Nirmatrelvir and Molnupiravir and Post-COVID-19 Condition in Older Patients

This observational cohort study assesses the occurrence of post–COVID-19 condition symptoms in Medicare enrollees prescribed nirmatrelvir and molnupiravir.

Two complementary AI approaches for predicting UMLS semantic group assignment: heuristic reasoning and deep learning

OBJECTIVE

Use heuristic, deep learning (DL), and hybrid AI methods to predict semantic group (SG) assignments for new UMLS Metathesaurus atoms, with target accuracy ≥95%.

MATERIALS AND METHODS

We used train-test datasets from successive 2020AA-2022AB UMLS Metathesaurus releases. Our heuristic "waterfall" approach employed a sequence of 7 different SG prediction methods. Atoms not qualifying for a method were passed on to the next method. The DL approach generated BioWordVec and SapBERT embeddings for atom names, BioWordVec embeddings for source vocabulary names, and BioWordVec embeddings for atom names of the second-to-top nodes of an atom's source hierarchy. We fed a concatenation of the 4 embeddings into a fully connected multilayer neural network with an output layer of 15 nodes (one for each SG). For both approaches, we developed methods to estimate the probability that their predicted SG for an atom would be correct. Based on these estimations, we developed 2 hybrid SG prediction methods combining the strengths of heuristic and DL methods.

RESULTS

The heuristic waterfall approach accurately predicted 94.3% of SGs for 1 563 692 new unseen atoms. The DL accuracy on the same dataset was also 94.3%. The hybrid approaches achieved an average accuracy of 96.5%.

CONCLUSION

Our study demonstrated that AI methods can predict SG assignments for new UMLS atoms with sufficient accuracy to be potentially useful as an intermediate step in the time-consuming task of assigning new atoms to UMLS concepts. We showed that for SG prediction, combining heuristic methods and DL methods can produce better results than either alone.

A practical strategy to use the ICD-11 for morbidity coding in the United States without a clinical modification

OBJECTIVE

The aim of this study was to derive and evaluate a practical strategy of replacing ICD-10-CM codes by ICD-11 for morbidity coding in the United States, without the creation of a Clinical Modification.

MATERIALS AND METHODS

A stepwise strategy is described, using first the ICD-11 stem codes from the Mortality and Morbidity Statistics (MMS) linearization, followed by exposing Foundation entities, then adding postcoordination (with existing codes and adding new stem codes if necessary), with creating new stem codes as the last resort. The strategy was evaluated by recoding 2 samples of ICD-10-CM codes comprised of frequently used codes and all codes from the digestive diseases chapter.

RESULTS

Among the 1725 ICD-10-CM codes examined, the cumulative coverage at the stem code, Foundation, and postcoordination levels are 35.2%, 46.5% and 89.4% respectively. 7.1% of codes require new extension codes and 3.5% require new stem codes. Among the new extension codes, severity scale values and anatomy are the most common categories. 5.5% of codes are not one-to-one matches (1 ICD-10-CM code matched to 1 ICD-11 stem code or Foundation entity) which could be potentially challenging.

CONCLUSION

Existing ICD-11 content can achieve full representation of almost 90% of ICD-10-CM codes, provided that postcoordination can be used and the coding guidelines and hierarchical structures of ICD-10-CM and ICD-11 can be harmonized. The various options examined in this study should be carefully considered before embarking on the traditional approach of a full-fledged ICD-11-CM.

A risk identification model for detection of patients at risk of antidepressant discontinuation

Purpose

Between 30 and 68% of patients prematurely discontinue their antidepressant treatment, posing significant risks to patient safety and healthcare outcomes. Online healthcare forums have the potential to offer a rich and unique source of data, revealing dimensions of antidepressant discontinuation that may not be captured by conventional data sources.

Methods

We analyzed 891 patient narratives from the online healthcare forum, "askapatient.com," utilizing content analysis to create PsyRisk-a corpus highlighting the risk factors associated with antidepressant discontinuation. Leveraging PsyRisk, alongside PsyTAR [a publicly available corpus of adverse drug reactions (ADRs) related to antidepressants], we developed a machine learning-driven algorithm for proactive identification of patients at risk of abrupt antidepressant discontinuation.

Results

From the analyzed 891 patients, 232 reported antidepressant discontinuation. Among these patients, 92% experienced ADRs, and 72% found these reactions distressful, negatively affecting their daily activities. Approximately 26% of patients perceived the antidepressants as ineffective. Most reported ADRs were physiological (61%, 411/673), followed by cognitive (30%, 197/673), and psychological (28%, 188/673) ADRs. In our study, we employed a nested cross-validation strategy with an outer 5-fold cross-validation for model selection, and an inner 5-fold cross-validation for hyperparameter tuning. The performance of our risk identification algorithm, as assessed through this robust validation technique, yielded an AUC-ROC of 90.77 and an F1-score of 83.33. The most significant contributors to abrupt discontinuation were high perceived distress from ADRs and perceived ineffectiveness of the antidepressants.

Conclusion

The risk factors identified and the risk identification algorithm developed in this study have substantial potential for clinical application. They could assist healthcare professionals in identifying and managing patients with depression who are at risk of prematurely discontinuing their antidepressant treatment.

Mapping 3 procedure coding systems to the International Classification of Health Interventions (ICHI): coverage and challenges

OBJECTIVE

To study the coverage and challenges in mapping 3 national and international procedure coding systems to the International Classification of Health Interventions (ICHI).

MATERIALS AND METHODS

We identified 300 commonly used codes each from SNOMED CT, ICD-10-PCS, and CCI (Canadian Classification of Health Interventions) and mapped them to ICHI. We evaluated the level of match at the ICHI stem code and Foundation Component levels. We used postcoordination (modification of existing codes by adding other codes) to improve matching. Failure analysis was done for cases where full representation was not achieved. We noted and categorized potential problems that we encountered in ICHI, which could affect the accuracy and consistency of mapping.

RESULTS

Overall, among the 900 codes from the 3 sources, 286 (31.8%) had full match with ICHI stem codes, 222 (24.7%) had full match with Foundation entities, and 231 (25.7%) had full match with postcoordination. 143 codes (15.9%) could only be partially represented even with postcoordination. A small number of SNOMED CT and ICD-10-PCS codes (18 codes, 2% of total), could not be mapped because the source codes were underspecified. We noted 4 categories of problems in ICHI-redundancy, missing elements, modeling issues, and naming issues.

CONCLUSION

Using the full range of mapping options, at least three-quarters of the commonly used codes in each source system achieved a full match. For the purpose of international statistical reporting, full matching may not be an essential requirement. However, problems in ICHI that could result in suboptimal maps should be addressed.

Prevalence and characteristics of long COVID in elderly patients: An observational cohort study of over 2 million adults in the US

BACKGROUND

Incidence of long COVID in the elderly is difficult to estimate and can be underreported. While long COVID is sometimes considered a novel disease, many viral or bacterial infections have been known to cause prolonged illnesses. We postulate that some influenza patients might develop residual symptoms that would satisfy the diagnostic criteria for long COVID, a condition we call "long Flu." In this study, we estimate the incidence of long COVID and long Flu among Medicare patients using the World Health Organization (WHO) consensus definition. We compare the incidence, symptomatology, and healthcare utilization between long COVID and long Flu patients.

