires much checking back and forth at data over time. It was decided that we would like to create wiki-based tools that will make it easier for collectors to do these tasks. The tools would incorporate the diagnosis guidelines into some sort of checklist “machinery”. Trish will assign Laura and a few other coders to work on this.

Next Task Group Meeting: February 1, 2018 at 11:00 am


Database Task Group Meeting December 20, 2017

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

1. Follow-up on seeking eChart access for coder. Allan reported that after talking to a high-level administrator at Manitoba Health, that indeed we are not going to be allowed to get eChart access.

2. Regarding influenza coding. After a discussion we agreed on the following aspects of coding:

  • Unrelated to ICD-10, we will continue to record whether:
    • A lab test for influenza was done and whether it was positive or negative. If more than one test was done (at different times, or by different labs), it is considered positive during flu season IF ANY of them are reported as positive. Off of flu season it’s considered as positive only if the FINAL CADHAM result was positive (e.g. off of flu season flu tests done by the separate hospital labs are ignored).
    • Whether a “full course” of treatment was given for influenza. The usual course of oseltamivir (Tamiflu) is 5 days. But WHO and CDC recommend that in severe cases the drug should be continued until the infection is resolved or there is satisfactory clinical improvement.
  • Based on this, ICD-10 coding for influenza will be done as follows:
    • Use this website https://www.gov.mb.ca/health/publichealth/surveillance/influenza/index.html to iIdentify whether we’re in flu season or not.
    • During established flu season a person will be considered to have influenza if either of the following is true: (i) any lab test done for influenza was positive, or (ii) the patient was believed the the medical team to have influenza and given a full course of treatment (which could possibly have concluded after leaving the hospital).
    • Off flu season (before there is any reported flu in the province, and after the flu season has been declared to be over) a person can ONLY be diagnosed as having influenza if the FINAL CADHAM test result is positive. In the absence of such a final Cadham result, clinical suspicion, treatment for flu, and positive rapid tests will be considered as NOT INFLUENZA.
  • Allan put all this info into the wiki article called “Influenza in ICD10” -- done.

3. Further discussion of the ICD10 diagnosis of Palliative care.

  • We agreed to operationalize using the ICD10 code of Palliative care as any of the following 4 items:
    • 1. ACP-C status
    • 2. Had been on palliative care ‘’prior’’ to this hospital admission (i.e. at home or in the care facility)
    • 3. Is receiving active palliation. What is meant by this is (again) related to the intent of care --- so receiving aggressive symptom control measures (e.g. a morphine drip) does not consitute active palliation UNLESS the intent of the overall care at this point is control of symptoms and not cure or even prolongation of life.
    • 4. The Palliative Care Service (physician group) is seeing the patient in an ongoing fashion. This means that they have seen the patient at least twice during this admission, or that they wrote that they intended to follow but the patient died or left hospital before they could be seen a second time. Thus, if that consult team saw the patient in an initial consult but didn’t or didn’t plan to follow them longitudinally, then this item doesn’t apply.
  • Allan will modify the wiki article Palliative care to reflect this -- done.

4. Regarding discharge planning, paneling, etc.

5. Regarding the list of “reasons for CRRT” that is completed by the Nephro team.

  • After discussion, we agreed that whatever they code should find it’s way into our general list of diagnoses.
  • Towards this end, we agreed that that/those code(s) should be duplicated: (a) in the Temp file for the special purpose of identifying why they were on CRRT, and (b) in our general list of diagnoses. Allan will identify the ICD10 code(s) that correspond to those diagnoses used by Nephro for this purpose.

6. Switching over to ICD-10

  • Allan is continuing work on making a set of “quick codes” for both ICD-10 and CCI, that can be chosen as unique entities, to represent procedures that are common, or difficult to otherwise figure out how to code.
  • Regarding the list of acquired complications that previously were singled out to ensure that they were coded: Trish spoke with collectors who indicated that this list is of “high profile” entities that they don’t really think they need reminding about. So, we will bring this information back to the Steering Committee, with a recommendation that there is no need to continue using this list.

