We have detected that you are using Internet Explorer to visit this website. Internet Explorer is now being phased out by Microsoft. As a result, NHS Digital no longer supports any version of Internet Explorer for our web-based products, as it involves considerable extra effort and expense, which cannot be justified from public funds. Some features on this site will not work. You should use a modern browser such as Edge, Chrome, Firefox, or Safari. If you have difficulty installing or accessing a different browser, contact your IT support team.
BETA - Clinical Information Standards
Overview of Clinical Information Standards
Clinical Information Standards resources
Information standards for health and care organisations are published so that information about the health and care of individuals can be shared and compared across the health care sector, using data that are defined consistently.
With consistent data that conform to national and international standards, every subsequent step in the lifecycle (assembly, analysis, interpretation, distribution and change) is made easier for each system and each organisation involved.
Through the standardisation of data, we:
- improve patient care, patient outcomes and patient experience
- drive world class research to support the development of new treatments
- reduce the risk of harm resulting from information errors
- enable analysis to monitor, manage and improve population health
- provide intelligence to inform the design of effective and efficient services
- enable improved accuracy of information for funding and reimbursement
Governance arrangements for Standards
There is an alphabetical list of all current Data Coordination Board (DCB) and Information Standards Board (ISB) standards and collections. The list is updated on a monthly basis, following the approval of new items, and changes, by the DCB.
Recent publication activity can also be found on our publication and notification pages.
Data Set Change Notices (DSCNs) and Information Standards Notices (ISNs) published between 1991 and 31 March 2014 are available to view in an Excel document where the ISN is still current, it will also be listed in the publication and notification pages.
Data Coordination Board, approved Classifications and Terminologies Standards
The Secretary of State reaffirmed the requirement for healthcare systems to implement and use the DCB approved standards for:
ICD-10, 5th Edition is the current, approved standard for use in England. ICD permits the systematic analysis, interpretation and comparison of morbidity data collected in different areas. The specified purpose of the ICD-10 is to provide a means of classifying diagnoses and is defined as a system of categories to which morbid entries are assigned according to established criteria.
OPCS is an NHS Information Standard that supports various data collections, such as Central Returns and Commissioning Data Sets, as well as other secondary uses of information that are essential to planning and improving health care. The current, approved standard for use in England is OPCS-4.8
SNOMED CT is now the internationally pre-eminent comprehensive terminology for medicine. It is funded and managed by 34 subscribing countries and it offers more than 960,000 terms for over 360,000 different clinical notions covering:
- occupations, lifestyle and social circumstances
- signs and symptoms
dm+d provides a consistent description of medicines and medical devices, essential to ensure patient safety. dm+d is already an integral part of the electronic prescription service (EPS) and is currently being deployed in secondary care settings where it’s playing a key role in medicines management.
Technology reference data update distribution
NHS Digital uses the Technology Reference data Update Distribution (TRUD) mechanism to license and distribute reference data to interested parties for ICD, OPCS, dm+d and SNOMED CT. Access to this reference data requires registration. For more information on the reference data for these DCB approved classification and terminology standards see:
- ICD-10 eVersion and eVersion Book
- OPCS-4 eVersion Book
- SNOMED CT UK Edition
In addition to the standards approved by the DCB there are national standards for the structure and content of health and social care records. These cover, for example, hospital referral letters, handover communications, discharge summaries, and inpatient and outpatient letters. These standards have been developed by the Professional Records Standards Body, (PRSB), in partnership with the public, royal colleges and representatives from the professional bodies.
More information on the standards published by PRSB can be seen here.
The vision for NHS Clinical Information Standards
Our vision is that:
- clinicians can record, share and access patient information that retains the semantic fidelity (meaning) of clinical encounters
- commissioners, researchers and public health planners can access clinical information for ‘secondary uses’ with the rich detail of the clinical encounters
To achieve this vision, we need to build an 'ecosystem' where information standards work together to support the ‘end-to-end’ flow of patient information across the health and care system. Information standards must not be developed in isolation from each other, but must be designed, developed and implemented collectively.
Through the implementation of mutually reinforcing information standards, the NHS clinical information system will become a 'learning health system'. This will lead to improved patient outcomes, lower commissioning costs, improved population health and more effective research to drive innovation.
Figure 1: Patient information flow in the health and care system
The generalised flow of patient information across the health and care system is shown in Figure 1 and it is clear how the accurate representation, recording and sharing of patient information contributes to the improvement of patient outcomes.
Our vision of a 'learning ecosystem' treats the information recorded by health practitioners as an essential resource of the health and care system rather than a mere byproduct. Clinical Information Standards allow the potential of this essential resource to be exploited (recorded, packaged and shared) in ways that can be consumed by the processes and technologies that underpin a modern digital health and care system.
To achieve our vision we will focus on five key standards that will work together to allow information to flow through the system. These standards are:
There are many other standards that exist in the health and care system, but by driving the uptake of those listed above the NHS will become a high performing learning ecosystem.
Figure 2: Key standards for each part of the ecosystem
Figure 2 shows how the 5 information standards relate to the different parts of the ecosystem. Taken together they enable the consistent recording and sharing of detailed health and care information and the representation of that information in forms that are valid for statistical purposes.
NHS Digital will consult with the research, population and public health communities to determine what flows would benefit from increased granularity of information. A collaborative consultation will then begin to determine how the flows can be augmented with SNOMED CT level information or information at a lower level granularity that can be derived by aggregating SNOMED CT level data.
Getting there - achieving the vision
Our vision is founded upon the detailed and shareable electronic record of direct patient care now coming into being. It will only be achieved with considered, collaborative effort, and by incrementally building the capabilities across the system locally, regionally and nationally.
To illustrate the thinking on how to get to the stage where patient information captured at the point of care can influence patient outcomes nationally, we provide the following staged approach with key milestones demonstrating an iterative enhancement of capabilities.
