Requested CCMDB changes for the next version

see the Development Documentation Category for other development logs

Bug Fixes

all addressed for now. Ttenbergen 09:57, 28 October 2010 (CDT)

Functional Improvements

Implement mean BP calculation tool on laptop

It would be beneficial to put in a mean BP helper (like the ABG helper) so that we could enter our systolic and diastolic BP's and it would automatically calculate the mean BP and perhaps even indicate the higher apache score and then we could pick the worst one. This would make apache scores more accurate.

Full Apache Score

Add full apache score to the Viewer form

• That one would need thorough testing to make sure I got the score right, e.g. enter 10 patients and check if their apache scores in ccmdb are what you would have calculated. Any volunteers for the testing? If it won't be validated I'd rather not add this to confuse anyone. Ttenbergen 11:05, 9 June 2010 (CDT)
  • Marie Laporte offered to help with this. I will let her know once I have implemented this. Ttenbergen 16:01, 2 July 2010 (CDT)

Background/IT/Coding Improvements

Logging errors in the CCMDB.mdb

Add a table to the CCMDB that records every time an error is generated with the following details:

• Date/time of error
• patient ID (Hosp Loc Serial)
• Field Identifier
• error descriptor (text)
• erroneous value

Make that table sync to a central repository on send.

Changing the local data structure automatically

Need a way to make changes to data that is stored in \ccmdb_data\ccmdb.mdb, such as adding room numbers. By Access design, \ccmdb_program\ccmdb.mdb can not touch a linked table's structure. Will need to copy down new data structure and transfer existing data into that. This gives the opportunity to rename the data version into something less ambiguous. If it is renamed, I'll also need to change news and backup file.

Specific changes to make to the data file

• update Patient List when room nr is rolled out.
• change L_Settings stuff to go to _info instead to be able to get rid of legacy content
• update or separate article when the BP helper using L_BPs is implemented
• add s_FinalChecks, fix all spots that use FinalCheck (or make a function to check it)
• update when done: CCMDB.mdb; "News and backup" batch file; Settings.mdb
• change error code to populate L_Errors

Changes Under Discussion

These requested changes can not be processed until the questions about them have been answered. Wherever possible the discussion should conclude in the home article of the topic. If a topic has no home article, it can be listed here.

Help to manage Serial numbers

If I understand right data collectors currently collect one or more locations from one or more pools of serial number pools. Details of the serial number system are located in Serial number. This has been flagged as one reason why people find paper printouts of the most recent patients sent useful.

Serial numbers have to be unique for a collection location, so if a ward like HSC_H4H is collected on by multiple collectors then a serial number must never be re-used. One way this is achieved is to assign blocks of numbers for each 100 possible serial numbers to a given collector.

A study number must be entered before any other data since access uses it to set the relationships for the data. To accomplish that, function new_pat_id() opens a window requesting the serial number pops up when "add new patient" is clicked on the patient list.

Added question to Serial_number#Special-Use_Serial_Numbers to find out if any serial numbers need to be "reserved".

Possible Solution

The program would provide a default serial number. For locations with only one serial pool, this would be a one-step, transparent process. Collectors with more than one serial number pools would get a dropdown list of possible pools; upon choosing one the program would default to serial (most recent+1). Collectors using only blocks of serial numbers would have to keep track if they are "leaving" their block (e.g. if you are using only 20-39 and the program defaults to 1140, you would have to realize that and change it to 1220).

To accomplish this, the _info table would store

• variable: serial_pools, value: comma-separated list of wards
• for each ward, a variable "Serial_pool" & <ward> (e.g. Serial_pool_HSC_H4H) with the most recent value used, updated automatically by the serial wizard

• Any thoughts about this? Do you think it would/wouldn't work or be helpful? Especially, do you think collectors would consistently "catch" the ends of blocks?