METHODS AND FINDINGS

This is a cohort study of Medicare (the US federal health insurance program) beneficiaries over 65. ICD-10-CM codes were used to capture COVID-19, influenza, and residual symptoms. Long COVID was identified by (a) the designated long COVID code B94.8 (code-based definition), or (b) any of 11 symptoms identified in the WHO definition (symptom-based definition), from 1 to 3 months post-infection. A symptom would be excluded if it occurred in the year prior to infection. Long Flu was identified in influenza patients from the combined 2018 and 2019 Flu seasons by the same symptom-based definition for long COVID. Long COVID and long Flu were compared in 4 outcome measures: (a) hospitalization (any cause); (b) hospitalization (for long COVID symptom); (c) emergency department (ED) visit (for long COVID symptom); and (d) number of outpatient encounters (for long COVID symptom), adjusted for age, sex, race, region, Medicare-Medicaid dual eligibility status, prior-year hospitalization, and chronic comorbidities. Among 2,071,532 COVID-19 patients diagnosed between April 2020 and June 2021, symptom-based definition identified long COVID in 16.6% (246,154/1,479,183) and 29.2% (61,631/210,765) of outpatients and inpatients, respectively. The designated code gave much lower estimates (outpatients 0.49% (7,213/1,479,183), inpatients 2.6% (5,521/210,765)). Among 933,877 influenza patients, 17.0% (138,951/817,336) of outpatients and 24.6% (18,824/76,390) of inpatients fit the long Flu definition. Long COVID patients had higher incidence of dyspnea, fatigue, palpitations, loss of taste/smell, and neurocognitive symptoms compared to long Flu. Long COVID outpatients were more likely to have any-cause hospitalization (31.9% (74,854/234,688) versus 26.8% (33,140/123,736), odds ratio 1.06 (95% CI 1.05 to 1.08, p < 0.001)), and more outpatient visits than long Flu outpatients (mean 2.9(SD 3.4) versus 2.5(SD 2.7) visits, incidence rate ratio 1.09 (95% CI 1.08 to 1.10, p < 0.001)). There were less ED visits in long COVID patients, probably because of reduction in ED usage during the pandemic. The main limitation of our study is that the diagnosis of long COVID in is not independently verified.

CONCLUSIONS

Relying on specific long COVID diagnostic codes results in significant underreporting. We observed that about 30% of hospitalized COVID-19 patients developed long COVID. In a similar proportion of patients, long COVID-like symptoms (long Flu) can be observed after influenza, but there are notable differences in symptomatology between long COVID and long Flu. The impact of long COVID on healthcare utilization is higher than long Flu.

Sequential Mapping – A Novel Approach to Map from ICD-10-CM to ICD-11

Background:

ICD-11 will be used to report mortality statistics by WHO member countries starting in 2022. In the US, ICD-10-CM will likely continue to be used for morbidity coding for a long period of time. A map between ICD-10-CM and ICD-11 will therefore be useful for interoperability purpose between datasets coded with ICD-10-CM and ICD-11.

Objectives:

The objective of this study is to explore novel approaches to automatically derive a map between ICD-10-CM and ICD-11 through the sequential use of existing maps.

Methods and results:

Sequential mapping through ICD-10 yielded better coverage and accuracy compared to mapping through SNOMED CT.

Conclusions:

Sequential mapping is useful in automatically creating a draft map from ICD-10-CM to ICD-11 and would reduce manual curation efforts in creating the final map. The various approaches offer different trade-offs among coverage, recall and precision.

Effect of common maintenance drugs on the risk and severity of COVID-19 in elderly patients

Background

Maintenance drugs are used to treat chronic conditions. Several classes of maintenance drugs have attracted attention because of their potential to affect susceptibility to and severity of COVID-19.

Methods

Using claims data on 20% random sample of Part D Medicare enrollees from April to December 2020, we identified patients diagnosed with COVID-19. Using a nested case-control design, non-COVID-19 controls were identified by 1:5 matching on age, race, sex, dual-eligibility status, and geographical region. We identified usage of angiotensin-converting enzyme inhibitors (ACEI), angiotensin-receptor blockers (ARB), statins, warfarin, direct factor Xa inhibitors, P2Y12 inhibitors, famotidine and hydroxychloroquine based on Medicare prescription claims data. Using extended Cox regression models with time-varying propensity score adjustment we examined the independent effect of each study drug on contracting COVID-19. For severity of COVID-19, we performed extended Cox regressions on all COVID-19 patients, using COVID-19-related hospitalization and all-cause mortality as outcomes. Covariates included gender, age, race, geographic region, low-income indicator, and co-morbidities. To compensate for indication bias related to the use of hydroxychloroquine for the prophylaxis or treatment of COVID-19, we censored patients who only started on hydroxychloroquine in 2020.

Results

Up to December 2020, our sample contained 374,229 Medicare patients over 65 who were diagnosed with COVID-19. Among the COVID-19 patients, 278,912 (74.6%) were on at least one study drug. The three most common study drugs among COVID-19 patients were statins 187,374 (50.1%), ACEI 97,843 (26.2%) and ARB 83,290 (22.3%). For all three outcomes (diagnosis, hospitalization and death), current users of ACEI, ARB, statins, warfarin, direct factor Xa inhibitors and P2Y12 inhibitors were associated with reduced risks, compared to never users. Famotidine did not show consistent significant effects. Hydroxychloroquine did not show significant effects after censoring of recent starters.

Conclusion

Maintenance use of ACEI, ARB, warfarin, statins, direct factor Xa inhibitors and P2Y12 inhibitors was associated with reduction in risk of acquiring COVID-19 and dying from it.

Evaluation of the International Classification of Health Interventions (ICHI) in the coding of common surgical procedures

OBJECTIVE

: To evaluate the International Classification of Health Interventions (ICHI) in the clinical and statistical use cases.

MATERIALS AND METHODS

: We identified 300 most-performed surgical procedures as represented by their display names in an electronic health record. For comparison with existing coding systems, we coded the procedures in ICHI, SNOMED CT, International Classification of Diseases (ICD)-10-PCS, and CCI (Canadian Classification of Health Interventions), using postcoordination (modification of existing codes by adding other codes), when applicable. Failure analysis was done for cases where full representation was not achieved. The ICHI encoding was further evaluated for adequacy to support statistical reporting by the Organisation for Economic Co-operation and Development (OECD) and European Union (EU) categories of surgical procedures.

RESULTS

: After deduplication, 229 distinct procedures remained. Without postcoordination, ICHI achieved full representation in 52.8%. A further 19.2% could be fully represented with postcoordination. SNOMED CT was the best performing overall, with 94.3% full representation without postcoordination, and 99.6% with postcoordination. Failure analysis showed that "method" and "target" constituted most of the missing information for ICHI encoding. For all OECD/EU surgical categories, ICHI coding was adequate to support statistical reporting. One OECD/EU category ("Hip replacement, secondary") required postcoordination for correct assignment.

CONCLUSION

: In the clinical use case of capturing information in the electronic health record, ICHI was outperformed by the clinically oriented procedure coding systems (SNOMED CT and CCI), but was comparable to ICD-10-PCS. Postcoordination could be an effective and efficient means of improving coverage. ICHI is generally adequate for the collection of international statistics.

Feasibility of replacing the ICD-10-CM with the ICD-11 for morbidity coding: A content analysis

OBJECTIVE

The study sought to assess the feasibility of replacing the International Classification of Diseases-Tenth Revision-Clinical Modification (ICD-10-CM) with the International Classification of Diseases-11th Revision (ICD-11) for morbidity coding based on content analysis.