Phase 1: Sharing and linking
Phase one focuses on embedding established standards and promoting their linkage, for example, dm+d with tracking standards, or UTL with SNOMED CT. In addition, it supports and enables Transfer of Care Interoperability initiatives by means of well-curated SNOMED CT, dm+d or UTL content for the FHIR information models and associated application programming interface (API) specifications. These measures will improve the volume and quality of shareable information that is available to health practitioners at the point of care, therefore helping them provide patients with better care and incentivising their ongoing efforts to capture still more and better quality data.
Phase one focuses on the following key objectives:
- Embedding tracking standards: barcodes, RFID etc are established standards, that are used widely in the NHS. Aligning these standards to national initiatives such as Scan4Safety and compliance with the Falsified Medicines Directive, are essential to achieve large scale efficiencies and improvements in patient safety.
- Deploying dm+d to secondary care settings, realising significant interoperability benefits to other secondary care sites and primary care, as well as local benefits around medicines management. dm+d is used extensively by the Electronic Prescription Service, is a nationally published standard, and is used widely in primary care. dm+d is a key enabler for medicines interoperability, providing increased safety around dose syntax, it integrates with tracking technologies and standards (e.g. barcodes, RFID) for additional safety.
- Increasing the use and uptake of SNOMED CT by implementing FHIR profiles with SNOMED CT information model content. This will enable richer and consistent national information exchange, as described in the Transfers of Care section under Interoperability.
- Complete and deploy the Unified Test List (UTL) and the associated FHIR profile specifications for exchanging test request or their results, and work with the NHS suppliers, labs, hospitals and GP practices to migrate from the existing messaging and test coding products to UTL. The UTL initially covers Haematology and Clinical Chemistry, with genomics, echo, ECG, microbiology and endoscopy likely to be included in the next stage of the work. The work will cover the associated Data Models, Units of Measure, FHIR specifications, and will provide Code Lists/Catalogues for publication and use by the NHS.
- Identify a replacement for OPCS-4 (such as the WHO International Classification of Health Interventions currently being developed) that builds on the best features of OPCS-4 and complements the richness of ICD-11 and providing for better links with SNOMED CT codes captured at the point of care. The current structure and design increasingly prohibits the appropriate further development of the classification needed to reflect more recent changes in medical practice whilst retaining the core principles of a classification for statistical purposes. We will carry out an options assessment in consultation with professional bodies and the market to identify the replacement for OPCS-4. As OPCS-4 is a mandated standard and a requirement for national data sets, in England, Northern Ireland, Wales, Scotland, and a key component for CDS, National Service Frameworks, Care Pathways, Performance Indicators and other Central Returns, the replacement will need to be planned carefully and in collaboration with the NHS.
- Modify the national systems so that they also allow information coded in SNOMED CT, dm+d, UTL to flow to, be received by and processed at the centre in alignment with the National Data Architecture. Such systems include CDS, SUS, HES, and will eventually include CASEMIX (not in phase 1).
Phase 2: Optimising and augmenting
Phase two focuses on incrementally increasing the value that the health and care system can extract from having SNOMED CT coded patient-level data (we include here dm+d and UTL under SNOMED CT). We do this by offering incentives and using levers for system suppliers to natively implement SNOMED CT in their systems. This can only be done if professional bodies, NHS representatives and system suppliers collaborate on careful, iterative and setting-by-setting designs. It also requires tooling and national infrastructure services that allow effective management of SNOMED CT capabilities that would otherwise be daunting and burdensome for providers and suppliers, including some prototype advanced services to be provided centrally in the earlier years of the journey but that may in time migrate back to suppliers.
Phase two also focuses on developing the National Reporting stage of the ecosystem (see Figures 1 and 2) where the national returns begin to transition toward landing raw SNOMED CT data rather than post-transformed data. This is a significant element of the vision, which will eventually provide the increased granularity to secondary uses flows and therefore positively impact patient outcomes through more informed commissioning, research and public health planning. During Phase two we will be working closely with commissioners, researchers and public health professionals, with pilots targeted at those secondary uses flows that would greatly benefit from more granular information. We would then work with pilot sites from the NHS to develop the required enhancements of primary data capture and flows and then jointly evaluate the result with the requesting organisation to develop a value proposition for similar augmentation of other secondary uses flows.
The key objectives for Phase two are:
- Introduce detailed SNOMED CT coded care records natively into NHS trusts in collaboration with professional bodies, suppliers and NHS leaders. In parallel, provide advice or guidance on implementing natively SNOMED CT in different settings, and deliver a national capability that addresses the key technical challenges around Reference Sets.
- Continue implementing dm+d and UTL across the NHS and expand UTL to cover more areas under diagnostics.
- Under the national terminology services, develop a national capability for supporting the transformation of SNOMED CT coded records to ICD-11 (or the OPCS-4 replacement), using new technologies to increase the level of accuracy achieved through such semi-automation using national data flows to inform the transformation output and maintain the quality of information. Work closely with NHS Clinical Coders to establish the balance between National and local content management, in particular in cases where the derivation of classification codes is complex and requires knowledge of the patient’s full history.
- Deliver pilots and demonstrator sites using the new procedure classification standard with SNOMED CT. Subject to approval by the Data Coordination Board, publish the new standard and timetable the migration to it, and formal withdrawal of the existing OPCS-4 standard. The SNOMED CT returns will be mapped to the new classification standard at the national system level.
- Continue the delivery of enhanced national terminology services, including particularly the introduction of SNOMED post-coordination capability at the national level together with guidance, advice, and direction for implementing post-coordination into native SNOMED CT systems, in close collaboration with suppliers, users and professional bodies.
Phase 3: Enhancing commissioning flows
Phase two demonstrated the value of enhanced secondary uses flows, and established a SNOMED CT foundation layer in direct care settings, where increasing levels of detail in clinician-driven information creation drives and enables improved clinician-to-clinician interoperability and richer and more flexibly designed national returns and secondary uses. At the end of Phase two we will have a number of requests for enhanced flows for research and public health planning that the system will be able to deliver.