Priority numbering for labs

• is there still a need to prioritize lab tests?--TOstryzniuk 15:12, 26 October 2010 (CDT)
  • Yes, there is while there are more possible labs to enter than there are spots to send from. Could get rid of the priorities entered by collectors if we gave priorities to the labs on the list themselves, e.g. CBC is always more important to count than Chest Xray. Would be able to get rid of the priorities also once we go to the reduced labs list. Also, would be able to get rid of priorities for labs when we go to the new data repository where there are no restrictions on the number of labs collected. All this is true for drugs as well. Ttenbergen 15:21, 26 October 2010 (CDT)
    • Currently that is the issue but in the new lab list there are only 24 items, this is the total number of spaces allowed to send from. Have a look at the output files....24 labs spots available. Once we completely transition to the new lab list then assigning a priority is not needed. The plan is to start the New lab list January 1.11 The current options to migrate to new list is:

1. All new admission for Jan 1.11 would use new lab list and those admitted prior to Jan 1.11 would use old lab list until discharged,though this can drag on for long any stays.
2. new lab list would be applied to all discharges starting Jan 1.11. The long stays already have other labs collected and the efforts would be turfed. I have patient from March, April and June still in IICU and may still be there January.

• waiting for Julie's input on this.--TOstryzniuk 19:08, 27 October 2010 (CDT)
  • Option 1 is better since it is hard to predict the type of request in the future. In the past when there is a request, I generated the LAB Report from the TMSX which is based on admit date as the covered period. So for 2010, it will be nice to have a complete 2010 data when doing that report. JMojica 09:48, 28 October 2010 (CDT)

Keeping track of what has been reviewed

One of the reasons some people keep paper is that they track on paper when they have last looked at labs etc so they know where to count from next time they review the chart. This is not a problem at SBGH where data can be collected from the EPR after discharge since the chart is always available. Is there a way to facilitate this in CCMDB.mdb that would add the least manual entry overhead for the greatest facilitation and paper reduction?

Option 1 - "last collected" dates

I have had a suggestion that adding date/time fields to the program that would set to "now" upon double-clicking them would mean one more thing tracked in the .mdb rather than on paper. These fields would not be mandatory and they would not be sent in, they would just be for your own use. Suggested date/times were:

• LAST COUNTED LABS
• LAST COUNTED EKG
• LAST COUNTED CXR
• LAST COUNTED PHARM
• CHART REVIEWED

This seems the simplest option to add to CCMDB.mdb, but it may not be much use to those who "tick" or flowsheet at this time.

• thoughts?
  • I know what the collectors are thinking here but would this not be labor intensive inputting dates for each and every single lab item so you can remember where you left off? My thoughts are this, count lab retrospectively. I believe that is what most are doing anyway. Example if you are in on Monday Oct 25 you would count all labs up to midnight Oct 24. The next time you are in, say Wednesday Oct 27th, you would started counting labs from Oct 25 0001 hrs to Oct 26 midnight. You would only have to put one date to remind you when you last counted labs for a patient. Maybe that is what you have in mind and I have misunderstood. --TOstryzniuk 19:42, 27 October 2010 (CDT)
  • thoughts and further suggestion from Collectors? We are trying to make the data collection tool work for you.--TOstryzniuk 19:42, 27 October 2010 (CDT)

Option 2 - electronic flowsheet

We could make sheets in ccmdb.mdb that have columns for dates and lines for e.g. the drug or test in question.

• Any opinions? This is the most comprehensive, but I worry that this will cause a lot of overhead entering.

Option 3 - ticks in charts

Some locations (e.g. MWaschuk@Grace) put ticks in the charts up to how far they have collected. Other locations tell me that sometimes whole pages of records are missing. Would going through the chart chronologically and relying on ticks miss such data loss? Also, where "ticks" are used, what are people using for ticks? If we are going to have coverage across sites this should at least be consistent. Anyone else using purple yet? We could use purple ticks? Or some quick-drawn symbol, maybe. A tick seems like something that could arrive by accident...

