Page tree
Skip to end of metadata
Go to start of metadata

Update 10AM Pacific, Friday

  • We're putting on FHIR tutorials at UW today and a FHIR workshop tomorrow, so we won't have time to do a separate debrief
  • We did just present a summary in the Public Health Track Summary - which is recorded
  • Please feel free to contact lober@uw.edu , ivanc@uw.edu , jflow2@uw.edu for more info
  • Thanks - we had fun, and got good stuff done - thanks to partners and advisors...

Update 12:30AM Pacific time, Friday

  • We tend to go to bed a bit late, and work a bit late.
  • We'll see you 7:30 Pacific time, later this morning.  chat link below.

Update 7:30 AM Pacific time, Thursday

Update 10:15 PM Pacific time, Thursday

A few things:

Where we are in all this:

  • We hope to be testing with Epic tomorrow (we've done that in IHE connectathons in the past with similar apps)
  • We hope to show enough of what we're doing to gather interest and comments regardless of testing success
  • We hope to be successful in testing.
  • We want to meet folks and have fun...!
  • Our excuse, we've been pretty distracted this past couple of weeks, and that distraction resulted in approval by WA DOH security review for a production connection to the state's PMP (yes, we drop the D in Washington), through the state's HIE.  That, for us, was fantastic news and got a bit in the way of a fancier testing agenda.

Thanks,

Bill


Original submission

Opioid prescribing guidelines and state rules require a variety of individual level information, including their prescribing or dispensing history, clinical observations from an EMR such as lab test and clinical conditions,  and data which may or may not be part of their clinical record, such as non-fatal overdoses.  

COSRI extends the work of the AHRQ Clinical Pain Summary by combining that app’s FHIR-EHR and Clinical Quality Language (CQL) guidelines logic (using the CDC Guidelines FHIR IG) with a new FHIR interface for Prescription Drug Monitoring Program (PDMP) data and a richer set of CQL that incorporates elements of the Washington State Rules for opioid prescribing.

The app was developed with funding from the WA State Health Care Authority, under the CMS Support Act, as is intended for use to improve access to and utility of PDMP data in WA State.  However, it is freely distributable under and open source license.  The app runs both as a standalone tool and in a SMART on FHIR environment as a back-end or “confidential” app.

The app can access FHIR medication resources through either the NCPDP SCRIPT 10.6/2017071 standards, in which case an internal FHIR facade exposes those medications as FHIR resources, or it can directly access PMP meds as FHIR resources.

We are seeking the participation of EHR vendors able to test the SMART on FHIR integration through implementation in their sandboxes or test instances, and PDMP vendors able to expose their data using either FHIR, or the existing NCPDP standards.   We are able to run against FHIR resources exposed from the WA State PMP test data, if no PDMP vendor participates.

Tools and Specifications

  • Smart on FHIR launch specifications from EHR/standalone launcher
  • CDC Opioid Guidelines FHIR IG.
  • CQL for WA State Rx rules

App Launch

The COSRI app, running as a SMART on FHIR app, uses the launch context of the EHR, or of a standalone launcher, for security/authentication, patient context, and clinical data on which the provider or CQL logic depend.  The token obtained for the EHR is then reused to provide secure access to API’s that support FHIR access to the PDMP, or other resources. This scenario will test the ability of the app, including it’s security model, to work in a variety of provider settings.

  • Action: The app is invoked during a patient encounter within an open session in an EHR through a link invoked by provider or provider staff.
  • Success Criteria: The app launches with correct patient context, and allows access to patient clinical data.  The opioid-relevant data will be displayed in the app.
  • Action:  The app is invoked directly in a standalone mode, thorugh a Smart on FHIR “launcher” or sandbox.  The 
  • Success Criteria:  The app launches with the correct patient context, and any stored opioid-relavent FHIR data will be displayed in the app.

PDMP queries exposed as FHIR resources 

The COSRI FHIR facade exposed the results of NCPDP SCRIPT queries as FHIR resources. In addition, the app may work with medication records exposed directly by a PDMP system.  Both routes will be tested if possible.

  • Action: The app’s launch provides a security context that can be reused to access internal services, including the FHIR facade to expose NCPDP SCRIPT queries as FHIR resources.  An NCPDP query is triggererd for the current patient and the patient’s currently dispensed meds are exposed as FHIR resources.
  • Success Criteria: PDMP query data are exposed as FHIR resources for the current patient context, and are displayed in the app.
  • Action: The app’s launch provides a security context that can be reused to directly access FHIR resources exposed by the PDMP/
  • Success Criteria: PDMP query data are exposed as FHIR resources for the current patient context, and are displayed in the app.

Clinical Quality Language rules run against EHR and PDMP FHIR resources


The app implements CDC guidelines on opioid prescribing (from the CDC Guidelines FHIR IG), and some Washington State Rules on opioid prescribing, as CQL logic.  The app will be tested to ensure that it is able to execute CQL regardless of launch method and regardless of PDMP access method.

  • Action:  The app will reason on clinical data used by CDC guidelines and WA rules, expressed in CQL 
  • Success Criteria: The app is able to correctly reason on those data using CQL rules.  Correct reasoning will be determined by comparison with Morphine Equivilent Dose calculated results and guidelines flags determined by manual inspection of hte data.
  • No labels