RE: Lab Normal Value Processing
Hi Phillip, Here’s an example of the code we use at Marshfield to extract numbers from messy lab values. I’m not sure if it would work with ranges in the ‘1.0-10.0’ format, but this could be a start? REGEXP_EXTRACT ( TRIM ( REGEXP_REPLACE ( TRIM(UPPER(LOW)), '(^[>])|(^[> ])|(^[<])|(^[< ])|(^[=])|(^[= ])|(^[*])|(^[* ])|([%]$)', '' ) ) ,'^[-+]?(([0-9]+[.]?[0-9]*)|([.][0-9]+))$' ) AS NORM_RANGE_LOW FYI, the functions above are user-defined functions native to our IBM PureData/Netezza data warehouse. Steff From: Gpc-dev On Behalf Of Phillip Reeder Sent: Tuesday, September 24, 2019 9:37 AM To: gpc-dev@listserv.kumc.edu Subject: [EXTERNAL] Lab Normal Value Processing We haven’t used the normal values associated with individual lab results in the past and while looking at the normal values, many of them are text with modifiers like <,> <=. Etc., along with some that are ranges like 1.0-10.0. Does anyone have SQL/regular expressions that will filter all of those out into the columns required for the PCORICDM? Thanks, Phillip 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. ___ Gpc-dev mailing list Gpc-dev@listserv.kumc.edu http://listserv.kumc.edu/mailman/listinfo/gpc-dev
Re: Lab Normal Value Processing
In our CDM, we don't seem to use normal ranges from individual results (yet) either; we us an age/sex adjusted table: https://github.com/kumc-bmi/i2p-transform/blob/db7cf8809d810385d2e6d4e4f1e66ac7cb02ed57/curated_data/labnormal.csv https://github.com/kumc-bmi/i2p-transform/blob/master/Oracle/lab_result_cm.sql In i2b2, we do mark Hi / Lo results on an individual basis, but it looks like we filter out things other than plain numbers. https://github.com/kumc-bmi/heron/blob/master/heron_load/epic_labs_transform.sql (private repository; see MultiSiteDev<https://informatics.gpcnetwork.org/trac/Project/wiki/MultiSiteDev> regarding access) FWIW, there is some regular expression stuff in there to handle result values such as ">40". It also uses https://github.com/kumc-bmi/heron/blob/master/heron_load/curated_data/lab_regex.csv From: Gpc-dev on behalf of Phillip Reeder Sent: Tuesday, September 24, 2019 9:36 AM To: gpc-dev@listserv.kumc.edu Subject: Lab Normal Value Processing We haven’t used the normal values associated with individual lab results in the past and while looking at the normal values, many of them are text with modifiers like <,> <=. Etc., along with some that are ranges like 1.0-10.0. Does anyone have SQL/regular expressions that will filter all of those out into the columns required for the PCORICDM? Thanks, Phillip UT Southwestern Medical Center The future of medicine, today. ___ Gpc-dev mailing list Gpc-dev@listserv.kumc.edu http://listserv.kumc.edu/mailman/listinfo/gpc-dev
Lab Normal Value Processing
We haven’t used the normal values associated with individual lab results in the past and while looking at the normal values, many of them are text with modifiers like <,> <=. Etc., along with some that are ranges like 1.0-10.0. Does anyone have SQL/regular expressions that will filter all of those out into the columns required for the PCORICDM? Thanks, Phillip UT Southwestern Medical Center The future of medicine, today. ___ Gpc-dev mailing list Gpc-dev@listserv.kumc.edu http://listserv.kumc.edu/mailman/listinfo/gpc-dev
