Hi Lindsay, I hope I am clear on what you are getting at here.
If we knew the complete list of research questions, then we wouldn't be building an i2b2 tool of this scale, but rather just extract 3 datasets and be done with it. If we are to answer questions that we do not yet anticipate, we need to include these types of data so that future extracts can more likely answer questions without re-extracting these (very large) datasets. As for research showing the utility of diagnoses applied under specific conditions (such as primary professional diagnoses or principal codes applied at hospital discharge), we have used them regularly here at Marshfield with good success. Aaron Aaron Miller, PhD | Analytics Informatician, Interim Research Analytics Manager | MCRF - BIRC | (715) 221-6429 From: Greater Plains Collaborative Software Development [mailto:[email protected]] On Behalf Of Lindsay Cowell Sent: Wednesday, February 26, 2014 5:02 PM To: [email protected] Subject: Re: modifier proposal from hackathon Furthermore, On Feb 26, 2014, at 3:53 PM, Dan Connolly wrote: In my experience, the YouArentGonnaNeedIt<http://c2.com/cgi/wiki?YouArentGonnaNeedIt> works well: Always implement things when you actually need them, never when you just foresee that you need them. I think this makes sense in the context of software development. Why build functionality you don't need, and may not need. On the other hand, the purpose (as I understand it) of our data repositories is to be able to answer unanticipated questions. I would think that means taking more care to accurately represent the data included, rather than just representing the data in a form that answers our current test cases. Thanks Lindsay The wiki page goes on to explain how adding them at a later date will not delay acquisition of high-quality datasets. We have to characterize ALS, breast cancer, and obesity cohorts. As to anything beyond that, I'm not inclined to track design issues until requirements that motivate them arise. >From the ALS break-out of the >hackathon<http://informatics.gpcnetwork.org/trac/Project/wiki/HackathonOne> in >particular, it seems we don't need modifiers for ALS diagnosis: * modifiers: billing? clinical? doesn't matter. * is past medical history too much? * no * is past medical history required? * no -- Dan ________________________________ From: Miller, Aaron W [[email protected]] Sent: Wednesday, February 26, 2014 2:31 PM To: Dan Connolly; [email protected]<mailto:[email protected]> Subject: RE: modifier proposal from hackathon Hi Dan, I am not aware that there are any research questions that have been stated explicitly. The emphasis of the data group has been to try to be prepared to answer a broad range of questions, and that is why these modifiers were identified. If you can point to representative (and broad) list of research questions that need to be answered, then we can trim this list down. Otherwise, I would suggest that we pursue intelligent ways to incorporate these types of data, as adding them at a later date will delay acquisition of high-quality datasets. Aaron Aaron Miller, PhD | Analytics Informatician, Interim Research Analytics Manager | MCRF - BIRC | (715) 221-6429 From: Dan Connolly [mailto:[email protected]] Sent: Wednesday, February 26, 2014 2:20 PM To: Miller, Aaron W; [email protected]<mailto:[email protected]> Subject: RE: modifier proposal from hackathon So we're not yet aware of any such research questions? Then email discussion is fine... it's interesting stuff... but this design discussion isn't on our critical path, so I don't need to track it with a ticket or put it on group teleconference agendas. -- Dan ________________________________ From: Greater Plains Collaborative Software Development [[email protected]] on behalf of Miller, Aaron W [[email protected]] Sent: Wednesday, February 26, 2014 2:11 PM To: [email protected]<mailto:[email protected]> Subject: Re: modifier proposal from hackathon Hi Dan, This list is intended to continue our discussion of diagnosis modifiers. When specific sources or types of diagnoses are necessary to answer research questions, these will greatly enhance our ability to respond with precision. Aaron Aaron Miller, PhD | Analytics Informatician, Interim Research Analytics Manager | MCRF - BIRC | (715) 221-6429 From: Greater Plains Collaborative Software Development [mailto:[email protected]] On Behalf Of Dan Connolly Sent: Wednesday, February 26, 2014 2:05 PM To: [email protected]<mailto:[email protected]> Subject: Re: modifier proposal from hackathon I guess I missed the motivation for these modifiers. Which research questions benefit from these distinctions? -- Dan ________________________________ From: Greater Plains Collaborative Software Development [[email protected]] on behalf of Wilson Nathan [[email protected]] Sent: Wednesday, February 26, 2014 1:21 PM To: [email protected]<mailto:[email protected]> Subject: Re: modifier proposal from hackathon All, I talked over the billing section of the outline with one of our business planning folks who works with hospital account data and based on that conversation I would recommend the following adjustments to the billing diagnosis modifiers. Original: BILLING DX DISCHARGE DX PRINCIPAL DISCHARGE DX NONPRINCIPAL PROFESSIONAL DX PRIMARY PROFESSIONAL DX NONPRIMARY Suggested: BILLING DIAGNOSIS ADMIT DIAGNOSIS PRINCIPAL ADMIT DIAGNOSIS SECONDARY ADMIT DIAGNOSIS PRESENT ON ADMISSION DISCHARGE DIAGNOSIS PRINCIPAL DISCHARGE DIAGNOSIS SECONDARY Reasons: The main set of diagnosis found on the bill are discharge diagnosis (at least from a CLARITY perspective) Non-principal diagnosis are collectively referred to as secondary For admitting diagnosis it is important to know if it is/was present on admission Does the group have any thoughts and/or other suggestions? Nathan Wilson From: Greater Plains Collaborative Software Development [mailto:[email protected]] On Behalf Of Dan Connolly Sent: Tuesday, February 25, 2014 11:55 AM To: [email protected]<mailto:[email protected]> Subject: modifier proposal from hackathon This was on the presentation machine at the hackathon. I gather it's a proposed modifier hierarchy. -- Dan ________________________________ The contents of this message may contain private, protected and/or privileged information. If you received this message in error, you should destroy the e-mail message and any attachments or copies, and you are prohibited from retaining, distributing, disclosing or using any information contained within. Please contact the sender and advise of the erroneous delivery by return e-mail or telephone. Thank you for your cooperation. ________________________________ The contents of this message may contain private, protected and/or privileged information. If you received this message in error, you should destroy the e-mail message and any attachments or copies, and you are prohibited from retaining, distributing, disclosing or using any information contained within. Please contact the sender and advise of the erroneous delivery by return e-mail or telephone. Thank you for your cooperation. ------------------------------------------------------------------- Lindsay G. Cowell, PhD Division of Biomedical Informatics Department of Clinical Sciences University of Texas Southwestern Medical Center at Dallas 5323 Harry Hines Blvd. Dallas, TX 75390-9066 F4.212b, MC9066 phone: 214-648-2289 fax: 214-648-2064 [email protected]<mailto:[email protected]> ________________________________ UT Southwestern Medical Center The future of medicine, today. ______________________________________________________________________ The contents of this message may contain private, protected and/or privileged information. If you received this message in error, you should destroy the e-mail message and any attachments or copies, and you are prohibited from retaining, distributing, disclosing or using any information contained within. Please contact the sender and advise of the erroneous delivery by return e-mail or telephone. Thank you for your cooperation.
