Logical Observation Identifiers Names and Codes for Laboratorians

Why do probabilistic clinical models fail to transport between sites

The rising popularity of artificial intelligence in healthcare is highlighting the problem that a computational model achieving super-human clinical performance at its training sites may perform substantially worse at new sites. In this perspective, we argue that we should typically expect this failure to transport, and we present common sources for it, divided into those under the control of the experimenter and those inherent to the clinical data-generating process. Of the inherent sources we look a little deeper into site-specific clinical practices that can affect the data distribution, and propose a potential solution intended to isolate the imprint of those practices on the data from the patterns of disease cause and effect that are the usual target of probabilistic clinical models.

Laboratory Test Names Matter: A Survey on What Works and What Doesn't Work for Orders and Results

CONTEXT.—

Health care providers were surveyed to determine their ability to correctly decipher laboratory test names and their preferences for laboratory test names and result displays.

OBJECTIVE.—

To confirm principles for laboratory test nomenclature and display and to compare and contrast the abilities and preferences of different provider groups for laboratory test names.

DESIGN.—

Health care providers across different specialties and perspectives completed a survey of 38 questions, which included participant demographics, real-life examples of poorly named laboratory orders that they were asked to decipher, an assessment of vitamin D test name knowledge, their preferences for ideal names for tests, and their preferred display for test results. Participants were grouped and compared by profession, level of training, and the presence or absence of specialization in informatics and/or laboratory medicine.

RESULTS.—

Participants struggled with poorly named tests, especially with less commonly ordered tests. Participants' knowledge of vitamin D analyte names was poor and consistent with prior published studies. The most commonly selected ideal names correlated positively with the percentage of the authors' previously developed naming rules (R = 0.54, P < .001). There was strong consensus across groups for the best result display.

CONCLUSIONS.—

Poorly named laboratory tests are a significant source of provider confusion, and tests that are named according to the authors' naming rules as outlined in this article have the potential to improve test ordering and correct interpretation of results. Consensus among provider groups indicates that a single yet clear naming strategy for laboratory tests is achievable.

An innovative technological infrastructure for managing SARS-CoV-2 data across different cohorts in compliance with General Data Protection Regulation

Background

The ORCHESTRA project, funded by the European Commission, aims to create a pan-European cohort built on existing and new large-scale population cohorts to help rapidly advance the knowledge related to the prevention of the SARS-CoV-2 infection and the management of COVID-19 and its long-term sequelae. The integration and analysis of the very heterogeneous health data pose the challenge of building an innovative technological infrastructure as the foundation of a dedicated framework for data management that should address the regulatory requirements such as the General Data Protection Regulation (GDPR).

Methods

The three participating Supercomputing European Centres (CINECA - Italy, CINES - France and HLRS - Germany) designed and deployed a dedicated infrastructure to fulfil the functional requirements for data management to ensure sensitive biomedical data confidentiality/privacy, integrity, and security. Besides the technological issues, many methodological aspects have been considered: Berlin Institute of Health (BIH), Charité provided its expertise both for data protection, information security, and data harmonisation/standardisation.

Results

The resulting infrastructure is based on a multi-layer approach that integrates several security measures to ensure data protection. A centralised Data Collection Platform has been established in the Italian National Hub while, for the use cases in which data sharing is not possible due to privacy restrictions, a distributed approach for Federated Analysis has been considered. A Data Portal is available as a centralised point of access for non-sensitive data and results, according to findability, accessibility, interoperability, and reusability (FAIR) data principles. This technological infrastructure has been used to support significative data exchange between population cohorts and to publish important scientific results related to SARS-CoV-2.

Conclusions

Considering the increasing demand for data usage in accordance with the requirements of the GDPR regulations, the experience gained in the project and the infrastructure released for the ORCHESTRA project can act as a model to manage future public health threats. Other projects could benefit from the results achieved by ORCHESTRA by building upon the available standardisation of variables, design of the architecture, and process used for GDPR compliance.