Phase three will enhance the commissioning information flows which have a number of specific requirements before changes are made. The level of granularity that SNOMED CT may introduce will need to be matched by the ability of NHS Improvement and NHS England to provide tariffs at the same level of granularity. Given that SNOMED CT is up to 100 times more granular than HRGs, this may not be feasible, practical or desirable. NHS Improvement, NHS England and NHS Digital will therefore need to collaborate on the development and implementation of new techniques capable of discarding the raw detail and richness that a SNOMED CT set of returns may offer, but in a controlled and understandable way. This new approach will need to be piloted and then run in parallel with the existing commissioning flows to ensure that it does not materially distort the existing commissioning flows in unanticipated ways.
By Phase three we expect that transforming from individual records of direct care to classifications (ICD-11 and the replacement for OPCS-4) will be implemented as a predominantly national service provided by us, which will receive returns and convert them to classifications codes for core reporting and robust statistics. Quality issues will be addressed through collaboration with local coding experts.
Clinical information standards indicative roadmaps
Achieving our vision will require collaborative work with the wider NHS, professional groups, suppliers, commissioners, regulators and the Department of Health and Social Care.
We present below an indicative set of roadmaps for each Clinical Information Standard and outline the key activities and initiatives we believe to be essential to achieve the vision. While all of these activities contribute towards the creation of a “learning ecosystem”, not all of them have been formally initiated. They are provided to illustrate the direction we want to travel, and the long term perspective required for clinical information standards for the health and care system.
Barcoding standards are sublicensed by NHS Digital to any NHS organisation requiring a unique scannable code for its assets, people and places. The most common example of these standards in action is the barcoding on patient bracelets, following the Global Service Relation Numbers (GSRNs) to identify patients and clinicians. The barcoding standards have been used locally by the majority of NHS organisations and recently have also become the focus of the DHSC initiative Scan4Safety. The project provides NHS trusts with the tools, financial backing and know-how to identify every patient, product and place within the healthcare setting using global standards. This provides additional visibility of how care is administered to the patient, when and by whom. This, in turn leads to improved patient care and experience, reduction in errors, and increased operational efficiency.
SNOMED CT was published as an NHS information standard in August 2011. SNOMED CT is an internationally funded and maintained clinical terminology and is now the pre-eminent vocabulary used by Health IT systems globally when seeking to document and support the direct management of the health and care of an individual. SNOMED CT enables the representation of clinically relevant information consistently and reliably in a way that can be processed by IT systems. This enables IT systems to:
- exchange data across the health and care environment
- provide clinical decision support tools
- undertake enhanced analytics to support effective delivery of high-quality healthcare to individual people and populations
SNOMED CT is already required to be used for communicating clinical content across health and care within the NHS standard contracts, and is also stated as the required terminology to support direct management of care within the policy document published by the National Information Board (NIB): Personalised Health and Care 2020: A Framework for Action.
The wording above specifies that SNOMED CT should be used when recording clinical content as part of direct care provision and when sharing information that contains clinical patient information. This contrasts with the current standard, which leaves to the NHS organisations whether the IT system they have uses SNOMED CT natively or whether it has its own local representation and then maps between this and SNOMED CT for sharing or extraction.
The standard is specific to the context of direct care and does not cover secondary uses returns or extracts or collections, which are governed by separate standards.
In making the vision of a “learning ecosystem” a reality, we are confident that all health care settings must be able to use SNOMED CT to record information and share information. We also believe that in some instances, settings and use-cases it is necessary to require that IT systems have native representations of SNOMED CT though in others a conversion from a local internal representation to SNOMED CT will be adequate.
We acknowledge that the current mechanisms of publishing SNOMED CT require advanced levels of digital maturity for providers and their IT systems. We will be introducing alternative mechanisms to reduce the technical burden of using SNOMED CT to represent clinical information in systems. Methods we are currently exploring are:
- A national terminology service which maintains the existing clinical information standards (SNOMED CT, dm+d, UTL, ICD-10, OPCS-4 and maps from SNOMED CT to ICD-10/OPCS-4) and allows the local system to request or search specific codes using APIs. This is a significant enabler as it offers providers a means of deriving SNOMED CT coding without having to fully implement the SNOMED CT ontology on their systems.
- New Technology methods of extracting SNOMED CT information from the electronic health record. Artificial Intelligence, Natural Language Processing and similar technologies will allow the extraction of some level of SNOMED CT coding from some types of clinical note or sound recording, and from a variety of other similarly unstructured documents. The extracted coding will require manual quality assurance by clinical coding experts where the derivation confidence is not high, or the clinical risk of even relatively low error rates is considered too high.
- Introducing methods to implement SNOMED’s “post-coordination” capability in specific settings/use-cases. Post-coordinated expressions in SNOMED CT allow novel or unusually precise clinical notions to be dynamically represented as a structured composite of several SNOMED CT concepts. This contrasts with “pre-coordination”, the current representation method in SNOMED CT and older terminologies, which only allows for a clinical notion to be captured if it is one of those already within the predefined and relatively static list of SNOMED CT concepts. While pre-coordination is easy to implement, it is not considered scalable, requires careful management and maintenance of subsets for each context, and does not take full advantage of the medical knowledge deliberately represented within the full SNOMED CT product precisely to support post-coordination.
In terms of sharing clinical information, we will be providing curated FHIR profiles which determine in their information models how SNOMED CT information should be used. The profiles will be curated collaboratively with NHS Subject Matter Experts, professional groups, suppliers and technical experts to ensure that design, clinical safety and effectiveness are key components of the design.
Direct care considerations
We believe that beyond the existing SCCI SNOMED CT standard we will need to define how the Health and Care system uses a hybrid of SNOMED CT representations including “hard” and “soft” implementations of SNOMED CT.
“Hard” SNOMED CT requires that the IT System is able to natively use SNOMED CT. “Hard” SNOMED CT adherence would be required in instances of itemised transactional activities (for example Lab Order-Comm, prescribing, and so on) where it is essential for patient safety, or for efficiency, to have absolute certainty that semantic equivalence could not be compromised as a result of mapping to SNOMED from an alternative internal representation. “Hard” SNOMED CT would establish references to limited SNOMED CT valuesets in tightly-defined models, similar to the Emergency Care Dataset (ECDS) returns, or a standard model for blood pressure or allergy representation.