MATERIALS AND METHODS

The most frequently used ICD-10-CM codes from each chapter covering 60% of patients were identified from Medicare claims and hospital data. Each ICD-10-CM code was recoded in the ICD-11, using postcoordination (combination of codes) if necessary. Recoding was performed by 2 terminologists independently. Failure analysis was done for cases where full representation was not achieved even with postcoordination. After recoding, the coding guidance (inclusions, exclusions, and index) of the ICD-10-CM and ICD-11 codes were reviewed for conflict.

RESULTS

Overall, 23.5% of 943 codes could be fully represented by the ICD-11 without postcoordination. Postcoordination is the potential game changer. It supports the full representation of 8.6% of 943 codes. Moreover, with the addition of only 9 extension codes, postcoordination supports the full representation of 35.2% of 943 codes. Coding guidance review identified potential conflicts in 10% of codes, but mostly not affecting recoding. The majority of the conflicts resulted from differences in granularity and default coding assumptions between the ICD-11 and ICD-10-CM.

CONCLUSIONS

With some minor enhancements to postcoordination, the ICD-11 can fully represent almost 60% of the most frequently used ICD-10-CM codes. Even without postcoordination, 23.5% full representation is comparable to the 24.3% of ICD-9-CM codes with exact match in the ICD-10-CM, so migrating from the ICD-10-CM to the ICD-11 is not necessarily more disruptive than from the International Classification of Diseases-Ninth Revision-Clinical Modification to the ICD-10-CM. Therefore, the ICD-11 (without a CM) should be considered as a candidate to replace the ICD-10-CM for morbidity coding.

The changing patterns of comorbidities associated with human immunodeficiency virus infection, a longitudinal retrospective cohort study of Medicare patients

The objective of this paper is to determine the temporal trend of the association of 66 comorbidities with human immunodeficiency virus (HIV) infection status among Medicare beneficiaries from 2000 through 2016.We harvested patient level encounter claims from a 17-year long 100% sample of Medicare records. We used the chronic conditions warehouse comorbidity flags to determine HIV infection status and presence of comorbidities. We prepared 1 data set per year for analysis. Our 17 study data sets are retrospective annualized patient level case histories where the comorbidity status reflects if the patient has ever met the comorbidity case definition from the start of the study to the analysis year.We implemented one logistic binary regression model per study year to discover the maximum likelihood estimate (MLE) of a comorbidity belonging to our binary classes of HIV+ or HIV- study populations. We report MLE and odds ratios by comorbidity and year.Of the 66 assessed comorbidities, 35 remained associated with HIV- across all model years, 19 remained associated with HIV+ across all model years. Three comorbidities changed association from HIV+ to HIV- and 9 comorbidities changed association from HIV- to HIV+.The prevalence of comorbidities associated with HIV infection changed over time due to clinical, social, and epidemiological reasons. Comorbidity surveillance can provide important insights into the understanding and management of HIV infection and its consequences.

Use of word and graph embedding to measure semantic relatedness between Unified Medical Language System concepts

Objective

The study sought to explore the use of deep learning techniques to measure the semantic relatedness between Unified Medical Language System (UMLS) concepts.

Materials and Methods

Concept sentence embeddings were generated for UMLS concepts by applying the word embedding models BioWordVec and various flavors of BERT to concept sentences formed by concatenating UMLS terms. Graph embeddings were generated by the graph convolutional networks and 4 knowledge graph embedding models, using graphs built from UMLS hierarchical relations. Semantic relatedness was measured by the cosine between the concepts’ embedding vectors. Performance was compared with 2 traditional path-based (shortest path and Leacock-Chodorow) measurements and the publicly available concept embeddings, cui2vec, generated from large biomedical corpora. The concept sentence embeddings were also evaluated on a word sense disambiguation (WSD) task. Reference standards used included the semantic relatedness and semantic similarity datasets from the University of Minnesota, concept pairs generated from the Standardized MedDRA Queries and the MeSH (Medical Subject Headings) WSD corpus.

Results

Sentence embeddings generated by BioWordVec outperformed all other methods used individually in semantic relatedness measurements. Graph convolutional network graph embedding uniformly outperformed path-based measurements and was better than some word embeddings for the Standardized MedDRA Queries dataset. When used together, combined word and graph embedding achieved the best performance in all datasets. For WSD, the enhanced versions of BERT outperformed BioWordVec.

Conclusions

Word and graph embedding techniques can be used to harness terms and relations in the UMLS to measure semantic relatedness between concepts. Concept sentence embedding outperforms path-based measurements and cui2vec, and can be further enhanced by combining with graph embedding.

The new International Classification of Diseases 11th edition: a comparative analysis with ICD-10 and ICD-10-CM

OBJECTIVE

To study the newly adopted International Classification of Diseases 11th revision (ICD-11) and compare it to the International Classification of Diseases 10th revision (ICD-10) and International Classification of Diseases 10th revision-Clinical Modification (ICD-10-CM).

MATERIALS AND METHODS

: Data files and maps were downloaded from the World Health Organization (WHO) website and through the application programming interfaces. A round trip method based on the WHO maps was used to identify equivalent codes between ICD-10 and ICD-11, which were validated by limited manual review. ICD-11 terms were mapped to ICD-10-CM through normalized lexical mapping. ICD-10-CM codes in 6 disease areas were also manually recoded in ICD-11.

RESULTS

Excluding the chapters for traditional medicine, functioning assessment, and extension codes for postcoordination, ICD-11 has 14 622 leaf codes (codes that can be used in coding) compared to ICD-10 and ICD-10-CM, which has 10 607 and 71 932 leaf codes, respectively. We identified 4037 pairs of ICD-10 and ICD-11 codes that were equivalent (estimated accuracy of 96%) by our round trip method. Lexical matching between ICD-11 and ICD-10-CM identified 4059 pairs of possibly equivalent codes. Manual recoding showed that 60% of a sample of 388 ICD-10-CM codes could be fully represented in ICD-11 by precoordinated codes or postcoordination.

CONCLUSION

In ICD-11, there is a moderate increase in the number of codes over ICD-10. With postcoordination, it is possible to fully represent the meaning of a high proportion of ICD-10-CM codes, especially with the addition of a limited number of extension codes.

Using Medicare Data to Assess the Proarrhythmic Risk of Non-Cardiac Treatment Drugs that Prolong the QT Interval in Older Adults: An Observational Cohort Study

Introduction

Serious cardiac arrhythmias caused by QT-prolonging drugs are difficult to predict based on physiological measurement and pre-approval clinical trials. Post-marketing surveillance and monitoring are important to generate safety data.

Objectives

To assess whether an observational study using Medicare claims data can detect the arrhythmogenic risk of QT-prolonging drugs.

Methods

We identified 17 QT-prolonging drugs with known risk of torsades des pointes (TdP) that were not used to treat cardiac arrhythmias. Amoxicillin and four serotonin-norepinephrine reuptake inhibitors (SNRIs) were used as controls. De-identified claims data of 1.2 million Medicare beneficiaries were accessed. Two separate Cox regressions were done for short-term and chronic-use drugs. The primary outcome was a composite of ventricular arrhythmias and/or sudden death, identified by ICD diagnostic codes. We explored the independent effect of each study drug on the outcomes. Other covariates included patient demographics, comorbidities, and known risk factors for drug-induced cardiac arrhythmia.