Improving Diagnostics with Deep Forest Applied to Electronic Health Records

An electronic health record (EHR) is a vital high-dimensional part of medical concepts. Discovering implicit correlations in the information of this data set and the research and informative aspects can improve the treatment and management process. The challenge of concern is the data sources' limitations in finding a stable model to relate medical concepts and use these existing connections. This paper presents Patient Forest, a novel end-to-end approach for learning patient representations from tree-structured data for readmission and mortality prediction tasks. By leveraging statistical features, the proposed model is able to provide an accurate and reliable classifier for predicting readmission and mortality. Experiments on MIMIC-III and eICU datasets demonstrate Patient Forest outperforms existing machine learning models, especially when the training data are limited. Additionally, a qualitative evaluation of Patient Forest is conducted by visualising the learnt representations in 2D space using the t-SNE, which further confirms the effectiveness of the proposed model in learning EHR representations.

Diabetes mellitus monitoring and control among adults in Australian general practice: a national retrospective cohort study

OBJECTIVES

This study investigated whether the monitoring and control of clinical parameters are better among patients with newly compared with past recorded diabetes diagnosis.

DESIGN

Retrospective cohort study.

SETTING

MedicineInsight, a national general practice database in Australia.

PARTICIPANTS

101 875 'regular' adults aged 18+ years with past recorded (2015-2016) and 9236 with newly recorded (2017) diabetes diagnosis.

MAIN OUTCOME MEASURES

Two different groups of outcomes were assessed in 2018. The first group of outcomes was the proportion of patients with clinical parameters (ie, glycated haemoglobin A1c (HbA1c), blood pressure (BP), total cholesterol, low-density lipoprotein cholesterol (LDL-C), high-density lipoprotein cholesterol, triglycerides, estimated glomerular filtration rate and albumin-to-creatinine ratio) monitored at least once in 2018. The second group of outcomes were those related to diabetes control in 2018 (HbA1c ≤7.0%, (BP) ≤140/90 mm Hg, total cholesterol <4.0 mmol/L and LDL-C <2.0 mmol/L). Adjusted ORs (OR_adj) and adjusted probabilities (%) were obtained based on logistic regression models adjusted for practice variables and patients' socio-demographic and clinical characteristics.

RESULTS

The study included 111 111 patients (51.7% men; mean age 65.3±15.0 years) with recorded diabetes diagnosis (11.0% of all 1 007 714 adults in the database). HbA1c was monitored in 39.2% (95% CI 36.9% to 41.6%) of patients with newly recorded and 45.2% (95% CI 42.6% to 47.8%) with past recorded diabetes (OR_adj 0.78, 95% CI 0.73 to 0.82). HbA1c control was achieved by 78.4% (95% CI 76.7% to 80.0%) and 54.4% (95% CI 53.4% to 55.4%) of monitored patients with newly or past recorded diabetes, respectively (OR_adj 3.11, 95% CI 2.82 to 3.39). Less than 20% of patients with newly or past recorded diabetes had their HbA1c, BP and total cholesterol levels controlled (OR_adj 1.08, 95% CI 0.97 to 1.21).

CONCLUSIONS

The monitoring of clinical parameters was lower among patients with newly than past recorded diabetes. However, diabetes control was similarly low in both groups, with only one in five monitored patients achieving control of all clinical parameters.

Electronic Health Record Optimization for Artificial Intelligence

Laboratory clinical decision support (CDS) typically relies on data from the electronic health record (EHR). The implementation of a sustainable, effective laboratory CDS program requires a commitment to standardization and harmonization of key EHR data elements that are the foundation of laboratory CDS. The direct use of artificial intelligence algorithms in CDS programs will be limited unless key elements of the EHR are structured. The identification, curation, maintenance, and preprocessing steps necessary to implement robust laboratory-based algorithms must account for the heterogeneity of data present in a typical EHR.

OUP accepted manuscript

Objective

Assess the effectiveness of providing Logical Observation Identifiers Names and Codes (LOINC®)-to-In Vitro Diagnostic (LIVD) coding specification, required by the United States Department of Health and Human Services for SARS-CoV-2 reporting, in medical center laboratories and utilize findings to inform future United States Food and Drug Administration policy on the use of real-world evidence in regulatory decisions.

