[ad_1]
Abstract
Study objectives Describe the implementation of an interoperable solution to support pharmacogenomic-guided prescribing in primary care in the National Health Service, England.
Methods We used an iterative approach to software development going through clinical workflow mapping, architecture design and development, and pilot-testing.
Results We configured a commercial health data management platform to store pharmacogenomic results in a structured format and created a knowledge base of pharmacogenomic guidance. This solution was deployed ‘as-a-service’ using an open application programming interface (API) specification, allowing third parties to receive pharmacogenomic results and guidance by querying the service using a patient identifier and medicine code. This was integrated with existing clinical decision support tools and presented contextual information to prescribers within their native electronic health record (EHR).
Discussion Pharmacogenomic results and guidance will be used across care settings and have greatest utility at the point of prescribing. This requires a solution, which separates the data from the applications, allowing integration with different EHRs through APIs.
Conclusions A vendor-agnostic standards-based solution can support the implementation of pharmacogenomic-guided prescribing across primary care.
Introduction
The safety and effectiveness of many medicines are impacted by common variations in an individual’s genome, a concept known as pharmacogenomics. Several gene-medicine relationships have been identified, and there is compelling evidence that pharmacogenomic-guided prescribing for common medicines such as statins and antidepressants improves outcomes, reduces adverse drug events and saves costs.1
The 2022 joint report by the Royal College of Physicians and the British Pharmacological Society, ‘Personalised Prescribing’, called for the implementation of pharmacogenomic testing in the National Health Service (NHS), recommending that results are made available to primary, secondary and tertiary care. This is a significant healthcare intervention and requires an interoperable solution, which can make results available to electronic health records (EHRs) across care settings, services and providers at the point of prescribing.2 3
This implementer report describes the design of a novel proof-of-concept solution to support pharmacogenomic-guided prescribing in primary care, including pilot-testing across two practices in the North West of England. This initiative was part of a wider project titled ‘pharmacogenetics roll out: gauging response to service’ (PROGRESS, ISRCTN15390784), based in Manchester, England.
Discussion
Pharmacogenomic results and guidance will be used across care settings and have greatest utility at the point of prescribing. We describe the implementation of a vendor-agnostic proof-of-concept capable of translating laboratory outputs into actionable insights, which can be surfaced within native EHRs as CDS alerts. This approach has the potential to add significant value to the prescribing of medicines by allowing personalisation based on individual genetic variation in routine practice.
Dolin et al described the use of CDS Hooks to prototype pharmacogenomic-guided prescribing within a commercial EHR. They identified several areas where FHIR or CDS Hooks lacked the semantic standardisation of terminologies, which could result in implementation ambiguity.5 We aimed to mitigate this by using established terminologies and only exchanging the patient identifier and medicine code between our solution and third parties. The design choice to not share the whole structured pharmacogenomic result significantly reduced complexity with no requirement to map multiple discrete data points between systems. However, given the importance of data standards, we are developing an open data model (using openEHR and HL7 FHIR) to express pharmacogenomic test results with the Global Alliance for Genomics and Health.6
Lessons learnt
Interoperability is a significant challenge for health systems globally, and sharing data across EHRs remains a significant barrier to the potential benefit of technology-enabled healthcare.7 In the context of the NHS, for novel clinical data, such as pharmacogenomic results, that have not been locked into proprietary data silos, there is an opportunity to implement solutions that are interoperable by design. A key contrast from previous data initiatives was our approach to minimise duplication of data across systems and avoid point-to-point integration. To scale this, there is a need to use an open standards strategy, which democratises data and fulfils the wide-ranging clinical applications, but also creates opportunities for population health and academic research. Critically, this requires decision-makers to recognise that storing data separate from applications will promote a culture of safe, secure and meaningful data sharing for patient and societal benefit.8 The approach described in this paper can be the foundation of developing formal solutions that enable the meaningful implementation of pharmacogenomics in the NHS and beyond.
Conclusions
An interoperable solution is critical to the successful implementation of pharmacogenomic-guided prescribing in primary care. The architecture should include a vendor-agnostic data repository and content management system to administer CDS alerts and guidance. The use of open standards, terminologies and APIs is a valid and feasible approach to develop such a solution.
[ad_2]
Source link