7. Regarding CCI.

  • Allan presented a greatly simplified version of CCI chapter 3 (Diagnostic imaging procedures). See new wiki article Procedure coding in ICD10/CCI’’’ (see CCI Collection instead) for evolving details.
  • We recognized that for all procedures an option (to be discussed more later) is to allow a “number done” to be associated with each procedure coded for each date. Thus, for example, if 4 CXRs were done on a given day, one would only need to enter it once, but with that count saved with the item (with default being 1). We will discuss this more later, after we’ve worked through all the chapters of CCI.
  • We began what will be a series of conversations about WHICH procedures we want to code, and creating guidelines/rules around them. We identified the benefits of simple rules, and the dynamic that goes along with simplicity is that sometimes we will collect procedures that we’re not that interested in. We also agreed that the same rules should apply to ICU and wards. Today we discussed Therapeutic Procedures. In coming meetings we will discuss Diagnostic Procedures. And after all this is done, we must decide whether, in addition to the picklists, to ALSO allow for menu-based coding of other procedures.
  • We agreed that a good goal would be to be able to include all the current items on the Tasks in CCI, and so be able to dispense with the Tasks (HD, PD, trach, isolation, NIV).
  • We still need to clarify isolation. Currently we are recording 3 different levels of isolation, and none of these relate to use of a negative pressure room. Allan will talk to Bojan and Nick to clarify the needs here.

8. Update on seeking data on PHIN validation data, hospitalization data, and mortality data from WRHA.

  • Allan is working on the new PIA.

9. Regarding Sepsis coding. Allan reported that our coding will follow the new Sepsis-3 definitions.

  • This greatly simplifies identification of Septic shock, eliminates SIRS from all sepsis-related definitions, and eliminates what was previously called “sepsis” i.e. SIRS without acute organ failure.
  • For what was previously called “severe sepsis”, the new guidelines merely call “sepsis”, which is identified as acute organ failure caused by proven or presumed infection. The official guidelines operationalize new organ failure as an increase in SOFA score by 2 or more points. Since SOFA scoring is nontrivial, there was concern about whether it is practical for the data collectors to try and do this. It was decided that Laura and some other ICU coders will look at the SOFA scoring grid and we’ll discuss this later.

10. Starting with CLI criteria, there was extensive discussion about the difficulties in adhering to criteria-based diagnoses. *Laura clarified that even more than CLI, VAP requires much checking back and forth at data over time. It was decided that we would like to create wiki-based tools that will make it easier for collectors to do these tasks. The tools would incorporate the diagnosis guidelines into some sort of checklist “machinery”. Trish will assign Laura and a few other coders to work on this.

Next Task Group Meeting: January 4, 2018 at 10:00 am

Database Task Group Meeting December 20, 2017

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

1. Follow-up on seeking eChart access for coders

  • Allan and Jodi Walker-Tweed have sent a letter to Christina Van Schindler, the WRHA Chief Privacy Officer, asking for permission. Awaiting a reply.