Materials and Methods

We compared gaps and similarities between diagnostic test manufacturers’ recommended LOINC® codes and the LOINC® codes used in medical center laboratories for the same tests.

Results

Five medical centers and three test manufacturers extracted data from laboratory information systems (LIS) for prioritized tests of interest. The data submission ranged from 74 to 532 LOINC® codes per site. Three test manufacturers submitted 15 LIVD catalogs representing 26 distinct devices, 6956 tests, and 686 LOINC® codes. We identified mismatches in how medical centers use LOINC® to encode laboratory tests compared to how test manufacturers encode the same laboratory tests. Of 331 tests available in the LIVD files, 136 (41%) were represented by a mismatched LOINC® code by the medical centers (chi-square 45.0, 4 df, P < .0001).

Discussion

The five medical centers and three test manufacturers vary in how they organize, categorize, and store LIS catalog information. This variation impacts data quality and interoperability.

Conclusion

The results of the study indicate that providing the LIVD mappings was not sufficient to support laboratory data interoperability. National implementation of LIVD and further efforts to promote laboratory interoperability will require a more comprehensive effort and continuing evaluation and quality control.

Telehealth-based diagnostic testing in general practice during the COVID-19 pandemic: an observational study

Background

Since the World Health Organization declared COVID-19 a pandemic on 11 March 2020, health technologies have been rapidly scaled up to ensure access to care. A significant innovation has been telehealth in general practice. Now widespread, it remains unknown how this shift to virtual care has impacted on quality-of-care indicators such as pathology testing and diagnosis.

Aim

To undertake a comparison of telehealth and face-to-face general practice consultations to: identify if there were differences in the proportion of pathology test referrals from 2019–2020; and quantify any change in pathology test collection and follow-up patterns.

Design & setting

Retrospective observational study of routinely collected electronic patient data from 807 general practices across New South Wales (NSW) and Victoria, Australia.

Method

Multivariate generalised estimating equation models were used to estimate the proportion of pathology test referrals for overall, face-to-face, and telehealth consultations. Pathology test follow-up was described through median (and interquartile range [IQR]) time.

Results

Pathology test referrals declined during periods of high COVID-19 cases, falling from 10.8% in February 2020 to a low of 4.5% during the first peak in April. Overall, pathology test referrals were lower for telehealth than face-to-face consultations. Median time between referral and test collection was 3 days (IQR 1–14) for telehealth and 1 day (IQR 0–7) for face to face.

Conclusion

For telehealth to become part of routine care, it is crucial that gaps in functionality, including difficulty in test referral processes, be addressed. Quality improvements supporting care practices will ensure clinicians’ workflows are supported and patients receive diagnostic testing.

Electronic Health Records and Genomics: Perspectives from the Association for Molecular Pathology Electronic Health Record (EHR) Interoperability for Clinical Genomics Data Working Group

The use of genomics in medicine is expanding rapidly, but information systems are lagging in their ability to support genomic workflows both from the laboratory and patient-facing provider perspective. The complexity of genomic data, the lack of needed data standards, and lack of genomic fluency and functionality as well as several other factors have contributed to the gaps between genomic data generation, interoperability, and utilization. These gaps are posing significant challenges to laboratory and pathology professionals, clinicians, and patients in the ability to generate, communicate, consume, and use genomic test results. The Association for Molecular Pathology Electronic Health Record Working Group was convened to assess the challenges and opportunities and to recommend solutions on ways to resolve current problems associated with the display and use of genomic data in electronic health records.