In contrast with these instances, there are many interoperability instances where a “soft” SNOMED CT adherence approach would be acceptable (for example record extracts, transfers and correspondence) and would offer a less onerous migration path for providers and their system suppliers. A “soft” SNOMED CT approach would support a general record structure with a number of sections and attributes, each constrained to content that may be structure or unstructured, but where coding is included, this must be done in SNOMED CT.
As SNOMED CT use increases across the NHS, there will be a variety of implementation types - “hard” vs “soft” vs “aspirational” (neither strictly hard nor soft, but working with mappings); precoordinated vs. post-coordinated implementations - we must resolve an additional risk that will emerge from this hybrid landscape, that of Semantic Equivalence. There may be different representations of the same clinical concept in SNOMED CT that arise through different adherence approaches.
For example, SNOMED CT already contains a single pre-coordinated code for “Appendectomy”. From Phase three, the same activity may also be described by a longer post-coordinated expression consisting of five other more elementary pre-coordinated codes “Procedure : Site = Appendix, Method = Excision”. Without a means to detect that these two descriptions are in fact interchangeably equivalent, a receiving system still designed to use only the shorter pre-coordinated representation paradigm may be unable to correctly represent the information transferred to it if that arrives instead only as the longer and post-coordinated expression. We propose to address this challenge by being explicit of the cases where “hard” vs “soft” SNOMED CT should be used and, at the same time, providing adequate implementation guidance when FHIR profiles are curated. We believe that FHIR profiles should incorporate curated SNOMED CT information models and specifications for all flavours of pair-wise systems interoperability set to be encountered as the NHS journeys through the phases.
SNOMED CT is increasingly used in national initiatives, applications and systems. It is currently being rolled out in primary care and is increasingly being used in datasets and collections (for example Mental Health Services Dataset, Emergency Care Dataset, etc) and as part of the specifications and requirements for the Local Health Care Record and Global Digital Exemplars (LHCREs and GDEs). SNOMED CT is also used (alongside OPCS-4) in the National Institute for Health and Care Excellence (NICE) Interventional Procedure Guidance and Medical Technology Guidance. The national Transfers of Care programme is also using SNOMED CT in its specifications for the exchange of information from Secondary Care to Primary Social Care. Increasingly, local and regional organisations with high digital maturity are implementing SNOMED CT. The CCIO7 initiatives place significant emphasis on using SNOMED CT to support laboratory tests requests and results flowing across the system.
We will be using the SNOMED CT coded, patient-level information captured and recorded at the point of care to derive classifications such as ICD-10/11. Maps in data files are already provided from SNOMED CT to ICD-10 and OPCS-4 to support semi-automation for simple and routine hospital activity but the full clinical notes are necessary to apply the classification business rules to ensure consistent and comparable data for analysis. A range of automation tools and services will be created to achieve this derivation (ranging from SNOMED CT to ICD-11 maps, to AI/NLP services to the use of a national Terminology Server, etc).
We therefore include in the vision for Clinical Information Services, the flow of SNOMED CT through to Secondary Uses. This is expected to take place in the following scenarios:
- datasets and collections will use SNOMED CT coding when they contain clinical information
- national systems (for example SPINE, ERS, CDS, SUS, HES) will be modified to flow SNOMED CT coded clinical information
- translation of SNOMED CT to classifications (for example ICD10/11) will be performed by national services, working closely with providers to resolve low confidence or high-risk translations
- new information standards will be published to introduce SNOMED CT to key datasets and returns that currently exclude SNOMED CT (for example CDS returns).
- commissioning, public health and research flows will all be derived from SNOMED CT, and will be able to provide granularity up to SNOMED CT level as appropriate thereby augmenting or replacing the existing level of nonreversible aggregation
Indicative SNOMED CT roadmap
The roadmap for SNOMED CT indicates the activities that will help establish a SNOMED CT interoperability-based foundation layer across the health and care system that will influence how secondary uses and therefore commissioning, public health and research can deliver improvements in patient outcomes.
Phase 1: sharing and linking
- NHS terminology server with SNOMED CT content and APIs
- FHIR curation process established and SNOMED CT only information models enacted
- Modify national systems (CDS, SUS, HES) to allow flow of SNOMED CT
- Establish RefSet management and maintenance preferred approach
- Establish "hard"/"soft" SNOMED CT adherence criteria
- Deliver improved RefSet maintenance and management services/tooling
- Pilot new technologies in deriving SNOMED CT at direct care settings
- Assess options for representing post-coordinated concepts in human readable form
- Begin amending non-compliant national standards to allow SNOMED CT returns
- Pilot/proof of concept for SNOMED CT adherence
- Enrich and harmonise analytical quality of SNOMED CT
- Introduce new standards on "hard"/"soft" SNOMED CT adherence
- DSP and new dataset extraction and collection only accepts SNOMED CT for clinical information
- Consult with public health, research and commissioning for SNOMED CT augmented Secondary Care Flows
- Establish approach for Equivalence Detection
- Collaborative design of Equivalence Detection with PRSB, Interopen and NHS reps
- Implement approach for Equivalence Detection
Unified test list
To provide a better understanding of why a Unified Test List standard is required, we provide some background on the existing NHS mechanisms for sharing pathology test requests and results.
In the mid 1990s, the Pathology Messaging Implementation Project, (PMIP), introduced universal electronic delivery of certain pathology results to UK GPs. It has two core subcomponents:
- A Pathology Bounded Code List (PBCL) - a fixed list of 3,352 orderable blood tests originally curated jointly with the Royal College of Pathologists. It is coded against a subset of the READ codes, a legacy standard clinical terminology whose maintenance ended in April 2016. The PBCL is therefore now necessarily static: no new tests can be added to it.
- EDIFACT - a text based but non-SGML electronic data interchange standard approved by the ISO in 1987 and still widely used in many industries.