2. Regarding influenza coding.

  • We had previously discussed two possible categories of influenza: (1) lab-confirmed and (2) suspected. But in further review of IDSA and CDC recommendations, it’s fairly complicated. The value of the tests in helping us figure out whether a person has influenza depends on: (a) how long after onset of symptoms the test was done, (b) whether the sample was upper or lower respiratory, and (c) whether the test is done in the midst of flu season, at the start of flu season, at the end of flu season, or not at all during flu season. Given all of the variation in lab interpretation, the idea of lab-confirmed vs. suspected is not being used.
  • Here is a workable approach:
    • During established flu season (defined as there has been laboratory-confirmed flu in the community), the diagnosis of influenza is mainly clinical. So, lab confirmation of any type is NOT needed to make the diagnosis. If the team says they think it’s flu, and they’re treating with a full course of tamiflu, then take that as true influenza, regardless of the results of any influenza tests, done in hospital labs or Cadham.
    • Completely off flu season (warm months, before there is any reported flu in the province, and after the flu season has been declared to be over) a person can ONLY be diagnosed as having influenza if the FINAL CADHAM test result is positive. In the absence of such a final Cadham result, clinical suspicion, treatment for flu, and positive rapid tests will be considered as NOT INFLUENZA.
    • The hard part is right at the start and end of typical flu season. In this case, use the following algorithm:
      • One or more lab tests were done (including rapid tests in hospital labs, and Cadham tests) and ‘’’any’’’ of them were positive -- then consider influenza to be present.
    • No lab tests were done, but the team says they think it’s flu, and they’re treating with a full course of tamiflu, then take that as influenza being present.
  • Allan put all this info into the wiki article called “Influenza in ICD10” -- done. Everyone is invited to take a look at this.

3. Further discussion of the ICD10 diagnosis of Palliative care.

  • This is different from the Palliative Care Service -- that refers to a group of physicians. Palliative care refers to whether the clinical plan for the patient is to provide comfort towards the end of life, not to prolong life.
  • Per Julie, the way this is being used is for reporting death rates -- specifically to remove people getting Palliative care from the denominator and numerator of death rates.
  • Accordingly, to figure out if a person should have the ICD10 diagnosis of Palliative care, we must figure out the INTENT of care. If the intent is aimed at cure and prolonging life, then the person is not in Palliative care. If the intent is primarily control of symptoms (whether the person currently has symptoms or not) and not cure or even prolongation of life, then the person is in Palliative care.
  • We will operationalize Palliative care as any of the following 4 items:
    • 1. ACP-C status
    • 2. Had been on palliative care ‘’prior’’ to this hospital admission (i.e. at home or in the care facility)
    • 3. Is receiving active palliation. What is meant by this is (again) related to the intent of care --- so receiving aggressive symptom control measures (e.g. a morphine drip) does not consitute active palliation UNLESS the intent of the overall care at this point is control of symptoms and not cure or even prolongation of life.
    • 4. The Palliative Care Service (physician group) is seeing the patient in an ongoing fashion, which means has seen them at least twice during this admission. So, if that consult team saw the patient in an initial consult but isn’t following them longitudinally, then this item doesn’t apply.
  • Allan will modify the wiki article Palliative care to reflect this -- done. Everyone is invited to take a look at this.

4. Regarding discharge planning, paneling, etc.

  • Discussion identified that there are reasons to separately classify time in hospital spent waiting to go to: long-term care facilities, home. Furthermore, with the advent of Riverridge 2, and the transition that is currently taking place in the WRHA to convert the Vic, Conc and Oaks to transition hospitals, there may be a near-future need by the Internal Medicine Program to know about time spent waiting to transfer to those. This all comes up because as of now, we have only a single ICD10 code to cover all of this. In the current database we track these via Temp, but the goal in general is (where possible) to incorporate Temp information into ICD10 and CCI codes.
  • Allan will see how we might do this in ICD10 -- here is a solution created by adapting a set of existing ICD10 codes:

Allan has put these into the Wiki. Everyone is invited to take a look.

5. Regarding the list of “reasons for CRRT” that is completed by the Nephro team.

  • After discussion, we agreed that whatever they code should find it’s way into our general list of diagnoses.
  • Towards this end, we agreed that that/those code(s) should be duplicated: (a) in the Temp file for the special purpose of identifying why they were on CRRT, and (b) in our general list of diagnoses. To operationalize this, Tina will send Allan the Wiki link to this list, and Allan will identify the ICD10 code(s) that correspond to those diagnoses used by Nephro for this purpose.
    • Link was provided and Allan replied in email 2017-11-18; contents integrated into CRRT#ICD10. Further discussion required since some did not have unambiguous equivalencies.

