Thanks to everyone who helped make the 2024 LOINC Conference a big success

In this presentation we would like to discuss the fact that LOINC realizes unity of language, but LOINC does not imply unity of test results (comparability). How can we make sure that laboratory test results are safely exchanged and interpreted? The method within LOINC can only partly cover this problem. In the Netherlands, we are currently discussing which additional information (context) is necessary to realize this safe exchange (e.g. more detail information about the specific laboratory test) and how we can realize a general and scalable solution to see if test results with a similar LOINC can or cannot be combined when they are produced in different laboratories. We plan to use data from EQA programs to identify exchangeable methods based on similar metrological traceability and similar selectivity.

This session will describe the use of pre and post coordination of LOINC codes to represent data in HL7 FHIR messages and other data exchange standards. Pre and post coordination is useful for laboratory data, clinical data, and document names. It allows greater specificity in coding with the use of fewer LOINC codes. The session will include examples of the use of pre and post coordination and also discuss what it means for how LOINC codes will be created.

The COVID-19 pandemic has made evident how our scientific knowledge depends on the possibility to analyse large amounts of data. Combining data from different research groups and from different countries is a process that usually requires time and resources due to the lack of standardization of variables across the different study groups. The European project ORCHESTRA has contributed to the standardization of semantics across several informational categories related to COVID-19. We would like to present the work performed with LOINC to standardize questionnaires that analyse social aspects of the pandemic such as mental health in the household, lifestyle during lockdown, unmet healthcare needs, prevention measures and trust in science. These questionnaires, developed within ORCHESTRA, are now available for worldwide use and identified with international LOINC codes. These questionnaires can support research to understand the social effects of the pandemic but also to increase preparedness for future health threats.

The PACIO (Post-Acute Care Interoperability) Project, established February 2019, is a collaborative effort between industry, government, and other stakeholders, with the goal of establishing a framework for the development of FHIR® (Fast Health Interoperability Resource) implementation guides (IGs) to facilitate health information exchange. This presentation will provide an overview of the new PACIO Project Transitions of Care (TOC) use case and FHIR® IG.

The TOC IG intentionally repurposes and reuses data elements and concepts collected in the standardized CMS Post-Acute Care (PAC) assessments to convey clinical information in a TOC composition including orders, diagnoses, conditions, procedures, observations, and medications. Documentation of these concepts is not consistent across settings and not discretely transmitted in transitions of care. The purpose of the TOC IG is to generate a standardized composition that will follow a patient who is leaving a PAC setting and moving to a new setting, which could be another PAC setting, Acute-Care, or Home and Community Based Services and Primary Care. The composition will inform a safe and efficient start of care in that new setting.

Role-specific clinician groups (e.g., MD/NP/PAs, RNs, dieticians, therapists) chose the concepts from the CMS Data Element Library (DEL) using a consensus-based process to determine whether a code could be leveraged for additional uses, such as functioning, cognition, or discipline-specific measures. Data elements in the DEL are encoded using the LOINC ontology. The group also included concepts they deemed essential that are not in the DEL using standardized terminology.

The value of establishing semantic interoperability, (i.e. the need for information to mean the same thing between sending clinician and receiving clinician any their IT systems) using LOINC played a large part in constraining the 2000+ DEL items into a meaningful composition.

In its third instalment since 2022, the present study (1) rounds up the LOINC code mapping algorithm previously presented at the 2022 and 2023 LOINC conferences and (2) addresses subsequent challenges in interoperability for clinical information shared via structured clinical notes.

We established in our previous study that our efforts in implementing LOINC (Logical Observation Identifiers Names and Codes) to map French laboratory tests at the national level (over 100 laboratories were used) have been essential in achieving interoperability of patient-related data and compatibility between health data management systems. As the French government has adopted the HL7 CDA R2 standard to share patient-related clinical data, this naturally points to the next step in our study: Evaluating the impact of using LOINC codes to describe clinical data using this standard. Addressing this issue reveals that LOINC code conversion is necessary, but in no way sufficient in achieving interoperability.

A critical analysis of document exchange using structured information according to the CDA R2 standard reveals that it can be prone to erroneous interpretations of laboratory test results if the previously described clinical data is simply replaced by its corresponding LOINC code. It is, in fact, equally necessary to consider other features such as:

• The Unified Code for Units of Measure (UCUM)
• The technique when it affects the clinical interpretation of test results
• The clinical analyser used and reference values.

In the course of our presentation we shall describe the factors that contribute to these erroneous interpretations and provide solutions to alleviate the problem. We conclude with lessons learnt and suggestions for guaranteeing perfect interoperability of laboratory test results within the HL7 CDA R2 standard.

