Difference between revisions of "PDDI-CDS FHIR IG Proposal"
|Line 114:||Line 114:|
<!-- 1-2 sentences describing the purpose/scope of the IG for inclusion in the registry -->
<!-- 1-2 sentences describing the purpose/scope of the IG for inclusion in the registry -->
This implementation guide drug-drug interaction . The a for and CDS ().
Revision as of 17:05, 8 August 2018
- 1 Potential Drug-Drug Interaction (PDDI) Clinical Decision Support (CDS)
- 1.1 Owning work group name
- 1.2 Committee Approval Date:
- 1.3 Contributing or Reviewing Work Groups
- 1.4 FHIR Development Project Insight ID
- 1.5 Scope of coverage
- 1.6 IG Purpose
- 1.7 Content location
- 1.8 Proposed IG realm and code
- 1.9 Maintenance Plan
- 1.10 Short Description
- 1.11 Long Description
- 1.12 Involved parties
- 1.13 Expected implementations
- 1.14 Content sources
- 1.15 Example Scenarios
- 1.16 IG Relationships
- 1.17 Timelines
- 1.18 When IG Proposal Is Complete
- 1.19 FMG Notes
Potential Drug-Drug Interaction (PDDI) Clinical Decision Support (CDS)
Owning work group name
Committee Approval Date:
TODO Please enter the date that the committee approved this IGproposal
Contributing or Reviewing Work Groups
- Clinical Quality Information
FHIR Development Project Insight ID
Scope of coverage
This project seeks to develop an implementation guide for potential drug-drug interaction (PDDI) clinical decision support (CDS). The implementation guide will specify both a knowledge representation format for PDDI logic and CDS services for PDDI with electronic health record (EHR) systems. Specifically, the implementation guide will include specifications for:
- How to represent PDDI logic in CQL and evidence in computational format using the FHIR Clinical Reasoning module.
- How to use CDS Hooks as a mechanism for EHRs to request PDDI CDS from CDS services.
The project may raise the need to create new FHIR resource(s) (e.g., a resource to represent drug interactions) and/or FHIR profile(s) (e,g, for PDDI context representation). The current scope is the universal realm. We will consider potential realm-specific work in the future.
New information regarding potential drug-drug interactions (PDDIs) is published every day in primary sources such as drug product labeling and the scientific literature. A PubMed search for publications indexed with the Medical Subject Headings keyword “Drug interactions” shows an average of 3,970 publications per year from 2000 through 2016. This suggests that the body of evidence about PDDIs is overwhelming and dynamic. Food and drug regulatory agencies in the United States, European Union, and Japan have issued guidance to industry recommending the use of drug product labeling to communicate potential involvements in drug interactions (Rekić et al. 2017). As it is impossible for clinicians to keep up with the PDDI evidence base, drug experts generate summaries of PDDI evidence from primary sources. These summaries bring PDDI knowledge to clinicians in the form of published drug information compendia, clinical decision support rules, and interaction checking applications. However, there are currently no broadly accepted standards to guide these experts in the knowledge representation and services implementation of PDDI information that would be most effective for clinical decision support. Without international coordination for the extensions and profiles used to express PDDI CDS, needless variation is certain to occur in implementations. This project will provide the opportunity for the international community to collaborate and come to consensus on patterns used in implementations.
This need for a standard representation of PDDI information was one of the topics addressed at two multi-stakeholder conference meetings/series (Hines et al. 2011; Scheife et al. 2015; Payne et al. 2015; Tilson et al. 2016). Attendees at both conferences included international stakeholders from drug information content providers, regulatory agencies, and academic organizations. Among the key recommendations was the following suggested set of core information that should be included for every PDDI mentioned in a clinically-oriented drug information resource (Payne et al. 2015):
- Drugs involved
- Clinical consequences
- Frequency of exposure to the interacting drug pair
- Frequency of harm for persons who have been exposed to the interacting drug pair
- Contextual information/modifying factors
- Mechanism of the interaction
- Recommended actions
- Seriousness rating
These core information elements are consistent with the results of a separate international Delphi study on how to improve the delivery of medication alerts within computerized physician order entry systems (Riedmann et al. 2011). This suggested list of core information elements includes some that are present in one or more of the 15 PDDI conceptual models analyzed in a recent comprehensive review by Herrero-Zazo, Segura-Bedmar, and Martínez (Herrero-Zazo et al. 2016). However, there is little commonality across the conceptual models on those elements that are included and no single conceptual model covers all 9 of the information elements. Moreover, the creation and maintenance of PDDI decision support generally requires considerable time and energy from highly trained domain experts. An additional need is to standardize the EHR context that is sent to CDS services. This includes context parameters that are currently not available but that, if present, would be useful for increasing the specificity of the CDS alerts. This is important to ensure that PDDI decision support can be implemented across a variety of systems.