Results

We were able to detect increased risk in 14 of 17 study drugs (82.3%), and none of the control drugs. Among the fluoroquinolones, ciprofloxacin was the safest. Azithromycin and clarithromycin were relatively safe compared to erythromycin. Compared to SNRIs, both citalopram and escitalopram had increased risk, more so with escitalopram than citalopram. Comorbidities associated with increased risk included ischemic heart disease, electrolyte imbalance, bradycardia, acute myocardial infarction, heart failure, and chronic kidney and liver disease.

Conclusion

Medicare data can be utilized for post-marketing surveillance and monitoring of the proarrhythmic risk of QT-prolonging drugs in older adults.

Supplementary Information

The online version contains supplementary material available at 10.1007/s40801-021-00230-1.

Evaluation of Research Accessibility and Data Elements of HIV Registries

BACKGROUND

Patient registries represent a long-term data collection system that is a platform for performing multiple research studies to generate real-world evidence. Many of these registries use common data elements (CDEs) and link data from Electronic Health Records.

OBJECTIVE

This study evaluated HIV registry features that contribute to the registry's usability for retrospective analysis of existing registry data or new prospective interventional studies.

METHODS

We searched PubMed and ClinicalTrials.gov (CTG) to generate a list of HIV registries. We used the framework developed by the European Medical Agency (EMA) to evaluate the registries by determining the presence of key research features. These features included information about the registry, request and collaboration processes, and available data. We acquired data dictionaries and identified CDEs.

RESULTS

We found 13 HIV registries that met our criteria, 11 through PubMed and 2 through CTG. The prevalence of the evaluated features ranged from all 13 (100%) having published key registry information to 0 having a research contract template. We analyzed 6 data dictionaries and identified 14 CDEs that were present in at least 4 of 6 (66.7%) registry data dictionaries.

CONCLUSION

The importance of registries as platforms for research data is growing and the presence of certain features, including data dictionaries, contributes to the reuse and secondary research capabilities of a registry. We found some features such as collaboration policies were in the majority of registries while others such as, ethical support, were in a few and are more for future development.

Sharing of Individual Participant Data from Clinical Trials: General Comparison and HIV Use Case

Sharing of individual participant data is encouraged by the International Committee of Medical Journal Editors. We analyzed clinical trial registry data from ClinicalTrials.gov (CTG) and determined the proportion of trials sharing de-identified Individual Participant Data (IPD). We looked at 3,138 medical conditions (as Medical Subject Heading terms). Overall, 10.8% of trials with first registration date after December 1, 2015 answered 'Yes' to plan to share de-identified IPD data. This sharing rate ranges between 0% (biliary tract neoplasms) to 72.2% (meningitis, meningococcal) when analyzed by disease that is focus of a study. Via a predictive model, we found that studies that deposited basic summary results data to CTG results registry, large studies and phase 3 interventional studies are most likely to declare intent to share IPD data. As part of an HIV common data element analysis project, we further compared a body of HIV trials (24% sharing rate) to other diseases.

The Use of Inter-terminology Maps for the Creation and Maintenance of Value Sets

Value sets are essential in activities such as electronic clinical quality measures (eCQM) and patient cohort definition. Creation and maintenance of value sets is labor intensive and error prone. Our method aims to use existing inter-terminology maps to improve the quality of value sets that are defined in more than one terminology. For 197 eCQM value sets defined in SNOMED CT plus ICD-9-CM and/or ICD-10-CM, the map-generated codes showed good overlap with the value set codes. Manual review showed that some new codes identified by mapping should probably be included in the value sets. This could potentially augment the ICD-9-CM codes by 45% (1.5 codes), ICD-10-CM codes by 25% (1.8 codes) and SNOMED CT codes by up to 42% (4.8 codes) per value set on average. The mapping between SNOMED CT and ICD-10-PCS did not perform as well because of the granularity discrepancy in the map.

A value set for documenting adverse reactions in electronic health records

Objective

To develop a comprehensive value set for documenting and encoding adverse reactions in the allergy module of an electronic health record.

Materials and Methods

We analyzed 2 471 004 adverse reactions stored in Partners Healthcare's Enterprise-wide Allergy Repository (PEAR) of 2.7 million patients. Using the Medical Text Extraction, Reasoning, and Mapping System, we processed both structured and free-text reaction entries and mapped them to Systematized Nomenclature of Medicine - Clinical Terms. We calculated the frequencies of reaction concepts, including rare, severe, and hypersensitivity reactions. We compared PEAR concepts to a Federal Health Information Modeling and Standards value set and University of Nebraska Medical Center data, and then created an integrated value set.

Results

We identified 787 reaction concepts in PEAR. Frequently reported reactions included: rash (14.0%), hives (8.2%), gastrointestinal irritation (5.5%), itching (3.2%), and anaphylaxis (2.5%). We identified an additional 320 concepts from Federal Health Information Modeling and Standards and the University of Nebraska Medical Center to resolve gaps due to missing and partial matches when comparing these external resources to PEAR. This yielded 1106 concepts in our final integrated value set. The presence of rare, severe, and hypersensitivity reactions was limited in both external datasets. Hypersensitivity reactions represented roughly 20% of the reactions within our data.

Discussion

We developed a value set for encoding adverse reactions using a large dataset from one health system, enriched by reactions from 2 large external resources. This integrated value set includes clinically important severe and hypersensitivity reactions.

Conclusion

This work contributes a value set, harmonized with existing data, to improve the consistency and accuracy of reaction documentation in electronic health records, providing the necessary building blocks for more intelligent clinical decision support for allergies and adverse reactions.

Map-Assisted Generation of Procedure and Intervention Encoding (Magpie): An Innovative Approach for ICD-10-PCS Coding

ICD-10-PCS coding is challenging because of the large number of codes, non-intuitive terms and paucity of the ICD-10-PCS index. We previously repurposed the richer ICD-9-CM procedure index for ICD-10-PCS coding. We have developed the MAGPIE tool based on the repurposed ICD-9-CM index with other lexical and mapping resources. MAGPIE helps the user to identify SNOMED CT and ICD-10-PCS codes for medical procedures. MAGPIE uses three innovative search approaches: cascading search (SNOMED CT to ICD-9-CM to ICD-10-PCS), hybrid lexical and map-assisted matching, and semantic filtering of ICD-10-PCS codes. Our evaluation showed that MAGPIE found the correct SNOMED CT code and ICD-10-PCS table in 70% and 85% of cases respectively, without any user intervention. MAGPIE is available online from the NLM website: magpie.nlm.nih.gov.

The PsyTAR dataset: From patients generated narratives to a corpus of adverse drug events and effectiveness of psychiatric medications

The "Psychiatric Treatment Adverse Reactions" (PsyTAR) dataset contains patients' expression of effectiveness and adverse drug events associated with psychiatric medications. The PsyTAR was generated in four phases. In the first phase, a sample of 891 drugs reviews posted by patients on an online healthcare forum, "askapatient.com", was collected for four psychiatric drugs: Zoloft, Lexapro, Cymbalta, and Effexor XR. For each drug review, patient demographic information, duration of treatment, and satisfaction with the drugs were reported. In the second phase, sentence classification, drug reviews were split to 6009 sentences, and each sentence was labeled for the presence of Adverse Drug Reaction (ADR), Withdrawal Symptoms (WDs), Sign/Symptoms/Illness (SSIs), Drug Indications (DIs), Drug Effectiveness (EF), Drug Infectiveness (INF), and Others (not applicable). In the third phases, entities including ADRs (4813 mentions), WDs (590 mentions), SSIs (1219 mentions), and DIs (792 mentions) were identified and extracted from the sentences. In the four phases, all the identified entities were mapped to the corresponding UMLS Metathesaurus concepts (916) and SNOMED CT concepts (755). In this phase, qualifiers representing severity and persistency of ADRs, WDs, SSIs, and DIs (e.g., mild, short term) were identified. All sentences and identified entities were linked to the original post using IDs (e.g., Zoloft.1, Effexor.29, Cymbalta.31). The PsyTAR dataset can be accessed via Online Supplement #1 under the CC BY 4.0 Data license. The updated versions of the dataset would also be accessible in https://sites.google.com/view/pharmacovigilanceinpsychiatry/home.