We are presenting here an approach to harmonize genetic and genomic testing with LOINC and other clinical standards, using a Terminology Server.

On January 23rd, the Spanish Ministry of Health presented the new Common Catalog of Genetic and Genomic Tests of the National Health System (NHS).

The catalog has started with the first prioritized areas: Adult Oncohematology, Pediatric Oncohematology, Pharmacogenomics, Cardiovascular diseases, disorders of the circulatory system, Ophthalmological diseases, Hereditary and mitochondrial metabolic diseases, Neurological and neuromuscular diseases, and Neurodevelopmental disorders, including neurocognitive deficits. Currently, the catalog consists of 672 tests.

For each disease, different information is provided: ICD-10-CM and ORPHA codes, indication, type of study, clinical utility, specimen, type of alteration detected, method of analysis, genes or regions studied, and associated treatment. International frameworks such as FHIR or GA4GH recommend LOINC for laboratory observations, and other specific standards must be considered to support genomics, such as SNOMED CT, OMIM, HPO, or ORPHA to normalize phenotypes, or HGNC and HGVS to normalize genes and variants.

Our objective is to demonstrate how LOINC can fit into this structure along with other genetic standards, allowing interoperability with other public, regional, or private Spanish catalogs that must be aligned.

Terminology server (TS) tools are needed to manage all this content and to keep all these terminologies updated and synchronized with the Standards Development Organizations. In addition, TS should include automatic coding functionalities to facilitate the mapping of local terms to the target terminologies.

Anil Patel from Canada Health Infoway will provide an overview of interoperability in the Canadian provinces and territories. This will segue into a full description of the LOINC Document Ontology and the benefits of its adoption in Canada including case studies.

TriNetX is a global health research network that connects the world of drug discovery and development from pharmaceutical company to study site and investigator to patient by sharing real-world data. In this network, there are more than 170 health care organizations located in 19 countries in North and South America, EMEA and Asia-Pacific. Laboratory data is standardized to LOINC and subsequently validated by TriNetX using in-house resources and in collaboration with the Universidad Politécnica de Madrid (UPM).

In it is well-known that the quality of laboratory mappings can be compromised by missing test details when performing the semantic standardization of local tests to LOINC. Quality problems have been already reported in literature.

In the presentation, we aim to quantify and classify the mis-mappings found in healthcare data provided for secondary uses by different health providers. We additionally want to propose a framework of quality measures to check the correctness of a laboratory dataset by using laboratory test result metadata (type of result, position in catalogue, etc). Lastly, we would like to provide a list of recommendations improve the quality when performing the semantic standardization.

With the growth of FHIR usage globally, and included in US EHR certification functionality, laboratory data and LOINC are key aspects of data exchanged with this standard. We’ll discuss which FHIR resources are used for exchange of laboratory data, their maturity, and challenges using real world examples. Cautions and considerations for users will be shared with a focus on the quality of exchanges and making interoperability easier.

This guide represents a significant milestone in the field of clinical laboratories in Chile, addressing the urgent need to establish quality criteria and consensus practices in the operation of Clinical Laboratory Information Systems (LIS). The creation of this guide arises in response to the lack of documents and agreements that would guide on the standards and best practices in LIS in the national context, despite the high level of technologization in Chilean clinical laboratories.

The main objective of this initiative is to provide laboratory users with the necessary information to select, evaluate, implement, maintain, and continuously improve LIS in their workplaces. Directed at various stakeholders in the clinical laboratory ecosystem, including professionals responsible for laboratory management, personnel in charge of LIS, providers of LIS and associated hardware, IT personnel responsible for LIS, and network administrators, this guide is based on international standards and best practices, adapted to national needs and realities. The methodology used for its development included a comparative diagnosis of international standards and best practices in LIS, the proposal of an initial document, validation through public consultation with stakeholders in the ecosystem, and review of feedback to generate a consensus version. The guide addresses fundamental aspects such as generalities of laboratory information systems, recommendations for proper management, quality control standards, the importance of data, interoperability, and the use of semantic and syntactic standards for a common language.

To date, the guide has had a notable impact, with hundreds of downloads of the document, highlighting its relevance and usefulness to the clinical laboratory community in Chile.

The Institute of Informatics and Telematics of the National Research Council hosts the working group LOINC Italia, which is the official LOINC reference in Italy. It carries out education and training activities on LOINC and validates mappings from local test catalogues to LOINC.