PBCL codes identify the named or coded parts of some lab results that are sent electronically back to primary care from labs. Coverage is restricted chiefly to mainstream blood sciences and so excludes for example microbiology and histopathology. Following several years of PMIP design work, EDIFACT was adopted around 2003 for use in transfer of Pathology and Laboratory messages from labs to primary care. It has since been used to carry mostly lab results to primary care, but has also been extended for use in Cancer Outcomes Services Dataset. The current PMIP-EDIFACT specification can only carry READ codes, and this is therefore a problem since no new READ codes - and therefore no new tests - can now be added to the PBCL.
Since PMIP, modern design has moved to new paradigms like Representational State Transfer Application Programming Interfaces (REST APIs) based on JSON. This current estate based on PBCL and EDIFACT is therefore to be replaced by a similar combination of a curated Unified Test List drawn from SNOMED and with a broader scope (for example including microbiology) and a messaging syntax based on HL7 FHIR.
The current programme of development work under the CCIO7 portfolio is driven by user needs, including:
- clinical need - the ability to share basic pathology results across health and care
- accounting and commissioning - the ability to unambiguously identify tests and associated results to support commissioning and accounting
- patient safety - the ability to interpret and analyse aggregated lab results where the same test can be returned to different clinicians in more than one unit of measure
This broad vision can be explained using an example. Imagine a clinical statement or observation below needs to be transferred:
Haemoglobin concentration of 80 gm/L
This statement can be split into the following aspects shown in the roadmap below.
- Coded Test Result - 1022431000000105 Haemoglobin estimation.
- Value - 80 (a quantitative measure – but tests can also have qualitative results like ‘high’, ‘low).
- Units of measure - gm/L (this could be expressed as gm/dL, which would result in a corresponding change in the value from 80 to 8).
- A message wrapper that contains this statement which can be then passed from one system to another.
The requirement to satisfy these broad use cases and the collection and coding of all test names and associated results under the UTL is only one aspect of the broader CCIO7 Pathology project. NHS Digital is working on three distinct areas around standards and interoperability:
- Unified Test List
- FHIR Message Profiles
- Standardised Units of Measure
Since the scope of diagnostic medicine is quite broad, this project under the guidance of the CCIO was currently scoped to include only basic pathology - haematology and clinical chemistry – and to build sound informatics principles for lab medicine in this basic pathology context. Since these specialities form the bulk of lab messages, they are also an appropriate place to start. With sound principles in place it should be possible to incrementally extend the UTL and the associated FHIR standards in a similarly principled way to support other specialities including Microbiology, Histopathology, genomics, echo, ECG, microbiology and endoscopy.
The indicative roadmap for the UTL
Phase 1: Sharing and linking
- Early version of Units of Measure list
- Limited, alpha version of UTL available for review with stakeholders
- Cross-system curation of FHIR profile specification
- Publish validated FHIR specification for pathology
- Share with key stakeholders early draft of the UTL as part of a published FHIR specification
- Publish version 0.01 ot UTL, FHIR profile specifications, Units of Measure
- Early discovery and first-of-type testing of UTL and FHIR profile specifications with NHS organisations
- Early availability of PathologyConnect portal
- Publish version 1.00 of UTL, FHIR profile specification, Units of Measure and corresponding APIs
- UTL published as a national standard
- UTL is available through the National Terminology services
- NHS organisations and suppliers implement the UTL standard
- Publication of implementation support, use-cases, webinars
- Microbiology content complete: Data model FHIR profile code list
- Histopathology content complete: Data model FHIR profile code list
The UTL is expected to be developed and published at least for Haematology and Clinical Chemistry during Phase one of the overarching Clinical Information Standards roadmap, in time allowing the graceful withdrawal of PBCL/EDIFCACT. Subject to there being a confirmed user need to develop Microbiology and Histopathology content in UTL, the enhanced versions of the UTL standard can also be delivered within Phase one.
We are taking a pragmatic, implementation focussed approach to the development and delivery of the UTL and associated FHIR specifications. We will co-develop all standards-based specifications by working closely with stakeholders including professional bodies, INTEROPen, suppliers, clinicians and informaticians.
This approach will be supplemented by iterative development of the specifications themselves, taking an Agile approach. One key aspect of this work will be to work closely with suppliers and trusts to do discovery work and prototype testing. We have also identified such work that will be done in conjunction with LHCRE sites (Greater Manchester), GDE sites (Cambridge) and middleware suppliers (X-Labs/NPEx). All feedback gathered will be fed into the iterative development process, so the UTL and associated standards can be improved. One key aspect of our approach will also be to engage with the Royal College of Pathology (RCPath) via PRSB to ensure adequate governance of the UTL. NHS Digital will also be keen to adopt and adapt as needed learning from the RCPath on experiences from the National Laboratory Medicine Catalogue (NLMC). We believe that input from RCPath and the feedback from LCHRE, GDE sites and early adopters will ultimately shape the content and scope of the UTL making it a standard better suited for use in NHS.
Enablers, collaborations and dependencies
UTL development will be guided by its stakeholder groups, and its success further enabled by ensuring that the relevant sections of the FHIR specification for exchanging pathology result data is co-curated within INTEROPen by representatives from the supplier, informatics and professional groups communities. To ensure that the resulting standards work for all of the NHS, the FHIR specifications will need validation by all those who need to exchange results of any kind.
There is interest in the UK and abroad in ensuring SNOMED CT supports diagnostic medicine. Since the UTL will be based on SNOMED CT, several other countries are looking to the UK for leadership and direction. NHS Digital are collaborating with international partners including Norway, Sweden, New Zealand and Australia to ensure that the principles we adopt in the development of UTL are applicable internationally. This is of particular relevance to emerging disciplines like genomics and for any FHIR specifications we publish. In addition to establishing the UK as a leader in this area, it also ensures that the UTL and FHIR designs we adopt will be portable internationally, therefore making them easier to implement by the international IT system suppliers.
ICD-11 and the new Procedure Based Classification (PBC)
ICD-11 and the new Procedure Based Classification (PBC)
ICD-11 and the new Procedure Based Classification (PBC) are used to classify patient episodes according to diagnosis or procedure, for performance management, reporting, remuneration, and also population health and public health planning.