Using SNOMED CT-encoded problems to improve ICD-10-CM coding-A randomized controlled experiment

OBJECTIVE

Clinical problems in the Electronic Health Record that are encoded in SNOMED CT can be translated into ICD-10-CM codes through the NLM's SNOMED CT to ICD-10-CM map (NLM Map). This study evaluates the potential benefits of using the map-generated codes to assist manual ICD-10-CM coding.

METHODS

De-identified clinic notes taken by the physician during an outpatient encounter were made available on a secure web server and randomly assigned for coding by professional coders with usual coding or map-assisted coding. Map-assisted coding made use of the problem list maintained by the physician and the NLM Map to suggest candidate ICD-10-CM codes to the coder. A gold standard set of codes for each note was established by the coders using a Delphi consensus process. Outcomes included coding time, coding reliability as measured by the Jaccard coefficients between codes from two coders with the same method of coding, and coding accuracy as measured by recall, precision and F-score according to the gold standard.

RESULTS

With map-assisted coding, the average coding time per note reduced by 1.5 min (p = 0.006). There was a small increase in coding reliability and accuracy (not statistical significant). The benefits were more pronounced in the more experienced than less experienced coders. Detailed analysis of cases in which the correct ICD-10-CM codes were not found by the NLM Map showed that most failures were related to omission in the problem list and suboptimal mapping of the problem list terms to SNOMED CT. Only 12% of the failures was caused by errors in the NLM Map.

CONCLUSION

Map-assisted coding reduces coding time and can potentially improve coding reliability and accuracy, especially for more experienced coders. More effort is needed to improve the accuracy of the map-suggested ICD-10-CM codes.

Re-purposing the ICD-9-CM Procedures Index for Coding in ICD-10-PCS and SNOMED CT

Compared to the ICD-10-PCS index, the ICD-9-CM Procedure (ICD9V3) Index is richer and contains more clinician-friendly terms including abbreviations and eponyms. We re-purposed the ICD9V3 index by mapping the index terms to SNOMED CT and the ICD-9-CM codes to ICD-10-PCS codes through the General Equivalent Mappings. The re-purposed index outperformed the ICD-10-PCS index in the retrieval of ICD-10-PCS codes using a list of commonly used procedure names, with significantly higher recall, precision and F-score. We also derived a SNOMED CT to ICD-10-PCS map from the re-purposed ICD-9-CM index, which had a higher coverage of SNOMED CT concepts and comparable accuracy compared to a map derived from the ICD-10-PCS index. The re-purposed index will be a useful resource for ICD-10-PCS coders and for mapping between SNOMED CT and ICD-10-PCS.

Identifying the Underlying Factors Associated With Patients' Attitudes Toward Antidepressants: Qualitative and Quantitative Analysis of Patient Drug Reviews

BACKGROUND

Nonadherence to antidepressants is a major obstacle to deriving antidepressants' therapeutic benefits, resulting in significant burdens on the individuals and the health care system. Several studies have shown that nonadherence is weakly associated with personal and clinical variables but strongly associated with patients' beliefs and attitudes toward medications. Patients' drug review posts in online health care communities might provide a significant insight into patients' attitude toward antidepressants and could be used to address the challenges of self-report methods such as patients' recruitment.

OBJECTIVE

The aim of this study was to use patient-generated data to identify factors affecting the patient's attitude toward 4 antidepressants drugs (sertraline [Zoloft], escitalopram [Lexapro], duloxetine [Cymbalta], and venlafaxine [Effexor XR]), which in turn, is a strong determinant of treatment nonadherence. We hypothesized that clinical variables (drug effectiveness; adverse drug reactions, ADRs; perceived distress from ADRs, ADR-PD; and duration of treatment) and personal variables (age, gender, and patients' knowledge about medications) are associated with patients' attitude toward antidepressants, and experience of ADRs and drug ineffectiveness are strongly associated with negative attitude.

METHODS

We used both qualitative and quantitative methods to analyze the dataset. Patients' drug reviews were randomly selected from a health care forum called askapatient. The Framework method was used to build the analytical framework containing the themes for developing structured data from the qualitative drug reviews. Then, 4 annotators coded the drug reviews at the sentence level using the analytical framework. After managing missing values, we used chi-square and ordinal logistic regression to test and model the association between variables and attitude.

RESULTS

A total of 892 reviews posted between February 2001 and September 2016 were analyzed. Most of the patients were females (680/892, 76.2%) and aged less than 40 years (540/892, 60.5%). Patient attitude was significantly (P<.001) associated with experience of ADRs, ADR-PD, drug effectiveness, perceived lack of knowledge, experience of withdrawal, and duration of usage, whereas oth age (F_4,874=0.72, P=.58) and gender (χ²₄=2.7, P=.21) were not found to be associated with patient attitudes. Moreover, modeling the relationship between variables and attitudes showed that drug effectiveness and perceived distress from adverse drug reactions were the 2 most significant factors affecting patients' attitude toward antidepressants.

CONCLUSIONS

Patients' self-report experiences of medications in online health care communities can provide a direct insight into the underlying factors associated with patients' perceptions and attitudes toward antidepressants. However, it cannot be used as a replacement for self-report methods because of the lack of information for some of the variables, colloquial language, and the unstructured format of the reports.

Achieving Logical Equivalence between SNOMED CT and ICD-10-PCS Surgical Procedures

Surgical procedures are coded in SNOMED CT in the electronic health record and in ICD-10-PCS in administrative systems. We compared the logical definitions of SNOMED CT concepts to the ICD-10-PCS axial components to identify overlap and gaps. The biggest discrepancy was in the surgical approach which was specified in all ICD-10-PCS codes but only in 8.7% of SNOMED CT surgical procedures. Among the top 100 commonly used ICD-10-PCS codes, 25% could be matched fully in meaning and logical definition to pre-coordinated SNOMED CT concepts. Using post-coordination, it was possible to represent the full meaning of 86% of ICD-10-PCS codes. Logical mapping between SNOMED CT and ICD-10-PCS is feasible but will be more productive if more SNOMED CT concepts can become fully-defined. Short of full logical matching, partial logical matches can also be useful in suggesting candidate maps for expert review and to support interactive post-coordination.

Comparison of three commercial knowledge bases for detection of drug-drug interactions in clinical decision support

Objective

To compare 3 commercial knowledge bases (KBs) used for detection and avoidance of potential drug-drug interactions (DDIs) in clinical practice.

Methods

Drugs in the DDI tables from First DataBank (FDB), Micromedex, and Multum were mapped to RxNorm. The KBs were compared at the clinical drug, ingredient, and DDI rule levels. The KBs were evaluated against a reference list of highly significant DDIs from the Office of the National Coordinator for Health Information Technology (ONC). The KBs and the ONC list were applied to a prescription data set to simulate their use in clinical decision support.