• thoughts? Ttenbergen 15:37, 25 October 2010 (CDT)

Trish (or anyone): Can you see any concerns if this were implemented? Would we automate something that could be done easier, or maybe doesn't need to be done at all? Any thoughts along that line? Ttenbergen 15:37, 22 October 2010 (CDT)

Facilitate exchange of patient discharge times

We discussed at the last collection meeting that I would set up a program to provide the collectors with admit/discharge info. We discussed whether to include Comos etc. in this, and it was decided not to - Julie will from now on treat ANY como on this admisssion as a como for all subsequent ward stays on that admission. (see also Admit, Transfer and Discharge date and time) Template:Discussion

• Trish - did I get that right, shall we proceed? Ttenbergen 14:59, 20 October 2010 (CDT)
  • I thought you, Tina, was going to set up something so we could at least track discharge times and not have to email coworkers esp in other sites. DKlopick
    • That was what I was trying to confirm. I re-worded it to be clearer. I need to know if we did actually decide that you would no longer compare comorbids. Trish? Ttenbergen 10:02, 25 October 2010 (CDT)

Limited Labs List

article that deals with this is S AllLabsTable Ver2.

Make a better labs entry system

Once reduction of number of labs collected is official and down to as many as can be, I will make it so the labs are entered in a pre-populated list. Ttenbergen 15:16, 2 July 2010 (CDT)

• Can you make it so the lab list does not mix up the priority numbers. Makes it hard to add new ones without using same number twice.LBilesky 08:23, 22 October 2010 (CDT)
  • I am not sure what you mean. How is it mixing up any priorities? Ttenbergen 13:19, 22 October 2010 (CDT)
    • I have run into the same problem. Seems once the file is closed and then reopened the lab list has sorted the labs alphabetically (not numerically by priority number)Mlaporte 15:16, 22 October 2010 (CDT)

Tasks

For a long time now there have been requests to make the Tasks more of a drop-down, similar to the ADLs, or a checklist like the greensheets used to be. This is a mess to implement, since it doesn't correspond to the data structure this is stored in. An alternative might be to automatically enter the five tasks as "blanks" for every new medicine patient. I would make 5 new tasks entries that would sort before each of the 5 existing options. This would be similar to a checklist in that the records are already there, and only would need to be edited. Template:Discussion I can not think of any data or process drawbacks of doing this... can anyone else? If I hear no issues by the time I get to this round of changes I will implement this.Ttenbergen 11:09, 20 October 2010 (CDT)

• Do you mean they will automatically pop up as negatives ie:no trach, no hemodialysis etc. and if the pt is on anything then we would change the task to reflect this? If so I think this would help.--CMarks 14:50, 21 October 2010 (CDT)
  • That is what I mean. Trish/Julie, do you think there would be a problem with this defaulting? Ttenbergen 13:20, 22 October 2010 (CDT)

Getting rid of old buttons

Unless I hear that there is still use for this button, I will remove both the "Edit Patients" button and the filter button from CCMDB.mdb in the next round of changes. Speak now or forever hold your peace if you think there would be a problem with this. Note: Sending_Patients#Historical:Filtering_for_complete_patients needs to be fixed after this change. Template:Discussion

• We use these buttons before sending files here at HSC otherwise we are unable to edit patients and filter completes.--CMarks 08:08, 18 October 2010 (CDT)
  • We use them at STB also it is handy to filter the completes especially when you are sending many profiles. --LKolesar 11:55, 18 October 2010 (CDT)
    • I use the FILTER button as a back up to know how many files are being sent prior to sending --PStein 08:31, 25 October 2010 (CDT)
      • I have added a final-check sort button to the patient list to v1.995. If you sort by final-check and then click on the last record with final-check checked, you will get the number (i.e. count) of that record at the bottom of the screen. So, any other reasons we would still need the button? Ttenbergen 09:56, 25 October 2010 (CDT)

When we had the PDAs, we needed to check the complete patients only at sending time since the PDAs could not do it as you go along. Now that we are collecting in Access all along, we no longer need to wait and batch this check for send day. Is there any reason to not set a patient to "final checked" when you are finished collecting them? When final-check is clicked a bunch of tests run, so doing this when you still have the chart in front of you would save a trip back later. Checking patient over at send time seems like an extra step that doesn't add any value any longer. Should we continue to do this? Template:Discussion Trish, this probably requires input from you. Ttenbergen 11:09, 20 October 2010 (CDT)

Facilitating diagnosis entries

I have had various requests to make it faster or easier to enter diagnoses. A lot of those requests would be addressed with more training on the various ways diagnoses can be entered. I have added a fair bit of functionality, but there may not have been enough training on this yet. Trish, before I add much functionality here, could you put me on the next meeting agenda for 15 minutes to show/tell/discuss diagnoses entries. Template:Discussion

I discussed with Fran and Con that, if there were groups of diagnoses that often go together, it might be possible to combine them somehow. I could add a limited number of buttons to e.g. put in "cardiovascular- Hypertension, MI, hypercholesterol, etc." as requested by Con. Could collectors provide a list of dx groups for which this would be useful?