The United Kingdom has a long history of using the International Classification of Diseases and revisions and as a World Health Organization (WHO) Member State has a mandatory obligation to collect and submit ICD morbidity and mortality data to the WHO for the production of international statistical and epidemiological data (WHO Nomenclature Regulations).
WHO’s constitution mandates to establish and revise as necessary international nomenclatures of diseases, of causes of death and of public health practices, and requests Member States to submit annual reports on health status and actions taken to improve health.
Accurate and timely data are an essential resource for Member States to achieve the Sustainable Development Goal (SDGs) targets and goals for Universal Healthcare Coverage (UHC), health emergencies and healthier populations. WHO is the steward and custodian for monitoring the health-related SDGs. Data are needed to measure performance, improve programme decisions and increase accountability.
For this purpose, WHO sets and maintains data collection standards that are needed by Member States, including the WHO Family of International Classifications that include the ICD as well as the International Classification of Functioning, Disability and Health (ICF) and the International Classification of Procedures (ICHI) under development
Morbidity statistics: ICD is a mandated NHS Standard (SCCI0021) published under Section 250 of the Health and Social Care Act 2012 to code the diagnoses of all inpatient episodes and day cases to support several data collections including Admitted Patient Care Commissioning Data Sets, Central Returns and Secondary Uses Service. ICD is a vital component of national data sets, such as Hospital Episodes Statistics (HES) in England, Hospital In-patient Statistics (HIS) in Northern Ireland, Patient Episode Data for Wales (PEDW), Scottish Morbidity Records (SMR), Cancer Registries, National Service Frameworks, Care Pathways, Performance Indicators, Commissioning Data Sets (CDS) and other Central Returns.
In England, ICD codes underpin the Healthcare Resource Groups (HRGs), to generate diagnoses-driven HRGs. The National Casemix Office (NCO) designs and refines the classifications used to describe NHS healthcare activity in England. These classifications underpin the national reimbursement system from costing through to payment and support local commissioning and performance management. NHS managers and health care professionals in secondary care also use it locally to support operational or strategic planning and performance management.
More information on the National Casemix Office and HRG tools can be seen here.
Maps from SNOMED CT to ICD-10 are currently provided to suppliers and technical professionals to incorporate the business rules for consistent code selection and sequencing into NHS patient administration and other systems to support Commissioning Data Sets, Central Returns and support the derivation of classifications from clinical records. There are four types of map, two of which may be automated for routine hospital activity, the other two types of map support the coding of complex activity and require manual intervention by coding experts. Utilisation and the opportunity for derivation of classifications for semi-automation is dependent on the maturity of the local clinical system and comprehensive, quality structured clinical information recorded in the system.
ICD updates are currently implemented on a three yearly cycle as determined by the Department of Health and Social Care.
Plans for future
The WHO launched the latest revision (ICD-11) which was released on 18 June 2018. This release is described as an advance preview that will allow countries to start building transition plans for implementation, prepare translations, and train health professionals all over the country. The priority for WHO was to provide ICD-11 for Mortality and Morbidity Statistics referred to as ICD-11-MMS.
The WHO Executive Board (EB 144) will receive a summary report in January 2019 on ICD-11 recommending ICD-11 submission to the World Health Assembly (May 2019) for adoption to come into effect on 1-Jan-2022. WHO expectation is that Member States will start reporting thereafter in ICD-11 but recognise the actual implementation date for each Member State will vary.
In England, ICD-10 is used in secondary care, and each NHS organisation must dedicate resources to review relevant clinical notes and patient histories and assign the appropriate ICD-10 codes. NHS Digital will provide guidance and business rules to support appropriate allocation of patient episodes to the appropriate codes which are subsequently used for remuneration under the Payment by Results (PbR) mechanism.
In planning for ICD-11, we need to recognise that ICD-10 code derivation from clinical notes and patient histories is a significant investment for many providers. In addition, uptake of SNOMED CT will add additional burden to upskill clinical coders, who will also then need to upgrade their knowledge of ICD-10 to ICD-11. We expect that this would represent a very expensive exercise, as it would require significant investment in training, upskilling, and also a significantly increased workload for clinical coders initially.
In considering the vision of a “learning ecosystem” it is important to aim for a future state where the majority of ICD-11 codes are derived from SNOMED CT codes that the health care practitioners have entered into the provider system, can link to patient history from a multitude of systems, and is supported by innovative technologies (for example natural language processing, Artificial Intelligence, etc) to automate as much the ICD-11 code derivation as possible.
We have yet to determine whether the derivation of ICD-11 codes would take place at the provider site or done centrally. It is clear that regardless of where the derivation takes place, the provider clinical coders will be key in ensuring the data quality of the derivations, in particular where the confidence of derivation is low.
An indicative roadmap for ICD-11 is shown in the Indicative roadmap for ICD-11 and the new PBC standard, found below, combined with the new Procedure Classification Standard. The key points to note on the roadmap are:
- Providing maps from SNOMED CT to ICD-11 to enable local use and derivation where it exists already or is part of local plans to derive ICD.
- ICD-11 migration will likely happen after the uptake of SNOMED CT by NHS providers has reached a “critical mass”.
- ICD-11 migration will be predominantly driven by the automatic derivation of the majority of ICD-11 codes, with quality assurance and resolution of low confidence derivations by the provider clinical coding experts.
- Initially, ICD-11 derivation will take place at the provider systems. We believe that as the technology and collaborative working practices improve in the NHS, the end-state of the ICD-11 migration will be that most derivation takes place in NHS Digital, with collaborative efforts from provider clinical coders for QA purposes and resolution of low confidence derivations.
- ICD-11 will be the basis for HRG derivation until the point that HRGs can reliably link to SNOMED CT coding.
- ICD-11 will continue to be the format for reporting to WHO.
- We will work with the Research community and Public Health England to determine whether secondary uses flows could be augmented with increased granularity, as SNOMED CT level data would be available. We are sensitive to the needs to maintain historical compatibility, and we would explore a number of options to allow the transition to more granular data.
