On September 27, 2016 at 1:13:02 PM, Block, Jason Perry,M.D. ([email protected]<mailto:[email protected]>) wrote:
Hi Russ. Thanks for emailing - no bother at all. I'm cc'ing a few others who have expertise in this issue (Charles Bailey at CHOP and Casie Horgan and Jessica Sturtevant on the DRNOC QF team). We are not including inpatient medications in our study. We made this decision for logistical reasons (we didn't think it would be available, or if available, it would be available in a highly variable way) and scientific (we are most focused on potentially modifiable antibiotic use, and inpatient use doesn't fit that categorization). The scientific point is a debatable one, and we have had some subtle disagreement on the team about this. But, ultimately, since it wasn't going to be available, we chose to pursue use of antibiotics prescribed in the outpatient setting. Casie and Jessica might be able to respond to how we excluded meds in ED and Inpatient settings, if they did that explicitly. Charlie has lots of experience with meds as well, so he might want to respond to this. We’ve adopted a similar position with regard to inpatient medication administration, and just omit those records when we build our PCORnet database. The biggest issue here, I think, is granularity: does one represent inpatient meds a dose at a time, to get the best estimate of actual exposure (this is what we do on the PEDSnet side), or roll up some span of doses into an order-level or era-level summary, which keeps the size of the table down, but may overestimate exposure (especially for prn meds)? We do feed discharge prescriptions forward into the PCORnet prescribing table, and will also capture some patient-reported medications presumably ascertained as part of medication reconciliation, since these may look like prescription records in some EHRs. I also think these are valuable information about patient exposures, though they are often lacking in specifics about frequency and the like. I could see how studies would like to include ED and inpatient med prescribing (or dispensing), and I would support inclusion of that in the CDM. We just won't use for our study, as it's currently structured. jb From: Russ Waitman [[email protected]] Sent: Tuesday, September 27, 2016 12:50 PM To: Block, Jason Perry,M.D. Cc: [email protected]; [email protected]; Christopher Forrest ([email protected]) Subject: Confirming medications sources for Antibiotic obesity study question Dear Jason, I hate to bug you but we wanted to get your quick check on something as the GPC sites are doing their CDM work in support of your study. Apologies for not being as close to your study but wanted to confirm something about how the CDM is being loaded and potential implications for your study as it’s the first big thing using medications from the CDM. The CDM http://pcornet.org/wp-content/uploads/2014/07/2015-07-29-PCORnet-Common-Data-Model-v3dot0-RELEASE.pdf on page 13 specifies that the prescribing tables and the dispensing tables should not contain - inpatient medication administrations - active medication lists resulting from medication reconciliation Prescribing should obviously have outpatient prescriptions but it’s unclear to some regarding physician inpatient orders. The table specification in written largely in language of outpatient prescriptions but the page 13 figure indicates that prescribing encompasses physician orders that spawn both dispenses and administrations. I would just be concerned that if sites are conservative in only loading ambulatory meds, we’re going to have limited CDM data to support studies that want to include inpatient and emergency room medication exposures. As you look at your early sites, what are you seeing? Does that match what exposures you wish you had? Bottom line: are you only looking at outpatient prescribing or dispensing exposure to antibiotics or does your study need inpatient and emergency room medication exposure? My personal bias is to exclude the admins and the med rec at this point but include both inpatient and outpatient medication orders written by the physician. Russ Waitman, PhD Director of Medical Informatics Associate Vice Chancellor for Enterprise Analytics Professor, Department of Internal Medicine University of Kansas Medical Center, Kansas City, Kansas 913-945-7087 (office) [email protected]<mailto:[email protected]> http://www.kumc.edu/ea-mi/ http://informatics.kumc.edu<http://informatics.kumc.edu/> http://informatics.gpcnetwork.org<http://informatics.gpcnetwork.org/> – a PCORnet collaborative The information in this e-mail is intended only for the person to whom it is addressed. If you believe this e-mail was sent to you in error and the e-mail contains patient information, please contact the Partners Compliance HelpLine at http://www.partners.org/complianceline . If the e-mail was sent to you in error but does not contain patient information, please contact the sender and properly dispose of the e-mail. — Regards, Charlie
_______________________________________________ Gpc-dev mailing list [email protected] http://listserv.kumc.edu/mailman/listinfo/gpc-dev
