Thanks, Alex. Which site is it that didn't include any data elements except the selection criteria?
To the sites that were still working on their submission as of HackathonThree<https://informatics.gpcnetwork.org/trac/Project/wiki/HackathonThree>, please let us know what this script says about your data. Justin, UMN, is the output of this script happy when given your file? If not, what issues do you see? George, MCW, how about you? Tom, WISC: you gave an ETA of 2-3 weeks. It's been 2 weeks; any luck? Phillip, UTSW - you reported some complications around your Children's Hospital, but that seems orthogonal to this sort of validation. What happens when you run it on data that you have readily available? Alex, I'm not sure I have a good list of which sites are pending. The new "All files OK?" column on the status report<https://redcap.gpcnetwork.org/redcap_v6.2.2/DataExport/index.php?pid=29&report_id=51> is only filled in for UTHSCSA. Back in November we started tracking status in #379<https://informatics.gpcnetwork.org/trac/Project/ticket/379>, but the last update there seems to be Jan 7, before HackathonThree. -- Dan ________________________________ From: [email protected] [[email protected]] on behalf of Alex Bokov [[email protected]] Sent: Monday, February 01, 2016 6:41 PM To: [email protected] Subject: DataBuilder validation script Just remembered, as of Hackathon3 I owe everyone the "validation script" I wrote for DataBuilder (script attached). If you pass it to sqlite3 on the command line, it will spit out a bunch of tables summarizing what's in your .db file. This is not dependent on what database you run your i2b2 on, it uses only SQLite. Here is how to invoke it on a Linux command line: sqlite3 -column -header YOURDATABUILDERFILE.db < dbuilder_validate.sql > YOURSITENAME.txt Probably same syntax on MacOS shell. Hopefully some Windows user will be kind enough to furnish the equivalent for cmd. Nothing sophisticated, but it can tell you if you are missing large numbers of EHR data elements. Hope it helps. Here is output from a good .db file (abridged for readability): tbl N N Distinct ------------------ ---------- ---------- MODIFIER_DIMENSION 0 0 CONCEPT_DIMENSION 5419 5376 PATIENT_DIMENSION 6561 6561 OBSERVATION_FACT 420970 VARIABLE 52 52 ---------- ------------------------------- ---------- Patients that have observations 6561 Observations that have patients 420970 ---------- Data elements redundant with selection criteria ----------------------------------------------- 9 --------------------------- Non redundant data elements concept_path name_char ----------------------------------------------------------------------------------------------------------------------------- \i2b2\labs loinc\LP29693-6\LP29697-7\LP71800-4\54037-7\54050-0\54053-4\24323-8\ 24323-8: Comprehensive metabolic 2000 ... \i2b2\labs loinc\LP29693-6\LP29697-7\LP7834-7\24348-5\ 24348-5: Free t4 and tsh panel - serum ... \i2b2\labs loinc\LP29693-6\LP29697-7\LP71800-4\54051-8\24362-6\ 24362-6: Renal function 2000 panel - ... ... many additional lines omitted ... TOTAL ELEMENTS 41 -------------------------------------------------------------- Non redundant data elements actually found in OBSERVATION_FACT concept_path name_char ----------------------------------------------------------------------------------------------------------------------------- \i2b2\labs loinc\LP29693-6\LP29697-7\LP71800-4\54037-7\54050-0\54053-4\24323-8\ 24323-8: Comprehensive metabolic 2000 ... \i2b2\labs loinc\LP29693-6\LP29697-7\LP7834-7\24348-5\ 24348-5: Free t4 and tsh panel - serum ... \i2b2\labs loinc\LP29693-6\LP29697-7\LP71800-4\54051-8\24362-6\ 24362-6: Renal function 2000 panel - ... ... many additional lines omitted ... TOTAL ELEMENTS 31 Here is output from a site that didn't include any data elements except the ones used to select the patient set in the first place (that's what I mean by non-redundant in the report above-- non redundant with what the researchers already know because it was part of the selection criteria for the survey... the point of the i2b2 pull is to get the labs, diagnoses, procedures, and meds specified at the bottom of GPC Obesity Query v1.3.pdf<https://informatics.gpcnetwork.org/trac/Project/raw-attachment/ticket/254/GPC%20Obesity%20Query%20v1.3.pdf> or in obesity_data_elements_keys_names_codes.xls<https://informatics.gpcnetwork.org/trac/Project/raw-attachment/ticket/254/obesity_data_elements_keys_names_codes.xls>): tbl N N Distinct ------------------ ---------- ---------- MODIFIER_DIMENSION 0 0 CONCEPT_DIMENSION 36 35 PATIENT_DIMENSION 17562 17562 OBSERVATION_FACT 2509960 VARIABLE 29 27 ---------- ------------------------------- ---------- Patients that have observations 17562 Observations that have patients 2509960 ---------- Data elements redundant with selection criteria ----------------------------------------------- 25 --------------------------- Non redundant data elements concept_path name_char -------------- ---------- TOTAL ELEMENTS 0 -------------------------------------------------------------- Non redundant data elements actually found in OBSERVATION_FACT concept_path name_char -------------- ---------- TOTAL ELEMENTS 0 Thanks. -- Alex -- Alex F. Bokov, Ph.D. Deputy Chief, Clinical Informatics Research Division UT Health Science Center at San Antonio
_______________________________________________ Gpc-dev mailing list [email protected] http://listserv.kumc.edu/mailman/listinfo/gpc-dev