ICD is embedded in secondary care hospitals as part of the Patient Administration Systems. The implementation date of ICD-11 will depend on:
- proper assessment of the impact on official health statistics
- availability of maps from SNOMED CT to ICD-11 to support semi-automatic derivation of ICD-11, dependent on the maturity of local EPR systems
- impact assessment on workforce transformation and supplier implementation, and to be conditional on the solution of any outstanding issues
Availability of links from SNOMED CT to ICD-11 to support derivation of ICD-11 dependent on the maturity of local EPR systems. Maps from SNOMED CT to OPCS-4 are currently provided to the NHS to support the derivation of classifications. This supports the semi-automation for routine hospital activity and manual intervention by coding experts for the complex activities. Utilisation is dependent on the maturity of the local electronic health record and extent and quality of clinical information provided in the electronic record.
Availability of the National Terminology Server which could carry out the SNOMED CT to ICD-11 derivation centrally via APIs from the local EPR systems.
Additional improvements that we will aim to deliver along with the introduction of ICD-11are:
- capture advances in health science and medical practice
- maintain Core classification features of jointly exhaustive and mutually exclusive and residual categories
- new chapters to better identify: conditions to sexual health, sleep-wake disorders, traditional medicine, disorders of immune system
- better use of the digital revolution - improved incorporation into electronic environments
- coding, browsing, translation, mapping and proposal online tools; remove dependency on paper-based publications
- improved usability that requires less training
- better links from SNOMED CT to support derivation of ICD-11 dependent on the maturity of local EPR systems
Indicative roadmap for ICD-11 and the new PBC standard
ICD-11 and new Procedure Based Classification Standard
Phase 1: sharing and linking
- ICD-11 published by WHO
- SNOMED CT to ICD-11 maps using computational algorithms as a starting point
- ICD-11 and OPCS-4 and maps (from SNOMED CT) are part of National Terminology Server
- AI framework published for deriving ICD-11 codes from SNOMED CT
- OPCS-4 replacement recommendation published and consulted on
- Pilot ICD-11 derivation from SNOMED CT using AI
- CDS modified to allow SNOMED CT returns
- Consultation with NHS on other use-cases requiring ICD-11 and OPCS-4
- Comparative analysis and testing of SNOMED CT returns instead of OPCS-4/ICD-10
Phase 2: optimising and augmenting
- New Procedure Based Classifications design/config and population complete
- Proof of concept for ICD-10 derived from SNOMED CT returns
- Parallel run of new PBC/ICD-11 and HRG
- Proof of concept for new Procedure Based Classification use from SNOMED CT returns
- Publish new DCB standard for PBC/ICD-11
- NHS switches from OPCS-4 to SNOMED CT returns
- Migrate to ICD-11 for HRG, research and public health flows
- Pilot and dual run HRG using SNOMED CT
- Switch to PBC is live
- HRG using SNOMED CT is live
The new Procedures Based Classification
The UK has a long history of using a procedure-based classification. The earliest classification of surgical operations in England was issued by the Medical Research Council in 1944. The OPCS-4 Classification of Interventions and Procedures, the latest procedure-based classification is currently used widely across the UK. OPCS-4 underwent a significant update in 2006 to support the Government’s reform of NHS funding, previously referred to as Payment by Results.
The classification comprises a list of alphanumeric codes organised into mainly anatomically based chapters to support aggregation of individual categories for statistical purposes. When OPCS-4 codes are linked to other data items, such as procedure data, commissioner and provider of health care services, diagnoses via ICD-10, consultant code and patient’s postcode, they become a powerful tool for managing and improving the care of persons and populations.
OPCS-4 is a mandated NHS Standard (SCCI0084) published under Section 250 of the Health and Social Care Act 2012 to code the interventions and procedures of all inpatient episodes and day cases to support several data collections including Admitted Patient Care Commissioning Data Sets, Central Returns and Secondary Uses Service.
OPCS-4 is a UK-only procedure classification standard and does not provide any links with other international procedure classification systems. It is adopted by other countries under OGL arrangements.
OPCS-4 is currently a vital component of national data sets, such as Hospital Episodes Statistics (HES) in England, Hospital In-patient Statistics (HIS) in Northern Ireland, Patient Episode Data for Wales (PEDW), Scottish Morbidity Records (SMR), National Service Frameworks, Care Pathways, Performance Indicators, Commissioning Data Sets (CDS) and other Central Returns. OPCS-4 is also used (alongside SNOMED CT) in the National Institute for Health and Care Excellence (NICE) Interventional Procedure Guidance and Medical Technology Guidance.
In England OPCS-4 codes underpin the Healthcare Resource Groups (HRGs) which are used for NHS reimbursement. NHS managers and health care professionals use it locally to support operational and strategic planning and performance management. OPCS-4 updates are implemented on a three yearly cycle as determined by the Department of Health and Social Care.
Maps from SNOMED CT to OPCS-4 are currently provided to suppliers and technical professionals to incorporate the business rules for consistent code selection and sequencing into NHS patient administration and other systems to support Commissioning Data Sets, Central Returns and to support the derivation of classifications from clinical records. There are four types of map, two of which may be automated for routine hospital activity. The other two types of map support the coding of complex activity and require manual intervention by coding experts. Utilisation and the opportunity for derivation of classifications for semi-automation is dependent on the maturity of the local clinical system and comprehensive, quality structured clinical information recorded in the system.
Plans for the future
The current code structure constrains the development of the classification to reflect changes in medical practice for aggregation purposes. A new procedure-based classification will be required that will replace OPCS-4, building on its best features and complementing the richness of ICD-11. This will provide better integration with and increased derivation from the SNOMED CT data captured at the point of care.
The World Health Organization is developing the International Classification of Health Interventions, (ICHI), which may provide a suitable replacement, subject to content development to support UK use cases.
The indicative roadmap for the replacement of OPCS-4 is shown in the Indicative roadmap for ICD-11 and the new PBC standard (above). Key points from the roadmap are:
- The analysis of options to replace OPCS-4 will begin in the next 12 months. The results of the options assessment will be socialised with NHS users, suppliers, professional groups and ALBs.