Results

The KBs contained 1.6 million (FDB), 4.5 million (Micromedex), and 4.8 million (Multum) clinical drug pairs. Altogether, there were 8.6 million unique pairs, of which 79% were found only in 1 KB and 5% in all 3 KBs. However, there was generally more agreement than disagreement in the severity rankings, especially in the contraindicated category. The KBs covered 99.8-99.9% of the alerts of the ONC list and would have generated 25 (FDB), 145 (Micromedex), and 84 (Multum) alerts per 1000 prescriptions.

Conclusion

The commercial KBs differ considerably in size and quantity of alerts generated. There is less variability in severity ranking of DDIs than suggested by previous studies. All KBs provide very good coverage of the ONC list. More work is needed to standardize the editorial policies and evidence for inclusion of DDIs to reduce variation among knowledge sources and improve relevance. Some DDIs considered contraindicated in all 3 KBs might be possible candidates to add to the ONC list.

Leveraging Lexical Matching and Ontological Alignment to Map SNOMED CT Surgical Procedures to ICD-10-PCS

In 2015 ICD-10-PCS replaced ICD-9-CM for coding medical procedures in the U.S. We explored two methods to automatically map SNOMED CT surgical procedures to ICD-10-PCS. First, we used MetaMap to lexically map ICD-10-PCS index terms to SNOMED CT. Second, we made use of the axial structure of ICD-10-PCS and aligned them to defining attributes in SNOMED CT. Lexical mapping produced 45% of correct maps and 44% of broader maps. Ontological mappings were 40% correct and 5% broader. Both correct and broader maps will be useful in assisting mappers to create the map. When the two mapping methods agreed, the accuracy increased to 93%. Reviewing the MetaMap generated body part mappings and using additional information in the SNOMED CT names and definitions can lead to better results for the ontological map.

Using an Ontology-Based Approach to Handle Author Affiliations in a Large Biomedical Citation Database

To handle differences in affiliation names submitted with biomedical journal articles, we build an affiliation knowledge base named Authority File for Affiliations (AFA) based on ontology principles. There are currently 113,700 affiliation concepts with about 583,700 affiliation names. The AFA becomes an essential tool in managing citation information and data analysis.

MetaMap Lite in Excel: Biomedical Named-Entity Recognition for Non-Technical Users

We developed an easy-to-use tool for non-technical biomedical researchers to conduct Named-Entity Recognition (NER) on biomedical text, in a familiar spreadsheet environment. The system is a simple, offline, easy to install, end-user front-end to the new MetaMap Lite. Early adopters found it to be a quick starting-point to incorporate NER in their investigations.

Preparing for the ICD-10-CM Transition: Automated Methods for Translating ICD Codes in Clinical Phenotype Definitions

Background:

The national mandate for health systems to transition from ICD-9-CM to ICD-10-CM in October 2015 has an impact on research activities. Clinical phenotypes defined by ICD-9-CM codes need to be converted to ICD-10-CM, which has nearly four times more codes and a very different structure than ICD-9-CM.

Methods:

We used the Centers for Medicare & Medicaid Services (CMS) General Equivalent Maps (GEMs) to translate, using four different methods, condition-specific ICD-9-CM code sets used for pragmatic trials (n=32) into ICD-10-CM. We calculated the recall, precision, and F score of each method. We also used the ICD-9-CM and ICD-10-CM value sets defined for electronic quality measure as an additional evaluation of the mapping methods.

Results:

The forward-backward mapping (FBM) method had higher precision, recall and F-score metrics than simple forward mapping (SFM). The more aggressive secondary (SM) and tertiary mapping (TM) methods resulted in higher recall but lower precision. For clinical phenotype definition, FBM was the best (F=0.67), but was close to SM (F=0.62) and TM (F=0.60), judging on the F-scores alone. The overall difference between the four methods was statistically significant (one-way ANOVA, F=5.749, p=0.001). However, pairwise comparisons between FBM, SM, and TM did not reach statistical significance. A similar trend was found for the quality measure value sets.

Discussion:

The optimal method for using the GEMs depends on the relative importance of recall versus precision for a given use case. It appears that for clinically distinct and homogenous conditions, the recall of FBM is sufficient. The performance of all mapping methods was lower for heterogeneous conditions. Since code sets used for phenotype definition and quality measurement can be very similar, there is a possibility of cross-fertilization between the two activities.

Conclusion:

Different mapping approaches yield different collections of ICD-10-CM codes. All methods require some level of human validation.

An exploration of the properties of the CORE problem list subset and how it facilitates the implementation of SNOMED CT

Objective Systematized Nomenclature of Medicine Clinical Terms (SNOMED CT) is the emergent international health terminology standard for encoding clinical information in electronic health records. The CORE Problem List Subset was created to facilitate the terminology’s implementation. This study evaluates the CORE Subset’s coverage and examines its growth pattern as source datasets are being incorporated.

Methods Coverage of frequently used terms and the corresponding usage of the covered terms were assessed by “leave-one-out” analysis of the eight datasets constituting the current CORE Subset. The growth pattern was studied using a retrospective experiment, growing the Subset one dataset at a time and examining the relationship between the size of the starting subset and the coverage of frequently used terms in the incoming dataset. Linear regression was used to model that relationship.

Results On average, the CORE Subset covered 80.3% of the frequently used terms of the left-out dataset, and the covered terms accounted for 83.7% of term usage. There was a significant positive correlation between the CORE Subset’s size and the coverage of the frequently used terms in an incoming dataset. This implies that the CORE Subset will grow at a progressively slower pace as it gets bigger.

Conclusion The CORE Problem List Subset is a useful resource for the implementation of Systematized Nomenclature of Medicine Clinical Terms in electronic health records. It offers good coverage of frequently used terms, which account for a high proportion of term usage. If future datasets are incorporated into the CORE Subset, it is likely that its size will remain small and manageable.

Coverage of rare disease names in standard terminologies and implications for patients, providers, and research

Small numbers of patients are a special challenge for rare diseases research. Electronic health record (EHR) data can facilitate research if patients with rare diseases can be reliably identified. We estimate the coverage of the names of a set of 6,519 rare diseases. Using the UMLS, 697 (11%) diseases were matched to ICD-9-CM, 1,386 (21%) to ICD-10-CM and 2,848 (44%) to SNOMED CT. Using published mappings from SNOMED CT to ICD, we further estimate additional broader matches of 2,569 (39%) rare diseases to ICD-9-CM and 1,635 (25%) to ICD-10-CM. The number of codes that match one and only one disease are 1,081 (62%) for ICD-9-CM, 1,403 (73%) for ICD-10-CM, and 3,311 (85%) for SNOMED CT. Our findings confirm that SNOMED CT has the greatest coverage and specificity needed to identify patients with a rare disease from EHR-data, and can facilitate research and evidence-based care.

Comparison of electronic pharmacy prescription records with manually collected medication histories in an emergency department

STUDY OBJECTIVE

Medication history is an essential part of patient assessment in emergency care. Patient-reported medication history can be incomplete. We study whether an electronic pharmacy-sourced prescription record can supplement the patient-reported history.

METHODS

In a community hospital, we compared the patient-reported history obtained by triage nurses to a proprietary electronic pharmacy record in all emergency department (ED) patients during 3 months.