Logical Observation Identifiers Names and Codes (LOINC®) Applied to Microbiology: A National Laboratory Mapping Experience in Taiwan

Background and Objective: Logical Observation Identifiers Names and Codes (LOINC) is a universal standard for identifying laboratory tests and clinical observations. It facilitates a smooth information exchange between hospitals, locally and internationally. Although it offers immense benefits for patient care, LOINC coding is complex, resource-intensive, and requires substantial domain expertise. Our objective was to provide training and evaluate the performance of LOINC mapping of 20 pathogens from 53 hospitals participating in the National Notifiable Disease Surveillance System (NNDSS). Methods: Complete mapping codes for 20 pathogens (nine bacteria and 11 viruses) were requested from all participating hospitals to review between January 2014 and December 2016. Participating hospitals mapped those pathogens to LOINC terminology, utilizing the Regenstrief LOINC mapping assistant (RELMA) and reported to the NNDSS, beginning in January 2014. The mapping problems were identified by expert panels that classified frequently asked questionnaires (FAQs) into seven LOINC categories. Finally, proper and meaningful suggestions were provided based on the error pattern in the FAQs. A general meeting was organized if the error pattern proved to be difficult to resolve. If the experts did not conclude the local issue’s error pattern, a request was sent to the LOINC committee for resolution. Results: A total of 53 hospitals participated in our study. Of these, 26 (49.05%) used homegrown and 27 (50.95%) used outsourced LOINC mapping. Hospitals who participated in 2015 had a greater improvement in LOINC mapping than those of 2016 (26.5% vs. 3.9%). Most FAQs were related to notification principles (47%), LOINC system (42%), and LOINC property (26%) in 2014, 2015, and 2016, respectively. Conclusions: The findings of our study show that multiple stage approaches improved LOINC mapping by up to 26.5%.

Influence of serum iron test results on the diagnosis of iron deficiency in children: a retrospective observational study

BACKGROUND AND OBJECTIVE

Serum iron results are not indicative of iron deficiency yet may be incorrectly used to diagnose iron deficiency instead of serum ferritin results. Our objective was to determine the association between serum iron test results and iron-deficiency diagnosis in children by general practitioners.

DESIGN, SETTING, PATIENTS AND MAIN OUTCOME MEASURES

A retrospective observational study of 14 187 children aged 1-18 years with serum ferritin and serum iron test results from 137 general practices in Victoria, Australia, between 2008 and 2018. Generalised estimating equation models calculating ORs were used to determine the association between serum iron test results (main exposure measure) and iron-deficiency diagnosis (outcome measure) in the following two population groups: (1) iron-deplete population, defined as having a serum ferritin <12 µg/L if aged <5 years and <15 µg/L if aged ≥5 years and (2) iron-replete population, defined as having a serum ferritin >30 µg/L.

RESULTS

3484 tests were iron deplete and 15 528 were iron replete. Iron-deplete children were less likely to be diagnosed with iron deficiency if they had normal serum iron levels (adjusted OR (AOR): 0.73; 95% CI 0.57 to 0.96). Iron-replete children had greater odds of an iron-deficiency diagnosis if they had low serum iron results (AOR: 2.59; 95% CI 1.72 to 3.89). Other contributors to an iron-deficiency diagnosis were female sex and having anaemia.

CONCLUSION

Serum ferritin alone remains the best means of diagnosing iron deficiency. Reliance on serum iron test results by general practitioners is leading to significant overdiagnosis and underdiagnosis of iron deficiency in children.

Data standardization implementation and applications within and among diagnostic laboratories: integrating and monitoring enteric coronaviruses

Every day, thousands of samples from diverse populations of animals are submitted to veterinary diagnostic laboratories (VDLs) for testing. Each VDL has its own laboratory information management system (LIMS), with processes and procedures to capture submission information, perform laboratory tests, define the boundaries of test results (i.e., positive or negative), and report results, in addition to internal business and accounting applications. Enormous quantities of data are accumulated and stored within VDL LIMSs. There is a need for platforms that allow VDLs to exchange and share portions of laboratory data using standardized, reliable, and sustainable information technology processes. Here we report concepts and applications for standardization and aggregation of data from swine submissions to multiple VDLs to detect and monitor porcine enteric coronaviruses by RT-PCR. Oral fluids, feces, and fecal swabs were the specimens submitted most frequently for enteric coronavirus testing. Statistical algorithms were used successfully to scan and monitor the overall and state-specific percentage of positive submissions. Major findings revealed a consistently recurrent seasonal pattern, with the highest percentage of positive submissions detected during December-February for porcine epidemic diarrhea virus, porcine deltacoronavirus, and transmissible gastroenteritis virus (TGEV). After 2014, very few submissions tested positive for TGEV. Monitoring VDL data proactively has the potential to signal and alert stakeholders early of significant changes from expected detection. We demonstrate the importance of, and applications for, data organized and aggregated by using LOINC and SNOMED CTs, as well as the use of customized messaging to allow inter-VDL exchange of information.