- The replacement of OPCS-4 codes in remuneration mechanisms will be a long-term activity due to long tariff setting cycles and the need to parallel run any changes to the HRG derivation methods.
- There will be a need to fully test SNOMED CT data for the CDS use case to ensure integrity of data.
- The migration of OPCS-4 uses to SNOMED CT can be de-coupled from the OPCS-4 replacement, by transitioning from using OPCS-4 in CDS to using SNOMED CT information collected at the point of care. We would therefore update the CDS returns to be able to flow SNOMED CT level data, which will require a change to the SCCI standard.
- The SNOMED CT to OPCS-4 derivation will be taking place predominantly at the national level, however we expect that we will need to work collaboratively with all consumers of OPCS-4 to design their migration path to SNOMED CT.
- We will explore the use of Research and Public Health flows that may contain OPCS-4 information to assess the optimal migration path to SNOMED CT, and at the same time explore any requirements for augmenting any of the existing flows to offer increased granularity that SNOMED CT would allow.
The dm+d standard arose in 1998 when the Government White Paper ‘Information for Health’ noted:
“There is a lack of standardisation in the UK in describing medicines, appliances and medical devices, in how such descriptions are organised, and in linking knowledge required for decision support to these descriptions”
dm+d is a clinical information standard in current use. dm+d is the recognised NHS Standard (SCCI0052) for uniquely identifying medicines and medical devices used in direct patient care.
The UK Clinical Products Reference Source (UKCPRS) programme was initiated to deliver a terminology to address this lack of standardisation.
The output of the UKCPRS project, the NHS dictionary of medicines and devices (dm+d), is designed to bring significant business benefits over existing terminologies, such as:
- reduced ambiguity in providing for selection of product terms by clinicians
- transfer of coded data between systems without translating data
- reduced cost of maintenance inherent with multiple code sets
dm+d is continually updated with new products to market, discontinuations, changes etc and the new data are released on a weekly basis. The underlying model for dm+d has been stable for many years and is represented below. All products entered into dm+d fit into this model (referred to as the ‘5-box model’).
Figure 3: The dm+d data model
The five classes in the dm+d data model have different use cases.
|Virtual Therapeutic Moiety||
Dose based prescribing
Recording of partial medication information
|Virtual Medicinal Product||
Product (generic) based prescribing
Product identification and selection for dispensing or administration
Recording of information within patient records
|Virtual Medicinal Product Pack||
Identification or selection of pack size for dispensing
To record dispensed items including pack size
Provides information for electronic reimbursement and secondary care tendering
|Actual Medicinal Product||
Product (brand or manufacturer) based prescribing
Product identification and selection for dispensing/administration
Recording of information within patient records
|Actual Medicinal Product Pack||
Identification of pack size and availability
Recording dispensed items including pack size and manufacturer or brand
Links to supply chain (for example, GTIN barcode mapping)
Due to its use in the Electronic Prescription Service (dm+d codes are the codes used in all electronic prescription messages) dm+d is widespread in primary care. The digital medicines programme (that the CCIO has included under his portfolio) work with the Secondary Care NHS organisations to increase the uptake of dm+d. The objective of the CCIO portfolio is to deliver on medicines interoperability. The ability to move medicines information safely and effectively within and across electronic systems without the need to rekey is fully dependent on a standardised way of coding medicines. It is therefore reliant on dm+d. Conversely, the drive to deliver on medicines interoperability will be vital to the uptake of dm+d in secondary care (for example, interoperable stock control and ePrescribing and Medicines Administration Systems; the ability to share patient medication information with other care settings such as General Practice, pharmacy and the ability to read and interpret a patient’s current medication to facilitate medication reconciliation on admission).
The roadmap below lists the key activities to enhance the value of dm+d, focusing on identifying secondary care needs for the standard, developing FHIR profiles and migration paths, and enhancing existing interoperability of the product.
Indicative dm+d roadmap
Phase 1: sharing and linking
- Engage with CCIO and secondary care NHS organisations to determine need for dm+d content
- Assess options to fully populate the dm+d and GTN mapping
- FHIR profiles for dm+d use cases developed
- Implement recommended option to fully populate dm+d and GTN mapping
- Publish the updated dm+d GTN mappings
- Scope linking IDMP to dm+d
- Develop the IDMP/dm+d link
- Pilot the IDMP/dm+d link
- dm+d is published with IDMP link active
Key migration components
Tracking standards – To fully deliver on medicines interoperability and negate the need to re-key in order to move information round the clinical system, there needs to be a link between the input (barcode/RFID for example) and the clinical codes (dm+d). A full map between dm+d and the GS1 GTIN (via the Global Trade Item Number (GTIN)) is required to support this and this is an initiative that we will be delivering during Phase one.
IDMP interaction – a European wide initiative to standardise on regulatory (licensing) drug coding – Identification of Medicinal Product (IDMP) standards (ISO 11238, 11239, 11240 11615, 11616) are being implemented by the European Medicines Agency in their regulatory database. Once implemented, IDMP is being looked at to provide a common language for medicines across Europe for wider use cases too. It is also being implemented by the FDA in the United States and therefore offers an opportunity for trans-Atlantic links.
To link dm+d into IDMP would mean to leverage pan-European and US connectivity. We will work with the relevant EU projects to ensure alignment of dm+d with the international connectivity standards and deliver dm+d and IDMP integration.
FHIR – as a terminology standard dm+d derives some of its greatest benefits from supporting interoperability. FHIR is a key messaging standard. Only by ensuring that dm+d is correctly messaged by FHIR will the interoperability benefits of dm+d be fully recognised. The migration to FHIR will need to carefully consider legacy systems that use HL7v3 messaging, as it is essential not to sacrifice existing interoperability.
Medicines regulation – wherever possible, dm+d uses categorical information from other national, European or international standards. For example, most medicine names come from an international standard (International Nonproprietary Name (INN)), and most medicine forms from a European standard (European Directorate for the Quality of Medicines and Healthcare (EDQM) Standard Terms). Many of these form parts of the way medicines are regulated. The close alignment with key regulatory frameworks allows dm+d to deliver significant patient safety benefits.