• (button name?): cardiovascular- Hypertension, MI, hypercholesterol,
• One idea might be to eliminate the use of "other categories" in the diagnosis sections. We don't really need this categorization any more. For example, instead of having to go into "other cardiovascular procedures" and find the item, one would just type in the item directly like permanent pacemaker so that tab and type would go faster. The "other" categories would need to be reviewed and revamped. The numerical system would be unchanged. This would make both entering comorbids and diagnostic codes immensely more efficient. I don't mind helping with this if you decide it could work.
• Another thing that I believe could be eliminated is acute care map in variable 2. Betty Lou had mentioned in a meeting a few years ago that this was no longer necessary. All patient who have acute MI are on the care map. This is just an unnecessary piece of data to add. --LKolesar 14:52, 27 October 2010 (CDT)--

Extra "pathogens"

--MWaschuk 17:46, 9 June 2010 (CDT) As part of an effort to reduce relying on paper notes during collection, Marie brought up the possibility to put "waiting for tests" and "review" as options into the pathogen dropdown and to conditional-format the field in orange if they are chosen. Of course, choosing them would also disable checking DX complete.

• Would anyone else find this useful? Should it be included in the Master DB? Ttenbergen 17:34, 9 February 2010 (CST)
  • Yes, I would find this useful
    • I usually put unknown pathogen in my pathogen slot if waiting for results, then when rechecking the profile upon discharge, I re-check on the EPR for results again. If still no results, I generally do not hold the profile longer. I wonder what our policy should be on this. How long should we hold a profile for possible results? This could back up our flow of profiles quite a bit. What does everyone think?--LKolesar 15:08, 10 June 2010 (CDT)

Changes to the ccmdb.mdb to reduce reliance on paper

The "extra pathogens" entry above makes me wonder: are there other features collectors can think of that would suddenly make it more desirable to enter data electronically in the first place rather than on paper? This question really only applies to the access program (e.g. to laptop users) since we are phasing out the PDAs and I am not going to make any modifications to them because they cause cascading problems.

• Any thoughts? Ttenbergen 17:34, 9 February 2010 (CST)

Bugs that need replicating

In order to troubleshoot a bug we need to know a way to "make it happen". This section contains bugs that have been reported but can not be demonstrated on demand.

WBC = 0 does not uncheck apache box

As per email from Pagasa, and confirmed by Tina, 28 jan 2010 Ttenbergen 14:32, 28 January 2010 (CST)

There may also still be a problem with sending very low numbers. Lois apparently thought she entered and sent WBC with a value of 0.02, which CCMDB.mdb should accept, but may not send properly. Ttenbergen 12:58, 29 January 2010 (CST)

How to demonstrate this bug

• Anyone?Ttenbergen 12:25, 12 April 2010 (CDT)
  • HSC-MICU-Batch 22 sent June 2.10. See Pagasa Email sent June 3.10. Will double check if it is the uploader program. If value 0.02.
  • I checked Lois original csv and WBC value entered was .02 but when I appended it tells me that incomplete apache score but after I appended it showing blank on Ed's program.PTorres[User:PTorres|PTorres]] --TOstryzniuk 17:27, 8 June 2010 (CDT)
    • Trish, have you been able to find out if this goes wrong in the uploader? Ttenbergen 16:19, 2 July 2010 (CDT)
      • Yes, PDA uploader for ICU is not reading in a WBC value of 0.02 for WBC. Problem shipped over Programmer today.--TOstryzniuk 14:52, 22 October 2010 (CDT)

Infection without pathogen subcode was sent to csv

Oaks - Patient coded for Cystitis Bladder Infection but no sub code? Admit date Jan 25/10-Jan. 30/10. Pagasa Torres.