6. Switching over to ICD-10

  • Allan is continuing work on making a set of “quick codes” for both ICD-10 and CCI, that can be chosen as unique entities, to represent procedures that are common, or difficult to otherwise figure out how to code.
  • Regarding the list of acquired complications that previously were singled out to ensure that they were coded: Trish spoke with collectors who indicated that this list is of “high profile” entities that they don’t really think they need reminding about. So, we will bring this information back to the Steering Committee, with a recommendation that there is no need to continue using this list.

7. Regarding CCI.

  • It was agreed that for simplicity & consistency of coding, that for the most common and tricky codes we need a picklist. Allan will work on that.
  • We began what will be a series of conversations about WHICH procedures we want to code, and creating guidelines/rules around them. We identified the benefits of simple rules, and the dynamic that goes along with simplicity is that sometimes we will collect procedures that we’re not that interested in. We also agreed that the same rules should apply to ICU and wards. Today we discussed Therapeutic Procedures. In coming meetings we will discuss Diagnostic Procedures. And after all this is done, we must decide whether, in addition to the picklists, to ALSO allow for menu-based coding of other procedures.
  • We agreed that a good goal would be to be able to include all the current items on the Tasks in CCI, and so be able to dispense with the Tasks (HD, PD, trach, isolation, NIV).
  • After discussion about rules around which Therapeutic Procedures (chapter 1 of CCI) to collect, we came up with:
    • Include all therapeutic procedures done outside the patient’s unit
    • Include all therapeutic procedures done using an endoscope (whether inserted through an orifice, incision or wound)
    • Code the following therapeutic procedures done in the patient’s unit -- but only the FIRST time it was done during the patient’s stay on that unit:
      • arterial catheter placement
      • PEG
      • hemodialysis
      • peritoneal dialysis
      • plasmapheresis
      • non-invasive mechanical ventilation -- includes CPAP, BiPAP, and classic NIV (where a mask is connected to a regular ventilator)
      • debridement
      • tracheostomy placement (i.e. done bedside)
  • We still need to clarify isolation. Currently we are recording 3 different levels of isolation, and none of these relate to use of a negative pressure room. Allan will talk to Bojan to clarify the needs here.

8. Update on seeking data on PHIN validation data, hospitalization data, and mortality data from WRHA.

  • Allan is working on the new PIA.

9. Regarding Sepsis coding. Allan reported that our coding will follow the new Sepsis-3 definitions.

  • This greatly simplifies identification of Septic shock, eliminates SIRS from all sepsis-related definitions, and eliminates what was previously called “sepsis” i.e. SIRS without acute organ failure.
  • For what was previously called “severe sepsis”, the new guidelines merely call “sepsis”, which is identified as acute organ failure caused by proven or presumed infection. The official guidelines operationalize new organ failure as an increase in SOFA score by 2 or more points. Since SOFA scoring is nontrivial, there was concern about whether it is practical for the data collectors to try and do this. It was decided that Laura and some other ICU coders will look at the SOFA scoring grid and we’ll discuss this more next time. -- Tabled today since Laura is not at this meeting.

10. Starting with CLI criteria, there was extensive discussion about the difficulties in adhering to criteria-based diagnoses. *Laura clarified that even more than CLI, VAP requires much checking back and forth at data over time. It was decided that we would like to create wiki-based tools that will make it easier for collectors to do these tasks. The tools would incorporate the diagnosis guidelines into some sort of checklist “machinery”. Trish will assign Laura and a few other coders to work on this.