This proposal aims to present the mapping experience of 5 hospitals’ laboratories in Friuli-Venezia Giulia Region. As they don’t have a unique regional test catalogue, they decided to proceed autonomously in the mapping process, and then try to standardize the laboratory services provided at a regional level through the LOINC codes.

The total amount of laboratory tests mapped to LOINC is over 7,000, covering various specialties. The mapping validation involved a dual review by clinical domain experts and LOINC Italia.

We would like to share pros and cons of this widespread mapping experience, focusing on peculiarities related to both the Italian context and some laboratory specialties.

Currently, clinical laboratories, healthcare providers, in vitro diagnostic (IVD) device manufacturers, and Laboratory Information Systems (LIS) have created various coding and data models to meet their own business needs. These variations in laboratory data from multiple sources along with gaps in standardized laboratory data capture, including IVD device data, can impact patient safety, affect interoperability, and generate misleading insights when aggregating and analyzing laboratory data at the population level. We aim to demonstrate a laboratory and IVD terminology ecosystem to create more reliable laboratory test data and to provide a more granular knowledge base for IVD tests, as well as efficient reporting systems for patient safety.

The Laboratory Interoperability Device Reference (LIDR) is a project under development that strives to be a resource to inform end-users how to populate their data dictionaries for specific laboratory tests on specific devices and provide additional information on laboratory tests and test results. This project curates LIDR data elements for modeling IVD devices and creates terminology attributes to represent the unique components of IVD devices and their characteristics utilizing LOINC® to enable consistent data representation. We will demonstrate inferred hierarchical relationships for LOINC® terms with IVD properties by linking LOINC® laboratory terms with the SNOMED CT® Observable Entity hierarchy and description logic definitions. To make this process turn-key for systems, the goal is for LIDR to be distributed in a computable HL7 FHIR format.

The Nursing Knowledge Big Data Science Knowledge Modeling & Encoding work group has been collaborating with LOINC to refine a set of heuristics to guide mapping for standardized assessment scales/surveys (SAS). Nurses use SAS extensively for physical and functional assessments. Scale observations are often poorly defined and are almost always pre-coordinated terms. Basic mapping heuristics are meant to align with FHIR, so normally all nursing observation terms are decomposed to a panel of unique terms--this is not possible with SAS, because the observations cannot be altered in any way. We also try to align with the LOINC policy for re-use of existing components and answers, when they are an exact match. The heuristics have become a bit complex. We will show examples of standardized assessment scales we submitted to LOINC, highlighting the mapping complexities, and we will share the mapping heuristics current state.

What could be EASIER than adopting LOINC having premade template and activities for your team to operate from? The Interoperability through LOINC Project Plan Template was first introduced in 2012 to the LOINC community. (Accessed through https://loinc.org/learn/recommended) The last update of the project management template for adopting LOINC was made in 2018. Now that SHIELD workgroups are in place in the US and LIVD files are starting to be distributed, what better time than now to update the plan template and provide activities to drive home the subtasks needs and decisions?

Power Skills as defined by Project Management Institute are specific behaviors and abilities helping project professionals gain success in their projects by working cohesively with others. Power Skills work in tandem with technical skills in that they enable PMs to bring all stakeholders along on the project journey. Project Management Institute involves all workplace endeavors and careers; not restrictive to IT, informatics or healthcare.

Walk through the template’s project stages of initiation, planning, executing and monitoring the adoption of LOINC in your system. Activities and worksheets build BOTH technical skills and Power Skills (specifically Communication, Problem-solving, Collaborative leadership and Strategic thinking) will be offered in the 2024 version of the Project Plan Template.

This session applies a Project Management Institute Pulse of the Profession survey findings ranking twelve Power Skills owned by a project manager and an updated version of the LOINC website will be made available.

Clinical documents created in the new Department of Veterans Affairs (VA) electronic health record (EHR) system contain a variety of narrative text information for patient care. The standardization of clinical document titles enables VA clinicians to search, filter, sort, or easily retrieve these documents at the point of care in a Joint Longitudinal Viewer (JLV) application system.

METHOD:
We reviewed 879 VA clinical document titles and mapped them to the pertinent LOINC axis (ontology class). After the classes were assigned, we manually mapped the document to the semantically equivalent or closest equivalent code in the LOINC document ontology (DO). A mapped document title was assigned a map category of “equivalent” or “broader”. In this project, broader mappings were allowed to support the use case for searching and filtering clinical document titles in JLV. A set of mapping and translational rules were developed during the process. For example, at least one LOINC axis value was required to assign a broader mapping. Another rule stated that clinical document titles that included the word “assessment” would map to “evaluation” in the LOINC document ontology (DO) Type of Service class.