Patient coded for Cystitis Bladder Infection but no sub code; Admit date Jan 25/10-Jan. 30/10.

Asked Wendy if there was any extraordinary edits for this pt, and she replied "Nope and I too am surprised as it has always been caught if I miss it in the Edit Patients pane??? I just simply forgot as I was awaiting the results...have no idea how it got sent???"

How to demonstrate this bug

• Anyone? Ttenbergen 16:59, 9 February 2010 (CST)

Tasks not being sent

Occasionally Medicine TASK items don't get sent to the task.mdb. Need a way to flag this. Ttenbergen 15:53, 2 July 2010 (CDT)

• Template:DiscussionHow frequent of a problem is this, has it recurred since this was documented? Ttenbergen 11:09, 20 October 2010 (CDT)
  • there has been no further reported problems from Pagasa or Julie since this was documented. --TOstryzniuk 15:24, 20 October 2010 (CDT)

How to demonstrate this bug

• Anyone? Ttenbergen 12:16, 22 October 2010 (CDT)

Previous requests that were not addressed (with reasons)

Default location field to main collection location

There have been requests to default the location field to the main (or only) collection location. I could set this up, but defaulting these would increase the chances of the data just being left in there if a patient is an exception. Upon consultation with Trish it was decided not to risk an increase in errors to make this more convenient.

• "There are a number of site collecting on 2 or more units using the same laptop. We get ward location entry error when there is a default location set." ... "--TOstryzniuk 12:25, 20 October 2010 (CDT)"

Cancel for Settings

change the settings form to have a save and a cancel button...

No can do. There are is a sb-form in settings form that would not be cancelled, so having a cancel would be misleading.

Off-line copy of wiki for laptop

We found a way to do this, but it would add overhead work to the collectors' job and it would deter from using the wiki as a tool owned and updated by all, so Trish and Tina decided not to pursue this further. Ttenbergen 15:01, 16 April 2010 (CDT)

• look into possibility of automatically downloading WIKI guide to laptop each day that it is synch'd. (portable version of guideline)--TOstryzniuk 20:11, 4 April 2009 (CDT)

• As moved from Diagnosis Coding:

A suggestion has been put forth to Tina is to find a away to allow a downloaded copy of the WIKI guide on laptops, the collection tool we are moving to soon. I think this would be an excellent solution if she can do this. Collectors would then be able to search for info no matter where they are in the collection field and each time they synch there laptop, they would get updated WIKI items. That is the plan............but waiting to here from IT support.

It is challenge right now in that the guideline information is not easily and quickly accessible while you are doing your work in the unit or in medical records. Most collectors must run back to their office or find a free computer to log in and look up guide info.

Keep in mind, guide information changes regularily, so what you print out one day is outdated the next for a number of articles. TOstryzniuk 11:25, 29 April 2009 (CDT)

• April 2010 - Trish and Tina reviewed the apparent option to have a COPY of wiki on laptop. Problem if implemented:

1. if collectors will not have the most update the copy on laptop because changes and updates are happening daily on the MAIN copy here, and thus changes to instruction will get missed.
2. 2. Don't want multiple copies of instruction and information on multiple laptops that are not up to date.
3. Collector questions and contribution to the Wiki cannot be added to copy of Wiki on laptop. Questions and contribution must be made centrally not locally so that information is in ONE PLACE and the same for everyone.--TOstryzniuk 19:11, 20 April 2010 (CDT)

APACHE lab value limit change request

• Apache upper limit for creatinine is 2968, which is based on the highest APACHE that was ever in database. STB has one pt with a creat of 3000 and has requested to change upper limit. Laura K. --TOstryzniuk 12:12, 6 July 2010 (CDT)
  • Decided to leave limits be and have collectors enter limit values and email Pagasa with real values when extreme values are encountered. Contacted Pagasa to let us know if such emails become too frequent. Ttenbergen 08:53, 12 July 2010 (CDT)