'There and Back Again'-Forward Genetics and Reverse Phenotyping in Pulmonary Arterial Hypertension

Although the invention of right heart catheterisation in the 1950s enabled accurate clinical diagnosis of pulmonary arterial hypertension (PAH), it was not until 2000 when the landmark discovery of the causative role of bone morphogenetic protein receptor type II (BMPR2) mutations shed new light on the pathogenesis of PAH. Since then several genes have been discovered, which now account for around 25% of cases with the clinical diagnosis of idiopathic PAH. Despite the ongoing efforts, in the majority of patients the cause of the disease remains elusive, a phenomenon often referred to as "missing heritability". In this review, we discuss research approaches to uncover the genetic architecture of PAH starting with forward phenotyping, which in a research setting should focus on stable intermediate phenotypes, forward and reverse genetics, and finally reverse phenotyping. We then discuss potential sources of "missing heritability" and how functional genomics and multi-omics methods are employed to tackle this problem.

Ethical Considerations on Pediatric Genetic Testing Results in Electronic Health Records

BACKGROUND

Advances in technology and access to expanded genetic testing have resulted in more children and adolescents receiving genetic testing for diagnostic and prognostic purposes. With increased adoption of the electronic health record (EHR), genetic testing is increasingly resulted in the EHR. However, this leads to challenges in both storage and disclosure of genetic results, particularly when parental results are combined with child genetic results.

PRIVACY AND ETHICAL CONSIDERATIONS

Accidental disclosure and erroneous documentation of genetic results can occur due to the nature of their presentation in the EHR and documentation processes by clinicians. Genetic information is both sensitive and identifying, and requires a considered approach to both timing and extent of disclosure to families and access to clinicians.

METHODS

This article uses an interdisciplinary approach to explore ethical issues surrounding privacy, confidentiality of genetic data, and access to genetic results by health care providers and family members, and provides suggestions in a stakeholder format for best practices on this topic for clinicians and informaticians. Suggestions are made for clinicians on documenting and accessing genetic information in the EHR, and on collaborating with genetics specialists and disclosure of genetic results to families. Additional considerations for families including ethics around results of adolescents and special scenarios for blended families and foster minors are also provided. Finally, administrators and informaticians are provided best practices on both institutional processes and EHR architecture, including security and access control, with emphasis on the minimum necessary paradigm and parent/patient engagement and control of the use and disclosure of data.

CONCLUSION

The authors hope that these best practices energize specialty societies to craft practice guidelines on genetic information management in the EHR with interdisciplinary input that addresses all stakeholder needs.

Challenges involved in establishing a web-based clinical decision support tool in community health centers

Implementation lessons: Establishing a shared 'hub-and-spoke,' web-based clinical decision support system (CDSS) in an EHR shared by >600 community health centers incurred a myriad of challenges, which are summarized here to guide others seeking to use similar CDSS. Legal and compliance challenges involved ensuring secure data exchanges, determining which entity maintains data records, and deciding which data are sent to the CDSS. Technical challenges involved using lab data from multiple sources and improving the CDSS' cache routine performance in its new setting. Clinical implementation challenges involved identifying optimal strategies for generating data on CDSS use rates, modifying the CDSS functionality for obtaining clinician/staff feedback, and customizing the risk thresholds that trigger the CDSS for the new setting.