RESULTS:
Nine mapping rules were developed. A total of 879 VA clinical document titles were reviewed for mapping. Of those, 815 (93%), were mapped to LOINC DO codes. Of the 815, 791 titles were assigned the map category of broader and only 24 (3%) were assigned the map category of equivalent. No appropriate mapping matches were found in the LOINC DO for 64 (7%) of the VA clinical document titles because there was no clear SMD identified.

CONCLUSION:
Mapping rules are required for aligning clinical documents to the LOINC DO. Document titles within the VA clinical systems are much more specific than the LOINC DO. Future LOINC DO enhancements may be necessary to support clinical document interoperability.

Kentucky’s Health Information Exchange (KHIE) has had a transformative impact in disease surveillance and public health activities. KHIE serves as a centralized repository for patient data, and has promoted interoperability, serving as an intermediary for many datatypes and databases used by the Kentucky Department for Public Health (KDPH). KHIE efficiently manages and filters electronic lab reports (ELR) by leveraging standardized vocabularies, such as LOINC codes. KHIE seamlessly identifies LOINC codes that are mapped as reportable or nationally notifiable and forwards the laboratory report to the National Electronic Disease Surveillance System (NEDSS). Within the NEDSS framework, epidemiologists and local health department staff can access timely and actionable laboratory reports. LOINC mappings and workflow decision support allow for automated system processes, such as condition assignment and review of negative laboratory results. This streamlined process not only saves valuable staff time but also facilitates proactive measures for the prevention and control of communicable diseases, ultimately promoting public health outcomes.

The Medical Imaging Data Resource Center (MIDRC) is an NIBIB-sponsored multi-institutional collaborative initiative that aims to make high quality medical imaging and clinical data publicly available to the research community for the purpose of accelerating development of machine learning models in medicine. MIDRC hosts a data commons (data.midrc.org), which is an online data resource featuring a data discovery UI used to select cohorts based on imaging and clinical data elements. It also features an open search API that allows users to send queries and access data programmatically.

Since MIDRC ingests data from many different sites, there is a need to harmonize data elements. Specifically, the DICOM attributes Study Description, Modality, and Body Part Examined are heavily used by researchers to select subsets of imaging studies for training ML models, but these attributes were found to contain a high degree of variability between contributing sites. Therefore, the MIDRC Data Quality and Harmonization subcommittee has sought to use the LOINC standard to harmonize the values of these attributes.

All incoming imaging studies in MIDRC are mapped to LOINC codes, which are then used to derive the LOINC system (body part examined), method (modality), long common name (study description), and contrast indicator. This process has vastly reduced the amount of variability in terminology for similar imaging studies, and thus has greatly enhanced the findability, usability, and interoperability of imaging studies in MIDRC. There are currently over 177,000 imaging studies with nearly 1,700 original, unharmonized study descriptions in MIDRC that have been mapped to 75 LOINC codes and corresponding long common names. which is a 22x reduction in study description variability.

I will present the LOINC mapping process and demonstrate how to use the LOINC attributes to perform programmatic cohort selection in the MIDRC data commons using a Jupyter Notebook demonstration.

At times, standard clinical observations of patient/provider interactions are recorded within the medical field with questionnaires. Premised on provider workflows, electronic medical record interfaces and at times, their overall lack of deployment, their impact to to drive care and glean insights into population health is marginalized. Moreover, when collected, they are oftentimes not represented in standard terminology such as LOINC. The LOINC terminology has a system to represent such answer lists and their terminology collections, making it easier to access, track, and share this data. For instance, when LOINC is used correctly, a PHQ-9 or PHQ-2 questionnaire can track a patient's depression severity across months and years and provide valuable insight to clinicians about medication usage, disease improvement, or progression. Using the LONIC lists can improve interoperability across care settings and continuity of care across the spectrum of health care and improve patient outcomes.

Discern Health uses LOINC answers and questions to support its Fall Prediction Model to help clinicians intervene when senior patients are trending toward catastrophic falls. Falls represent not just a physical stumble, but a profound threat to the health and independence of older adults. According to the CDC, falls are the leading cause of injury for U.S. adults over age 65, with more than one out of four seniors falling each year and over 3 million receiving treatment for fall injuries in emergency departments annually. Each fall treated in the outpatient setting incurs a cost of $9,000-13,000 to treat, while falls that require an inpatient stay double the cost, burdening the system with a $29 billion annual spend. Beyond the statistics, each fall represents a life-altering event for a senior, oftentimes caused by a preventable event and exacerbated health challenges unique to this population.