RESULTS

Of 9,426 triaged patients, 5,001 (53%) had at least 1 (mean 7.7) prescription medication in the full-year electronic pharmacy record. Counting only recent prescription medications (supply lasting to at least 7 days before the ED visit), 3,688 patients (39%) had at least 1 (mean 4.0) recent medication. After adjustment for possible false-positive results, recent electronic prescription medication record enriched the patient-reported history by 28% (adding 1.1 drugs per patient). However, only 60% of patients with any active prescription medications from either source had any recent prescription medications in their electronic pharmacy record.

CONCLUSION

The electronic pharmacy prescription record augments the manually collected history.

Extracting drug indication information from structured product labels using natural language processing

OBJECTIVE

To extract drug indications from structured drug labels and represent the information using codes from standard medical terminologies.

MATERIALS AND METHODS

We used MetaMap and other publicly available resources to extract information from the indications section of drug labels. Drugs and indications were encoded by RxNorm and UMLS identifiers respectively. A sample was manually reviewed. We also compared the results with two independent information sources: National Drug File-Reference Terminology and the Semantic Medline project.

RESULTS

A total of 6797 drug labels were processed, resulting in 19 473 unique drug-indication pairs. Manual review of 298 most frequently prescribed drugs by seven physicians showed a recall of 0.95 and precision of 0.77. Inter-rater agreement (Fleiss κ) was 0.713. The precision of the subset of results corroborated by Semantic Medline extractions increased to 0.93.

DISCUSSION

Correlation of a patient's medical problems and drugs in an electronic health record has been used to improve data quality and reduce medication errors. Authoritative drug indication information is available from drug labels, but not in a format readily usable by computer applications. Our study shows that it is feasible to use publicly available natural language processing resources to extract drug indications from drug labels. The same method can be applied to other sections of the drug label-for example, adverse effects, contraindications.

CONCLUSIONS

It is feasible to use publicly available natural language processing tools to extract indication information from freely available drug labels. Named entity recognition sources (eg, MetaMap) provide reasonable recall. Combination with other data sources provides higher precision.

Semantic interoperation and electronic health records: context sensitive mapping from SNOMED CT to ICD-10

An important case for successful deployment of a lifetime electronic health record is reuse of clinical data from the electronic health record (EHR) for epidemiology, reimbursement, and research. We report a collaboration between the IHTSDO and the WHO to develop knowledge-based tools supporting translation of data from SNOMED CT to the ICD-10 classification. These tools have been vetted by an international community and are available for system vendors to enhance the interoperability of their products. The maps we created are also informing the development of the next generation of classifications which will employ a common ontology base between SNOMED CT and ICD-11 to promote interoperability.

Handling age specification in the SNOMED CT to ICD-10-CM cross-map

A SNOMED CT-encoded problem list will be required to satisfy the Certification Criteria for Stage 2 "Meaningful Use" of the EHR incentive program. ICD-10-CM will be replacing ICD-9-CM as the reimbursement code set in the near future. Having a cross-map from SNOMED CT to ICD-10-CM will promote the use of SNOMED CT as the primary problem list terminology, while easing the transition to ICD-10-CM. This rule-based map will support semi-automatic generation of ICD-10-CM codes from SNOMED CT-encoded data. Among the different types of rules, the age rule is used to handle age-specific code assignment in ICD-10-CM. To supplement the manual process of creation of age rules, a special QA process was implemented to flag maps that were potentially missing age rules. The QA flagged 342 concepts for review (out of 7,277), of which 172 concepts (50.3%) were true positives. Without the special QA, many of the age rules would have been missed.

Synergism between the mapping projects from SNOMED CT to ICD-10 and ICD-10-CM

Two mapping projects are currently underway, creating maps from SNOMED CT to ICD-10 and ICD-10-CM respectively. Even though the two projects belong to different organizations, there has been a lot of synergism between them. The ICD-10-CM map project heavily re-used the mapping methodology, tools and map data developed in the ICD-10 map project. An algorithm was derived to generate candidate ICD-10-CM map records from the ICD-10 map. We evaluated the algorithm in 5,264 SNOMED CT concepts common to both maps. 4,317 ICD-10-CM candidate maps could be generated from the ICD-10 map and 3,341 (77%) of the generated maps agreed with the published map. By priming the mapping process with candidate maps generated from the other map project, significant saving in time and effort in future phases of the two projects can be anticipated. The reasons for the discordance between the generated map and published map were also analyzed.

Can SNOMED CT fulfill the vision of a compositional terminology? Analyzing the use case for problem list

We analyzed 598 of 63,952 terms employed in problem list entries from seven major healthcare institutions that were not mapped with UMLS to SNOMED CT when preparing the NLM UMLS-CORE problem list subset. We intended to determine whether published or post-coordinated SNOMED concepts could accurately capture the problems as stated by the clinician and to characterize the workload for the local terminology manager. From the terms we analyzed, we estimate that 7.5% of the total terms represent ambiguous statements that require clarification. Of those terms which were unambiguous, we estimate that 38.1% could be encoded using the SNOMED CT January 2011 pre-coordinated (published core) content. 60.4% of unambiguous terms required post-coordination to capture the term meaning within the SNOMED model. Approximately 28.5% of post-coordinated content could not be fully defined and required primitive forms. This left 1.5% of unambiguous terms which were expressed with meaning which could not be represented in SNOMED CT. We estimate from our study that 98.5% of clinical terms unambiguously suggested for the problem list can be equated to published concepts or can be modeled with SNOMED CT but that roughly one in four SNOMED modeled expressions fail to represent the full meaning of the term. Implications for the business model of the local terminology manager and the development of SNOMED CT are discussed.

Testing Three Problem List Terminologies in a simulated data entry environment

Three Problem List Terminologies (PLT) were tested using a web-based application simulating a clinical data entry environment to evaluate coverage and coding efficiency. The three PLTs were: the CORE Problem List Subset of SNOMED CT, a clinical subset extracted from the full SNOMED CT and the PLT currently used at the Mayo Clinic. Candidate problem statements were randomly extracted from free text problem list entries contained in two electronic medical record systems. Physician reviewers searched for concepts in one of the three PLTs that most closely matched a problem statement. Altogether 45 reviewers reviewed 15 problems each. The coverage of the much smaller CORE Subset was comparable to Clinical SNOMED for combined exact or partial matches. The CORE Subset required the shortest time to find a concept. This may be related to the smaller size of the pick lists for the CORE Subset.

The UMLS-CORE project: a study of the problem list terminologies used in large healthcare institutions

OBJECTIVE

To study existing problem list terminologies (PLTs), and to identify a subset of concepts based on standard terminologies that occur frequently in problem list data.

DESIGN

Problem list terms and their usage frequencies were collected from large healthcare institutions.

MEASUREMENT

The pattern of usage of the terms was analyzed. The local terms were mapped to the Unified Medical Language System (UMLS). Based on the mapped UMLS concepts, the degree of overlap between the PLTs was analyzed.

RESULTS

Six institutions submitted 76,237 terms and their usage frequencies in 14 million patients. The distribution of usage was highly skewed. On average, 21% of unique terms already covered 95% of usage. The most frequently used 14,395 terms, representing the union of terms that covered 95% of usage in each institution, were exhaustively mapped to the UMLS. 13,261 terms were successfully mapped to 6776 UMLS concepts. Less frequently used terms were generally less 'mappable' to the UMLS. The mean pairwise overlap of the PLTs was only 21% (median 19%). Concepts that were shared among institutions were used eight times more often than concepts unique to one institution. A SNOMED Problem List Subset of frequently used problem list concepts was identified.