Clinical thresholds for diagnosing iron deficiency: comparison of functional assessment of serum ferritin to population based centiles

Low serum ferritin is diagnostic of iron deficiency, yet its published lower cut-off values are highly variable, particularly for pediatric populations. Lower cut-off values are commonly reported as 2.5th percentiles, and is based on the variation of ferritin values in the population. Our objective was to determine whether a functional approach based on iron deficient erythropoiesis could provide a better alternative. Utilizing 64,443 ferritin test results from pediatric electronic health records, we conducted various statistical techniques to derive 2.5th percentiles, and also derived functional reference limits through the association between ferritin and erythrocyte parameters: hemoglobin, mean corpuscular volume, mean cell hemoglobin concentration, and red cell distribution width. We find that lower limits of reference intervals derived as centiles are too low for clinical interpretation. Functional limits indicate iron deficiency anemia starts to occur when ferritin levels reach 10 µg/L, and are largely similar between genders and age groups. In comparison, centiles (2.5%) presented with lower limits overall, with varying levels depending on age and gender. Functionally-derived limits better reflects the underlying physiology of a patient, and may provide a basis for deriving a threshold related to treatment of iron deficiency and any other biomarker with functional outcomes.

Experience in Developing an FHIR Medical Data Management Platform to Provide Clinical Decision Support

This paper is an extension of work originally presented to pHealth 2019—16th International Conference on Wearable, Micro and Nano Technologies for Personalized Health. To provide an efficient decision support, it is necessary to integrate clinical decision support systems (CDSSs) in information systems routinely operated by healthcare professionals, such as hospital information systems (HISs), or by patients deploying their personal health records (PHR). CDSSs should be able to use the semantics and the clinical context of the data imported from other systems and data repositories. A CDSS platform was developed as a set of separate microservices. In this context, we implemented the core components of a CDSS platform, namely its communication services and logical inference components. A fast healthcare interoperability resources (FHIR)-based CDSS platform addresses the ease of access to clinical decision support services by providing standard-based interfaces and workflows. This type of CDSS may be able to improve the quality of care for doctors who are using HIS without CDSS features. The HL7 FHIR interoperability standards provide a platform usable by all HISs that are FHIR enabled. The platform has been implemented and is now productive, with a rule-based engine processing around 50,000 transactions a day with more than 400 decision support models and a Bayes Engine processing around 2000 transactions a day with 128 Bayesian diagnostics models.

A Survey of LOINC Code Selection Practices Among Participants of the College of American Pathologists Coagulation (CGL) and Cardiac Markers (CRT) Proficiency Testing Programs

Context.—

Biomedical terminologies such as Logical Observation Identifiers, Names, and Codes (LOINC) were developed to enable interoperability of health care data between disparate health information systems to improve patient outcomes, public health, and research activities.

Objective.—

To ascertain the utilization rate and accuracy of LOINC terminology mapping to 10 commonly ordered tests by participants of the College of American Pathologists (CAP) Proficiency Testing program.

Design.—

Questionnaires were sent to 1916 US and Canadian laboratories participating in the 2018 CAP coagulation (CGL) and/or cardiac markers (CRT) surveys requesting information on practice setting, instrument(s) and test method(s), and LOINC code selection and usage in the laboratory and electronic health records.

Results.—

Ninety of 1916 CGL and/or CRT participants (4.7%) responded to the questionnaire. Of the 275 LOINC codes reported, 54 (19.6%) were incorrect: 2 codes (5934-2 and 12345-1) (0.7%) did not exist in the LOINC database and the highest error rates were observed in the property (27 of 275, 9.8%), system (27 of 275, 9.8%), and component (22 of 275, 8.0%) LOINC axes. Errors in LOINC code selection included selection of the incorrect component (eg, activated clotting time instead of activated partial thromboplastin time); selection of panels that can never be used to obtain an individual analyte (eg, prothrombin time panel instead of international normalized ratio); and selection of an incorrect specimen type.

Conclusions.—

These findings of real-world LOINC code implementation across a spectrum of laboratory settings should raise concern about the reliability and utility of using LOINC for clinical research or to aggregate data.