Rekić D, Reynolds KS, Zhao P, Zhang L, Yoshida K, Sachar M, Piquette Miller M, Huang SM, Zineh I. Clinical Drug-Drug Interaction Evaluations to Inform Drug Use and Enable Drug Access. J Pharm Sci. 2017 Apr 21. pii: S0022-3549(17)30249-6. doi: 10.1016/j.xphs.2017.04.016. [Epub ahead of print] PubMed PMID: 28435142.
Hines LE, Murphy JE, Grizzle AJ, Malone DC. Critical issues associated with drug-drug interactions: highlights of a multistakeholder conference. Am J Health Syst Pharm. 2011 May 15;68(10):941-6. doi: 10.2146/ajhp100440. PubMed PMID:21546646.
Scheife RT, Hines LE, Boyce RD, Chung SP, Momper JD, Sommer CD, Abernethy DR, Horn JR, Sklar SJ, Wong SK, Jones G, Brown ML, Grizzle AJ, Comes S, Wilkins TL, Borst C, Wittie MA, Malone DC. Consensus Recommendations for Systematic Evaluation of Drug-Drug Interaction Evidence for Clinical Decision Support. Drug Saf. 2015 Feb. 38(2):197-206 http://link.springer.com/article/10.1007%2Fs40264-014-0262-8. PubMed PMID: 25556085. PMCID: PMC4624322
Payne, Thomas H., et al. "Recommendations to improve the usability of drug-drug interaction clinical decision support alerts." Journal of the American Medical Informatics Association (2015): ocv011.
Tilson H, Hines LE, McEvoy G, Weinstein DM, Hansten PD, Matuszewski K, le Comte M, Higby-Baker S, Hanlon JT, Pezzullo L, Vieson K, Helwig AL, Huang SM, Perre A, Bates DW, Poikonen J, Wittie MA, Grizzle AJ, Brown M, Malone DC. Recommendations for selecting drug-drug interactions for clinical decision support. Am J Health Syst Pharm. 2016 Apr 15;73(8):576-85. doi: 10.2146/ajhp150565. PubMed PMID: 27045070; PubMed Central PMCID: PMC50649
Riedmann D, Jung M, Hackl WO, Ammenwerth E. How to improve the delivery of medication alerts within computerized physician order entry systems: an international Delphi study. J Am Med Inform Assoc. 2011 Nov-Dec;18(6):760-6. Doi: 10.1136/amiajnl-2010-000006. Epub 2011 Jun 22. PubMed
Proposed IG realm and code
This specification will be maintained by the CDS work group.
This implementation guide (IG) is targeted at stakeholders who seek to increase the specificity and clinical relevance of drug-drug interaction alerts presented through the electronic health record. The approach is service-oriented and uses Web standards, a minimum information model for potential drug interactions, and emerging Health Information Technology standards including CDS Hooks, Fast Health Interoperability Resources (FHIR), and Clinical Quality Language (CQL).
This implementation guide:
- Describes knowledge and decision points for two PDDI CDS artifacts:
- Warfarin + NSAIDs
- Digoxin + Cyclosporine
- Provides structured code for CDS artifacts using current Health Information Technology (HIT) specifications:
- HL7 – Fast Health Interoperability Resources (FHIR)
- CDS Hooks
- Clinical Quality Language (CQL)
- Proposes and provides guidance to optimize PDDI CDS artifacts
W3C Healthcare and Lifesciences Working Group – volunteer-based task force (https://www.w3.org/wiki/HCLSIG/DDI )
This implementation guide is expected to be piloted this September (2018) at the FHIR CDS Hooks Connectathon and subsequent Connectathons. Potential implementers include University of Pittsburgh Medical Center, Wolters Kluwer Health, and EHR vendors.
- CDS Hooks (one of the lead authors of the CDS Hooks specification is part of the PDDI IG team)
medication-prescribe hook: Check for Potential Drug-Drug Interactions
Action: With the medication-prescribe hook, the EHR sends a request that may include prefetch data to the CDS Service. The CDS Service may request additional data from the FHIR server then returns an array of Card(s) with associated suggestions.
Precondition: A prefetch or FHIR server query for contextual data element (below) resources (e.g., MedicationDispense, MedicationAdministration, MedicationRequest, MedicationStatement, Patient, Observation, Condition). Preference for obtaining medication resources with prefetch query.
Success Criteria: The EHR displays the appropriate Card response for the specified indicator element.
Bonus point: The EHR supports Card actions (i.e., create, update, delete) relevant to drug-drug interactions that may include canceling the current drug order, substituting the current drug order, discontinuing a historical drug, and adding an additional non-drug order (e.g., lab). The EHR supports parsing the DetectedIssue resource as an extension to the CDS Hooks Card Attributes. The EHR supports medication-select – defined as the step when the provider selects the drug product (i.e., ingredient and strength) during order entry.
Specific data elements for Connectathon use case include: Current medications, Historical medications, Age, Serum digoxin, Serum creatinine, Potassium, Magnesium, Calcium, and History of Upper Gastrointestinal Bleed.
We plan to submit the specification for ballot for comments in September 2018 (see complete timelines on Project Insight)
When IG Proposal Is Complete
When you have completed your proposal, please send an email to FMGcontact@HL7.org