CONCLUSIONS

Most of the frequently used problem list terms could be found in standard terminologies. The overlap between existing PLTs was low. The use of the SNOMED Problem List Subset will save developmental effort, reduce variability of PLTs, and enhance interoperability of problem list data.

RxTerms - a drug interface terminology derived from RxNorm

A good interface terminology is an essential component of any Computerized Provider Order Entry system. RxTerms is a drug interface terminology derived from RxNorm. By reorganizing the drug information into two dimensions as prescribers do when writing prescriptions and by eliminating drug names that are less likely to be needed in a prescribing environment, RxTerms helps the user to efficiently enter complete prescription orders. Preliminary evaluation of RxTerms using a list of most commonly prescribed drugs showed that its coverage was very good (99% for both generic and branded drug names). There was significant efficiency gain compared to using the unprocessed RxNorm names. RxTerms fills the gap for a free, up-to-date drug interface terminology that is linked to RxNorm, the U.S. designated standard for clinical drugs.

UMLS-based automatic image indexing

To date, most accurate image retrieval techniques rely on textual descriptions of images. Our goal is to automatically generate indexing terms for an image extracted from a biomedical article by identifying Unified Medical Language System (UMLS) concepts in image caption and its discussion in the text. In a pilot evaluation of the suggested image indexing method by five physicians, a third of the automatically identified index terms were found suitable for indexing.

Heterogeneous but "standard" coding systems for adverse events: Issues in achieving interoperability between apples and oranges

Monitoring adverse events (AEs) is an important part of clinical research and a crucial target for data standards. The representation of adverse events themselves requires the use of controlled vocabularies with thousands of needed clinical concepts. Several data standards for adverse events currently exist, each with a strong user base. The structure and features of these current adverse event data standards (including terminologies and classifications) are different, so comparisons and evaluations are not straightforward, nor are strategies for their harmonization. Three different data standards - the Medical Dictionary for Regulatory Activities (MedDRA) and the Systematized Nomenclature of Medicine Clinical Terms (SNOMED CT) terminologies, and Common Terminology Criteria for Adverse Events (CTCAE) classification - are explored as candidate representations for AEs. This paper describes the structural features of each coding system, their content and relationship to the Unified Medical Language System (UMLS), and unsettled issues for future interoperability of these standards.

Combining lexical and semantic methods of inter-terminology mapping using the UMLS

The need for inter-terminology mapping is constantly increasing with the growth in the volume of electronically captured biomedical data and the demand to re-use the same data for secondary purposes. Using the UMLS as a knowledge base, semantically-based and lexically-based mappings were generated from SNOMED CT to ICD9CM terms and compared to a gold standard. Semantic mapping performed better than lexical mapping in terms of coverage, recall and precision. As the two mapping methods are orthogonal, the two sets of mappings can be used to validate and enhance each other. A method of combining the mappings based on the precision level of sub-categories in each method was derived. The combined method outperformed both methods, achieving coverage of 91%, recall of 43% and precision of 27%. It is also possible to customize the method of combination to optimize performance according to the task at hand.

Who is using the UMLS and how - insights from the UMLS user annual reports

The NLM's UMLS resources are available to users free of charge under a license that requires submission of an annual report on their usage. A new web-based template was used to collect users' annual reports for the calendar year 2004. Out of 2,677 li-censees, 1,427 (53%) submitted their annual reports through the web template. This represented a five-fold increase in the reports submitted compared to previous years. The information collected via the web template was more structured, more complete and easier to analyze. The main results from the 2004 annual reports are summarized and discussed. They are being used to guide UMLS developments.

Integrating SNOMED CT into the UMLS: an exploration of different views of synonymy and quality of editing

OBJECTIVE

The integration of SNOMED CT into the Unified Medical Language System (UMLS) involved the alignment of two views of synonymy that were different because the two vocabulary systems have different intended purposes and editing principles. The UMLS is organized according to one view of synonymy, but its structure also represents all the individual views of synonymy present in its source vocabularies. Despite progress in knowledge-based automation of development and maintenance of vocabularies, manual curation is still the main method of determining synonymy. The aim of this study was to investigate the quality of human judgment of synonymy.

DESIGN

Sixty pairs of potentially controversial SNOMED CT synonyms were reviewed by 11 domain vocabulary experts (six UMLS editors and five noneditors), and scores were assigned according to the degree of synonymy.

MEASUREMENTS

The synonymy scores of each subject were compared to the gold standard (the overall mean synonymy score of all subjects) to assess accuracy. Agreement between UMLS editors and noneditors was measured by comparing the mean synonymy scores of editors to noneditors.

RESULTS

Average accuracy was 71% for UMLS editors and 75% for noneditors (difference not statistically significant). Mean scores of editors and noneditors showed significant positive correlation (Spearman's rank correlation coefficient 0.654, two-tailed p < 0.01) with a concurrence rate of 75% and an interrater agreement kappa of 0.43.

CONCLUSION

The accuracy in the judgment of synonymy was comparable for UMLS editors and nonediting domain experts. There was reasonable agreement between the two groups.

Utilizing the UMLS for semantic mapping between terminologies

An algorithm was derived to find candidate mappings between any two terminologies inside the UMLS, making use of synonymy, explicit mapping relations and hierarchical relationships among UMLS concepts. Using an existing set of mappings from SNOMED CT to ICD9CM as our gold standard, we managed to find candidate mappings for 86% of SNOMED CT terms, with recall of 42% and precision of 20%. Among the various methods used, mapping by UMLS synonymy was particularly accurate and could potentially be useful as a quality assurance tool in the creation of mapping sets or in the UMLS editing process. Other strengths and weaknesses of the algorithm are discussed.

Will decision support in medications order entry save money? A return on investment analysis of the case of the Hong Kong hospital authority

The computerized medications order entry system currently used in the public hospitals of Hong Kong does not have decision support features. Plans are underway to add decision support to this system to alert physicians on drug-allergy conflicts, drug-lab result conflicts, drug-drug interactions and atypical dosages. A return on investment analysis is done on this enhancement, both as an examination of whether there is a positive return on the investment and as a contribution to the ongoing discussion of the use of return on investment models in health care information technology investments. It is estimated that the addition of decision support will reduce adverse drug events by 4.2 - 8.4%. Based on this estimate, a total net saving of $44,000 - $586,000 is expected over five years. The breakeven period is estimated to be between two to four years.

The impact of mastectomy, breast-conserving treatment and immediate breast reconstructions on the quality of life of Chinese women

BACKGROUND

The psychosocial impact of breast surgery has been extensively studied in the Western population. There is a relative paucity of comparable data in Oriental women who are increasingly affected by cancer of the breast. The present study investigates the effects that different types of breast surgery have on the quality of life of Chinese women.

METHODS

Forty-nine Chinese women with early breast cancer were interviewed at 6 months-2 years following their primary surgery (breast-conserving treatment (BCT, 17 patients), mastectomy (15 patients) and mastectomy with immediate breast reconstruction (17 patients)). Aspects of quality of life measured included general psychological well-being, body image, sexual functioning and social functioning.

RESULTS

Patients who received BCT had significantly better body image scores compared to mastectomy patients. They were less worried about their appearance, had more freedom in the choice of clothing, felt less upset by the change in their body and felt more accepted by their partners. The three groups did not differ significantly in the other aspects of quality of life measured.

CONCLUSIONS

Compared to mastectomy or mastectomy and immediate breast reconstruction, the most significant benefit of BCT is the preservation of a better body image.