Next Task Group Meeting: January 4, 2018 at 10:00 am



Database Task Group Meeting December 8, 2017

  • Present: Allan Garland, Con Marks, Julie Mojica, Laura Kolesar, Tina Tenbergen
  • Absent: none Minutes prepared by: AG Action items in BOLD

1. Regarding the influenza coding:

  • We discussed the possibility of going forward with 2 categories of influenza: (1) lab-confirmed and (2) suspected.
    • For lab-confirmed: we will use only the FINAL CADHAM result for this. Thus, we will have to wait for it to be finalized before we can finalize influenza coding. Specifically, we will NOT use the hospital lab rapid influenza test results, or the preliminary Cadham result. Cadham results are on eChart, but not on the regular EMR lab computer. Allan will communicate with the WRHA privacy office to try and get all the collectors eChart access.
    • For suspected: here the clinical team wrote believed that it was influenza, gave a “full course” of anti-influenza drug PLUS either: (a) no swab was sent at all, or (b) swab was sent and FINAL CADHAM result negative. Before we finalize this concept, Allan will talk to infection control.
      • Regarding a “full course of tamiflu”: The usual course of oseltamivir is 5 days. But WHO and CDC recommend that in severe cases the drug should be continued until the infection is resolved or there is satisfactory clinical improvement.
  • We will talk about this more later, and finish the wiki article called “Influenza in ICD10”.

2. Agreed that our new goal for implementing ICD10 and CCI is April 1, 2018.

3. Agreed that high-flow oxygen will NOT be considered to be CPAP or any form of non-invasive ventilation.


ITEMS FROM PRIOR MEETING THAT WE DIDN’T GET TO YET


4. Regarding Palliative care/palliative service. There is an ICD10 code for Palliative care. Allan will update the Wiki on this code, to clarify that by this code we mean a person is getting active palliative care, which may or may not concide with being ACP-C, or having a consult from the Palliative care service.

5. Regarding the list of “reasons for CRRT” that is completed by the Nephro team.

  • After discussion, we agreed that whatever they code should find it’s way into our general list of diagnoses.
  • Towards this end, we agreed that that/those code(s) should be duplicated: (a) in the Temp file for the special purpose of identifying why they were on CRRT, and (b) in our general list of diagnoses. To operationalize this, Allan will identify the ICD10 code(s) that correspond to those diagnoses used by Nephro for this purpose.

6. Regarding Sepsis coding. Allan reported that our coding will follow the new Sepsis-3 definitions.

  • This greatly simplifies identification of Septic shock, eliminates SIRS from all sepsis-related definitions, and eliminates what was previously called “sepsis” i.e. SIRS without acute organ failure.
  • For what was previously called “severe sepsis”, the new guidelines merely call “sepsis”, which is identified as acute organ failure caused by proven or presumed infection. The official guidelines operationalize new organ failure as an increase in SOFA score by 2 or more points. Since SOFA scoring is nontrivial, there was concern about whether it is practical for the data collectors to try and do this. It was decided that Laura and some other ICU coders will look at the SOFA scoring grid and we’ll discuss this more next time.

7. Switching over to ICD-10

  • A number of coders, Tina and Allan are working on Wiki articles for the new ICD10 coding schema.
  • Allan is continuing work on making a set of “quick codes” for both ICD-10 and CCI, that can be chosen as unique entities, to represent procedures that are common, or difficult to otherwise figure out how to code.
  • Regarding the list of acquired complications that previously were singled out to ensure that they were coded: Trish spoke with collectors who indicated that this list is of “high profile” entities that they don’t really think they need reminding about. So, we will bring this information back to the Steering Committee, with a recommendation that there is no need to continue using this list.

8. Regarding switching over to CCI

  • 'Allan is continuing work on CCI coding.
  • There was discussion about whether procedure coding should comprise: (a) fully flexible coding whereby every procedure is constructed by data collectors as a combination of: body part + what was done to it + how it was done, (b) ready-make CCI codes that have been pre-constructed from the native components, or (c) a combination of ready-made codes for the most common procedures, augmented by component construction of codes for the less common procedures. Towards making this decision Allan will send Tina lists of the components to peruse and start sharing with data collectors.

9. Update on seeking data on PHIN validation data, hospitalization data, and mortality data from WRHA.

  • Allan is working on the new PIA.

10. Regarding CLI and VAP criteria: Laura has reworked the articles. Allan will take a look at them.

Next Task Group Meeting: December 20, 2017 at 11:00 am