Defining Disease Phenotypes in Primary Care Electronic Health Records by a Machine Learning Approach: A Case Study in Identifying Rheumatoid Arthritis

Management of Rheumatoid Arthritis in Primary Care: A Scoping Review

Rheumatoid arthritis (RA) can lead to severe joint impairment and chronic disability. Primary care (PC), provided by general practitioners (GPs), is the first level of contact for the population with the healthcare system. The aim of this scoping review was to analyze the approach to RA in the PC setting. PubMed, Scopus, and Web of Science were searched using the MESH terms "rheumatoid arthritis" and "primary care" from 2013 to 2023. The search strategy followed the PRISMA-ScR guidelines. The 61 articles selected were analyzed qualitatively in a table and discussed in two sections, namely criticisms and strategies for the management of RA in PC. The main critical issues in the management of RA in PC are the following: difficulty and delay in diagnosis, in accessing rheumatological care, and in using DMARDs by GPs; ineffective communication between GPs and specialists; poor patient education; lack of cardiovascular prevention; and increase in healthcare costs. To overcome these criticisms, several management strategies have been identified, namely early diagnosis of RA, quick access to rheumatology care, effective communication between GPs and specialists, active patient involvement, screening for risk factors and comorbidities, clinical audit, interdisciplinary patient management, digital health, and cost analysis. PC appears to be the ideal healthcare setting to reduce the morbidity and mortality of chronic disease, including RA, if a widespread change in GPs' approach to the disease and patients is mandatory.

A systematic review of clinical health conditions predicted by machine learning diagnostic and prognostic models trained or validated using real-world primary health care data

With the advances in technology and data science, machine learning (ML) is being rapidly adopted by the health care sector. However, there is a lack of literature addressing the health conditions targeted by the ML prediction models within primary health care (PHC) to date. To fill this gap in knowledge, we conducted a systematic review following the PRISMA guidelines to identify health conditions targeted by ML in PHC. We searched the Cochrane Library, Web of Science, PubMed, Elsevier, BioRxiv, Association of Computing Machinery (ACM), and IEEE Xplore databases for studies published from January 1990 to January 2022. We included primary studies addressing ML diagnostic or prognostic predictive models that were supplied completely or partially by real-world PHC data. Studies selection, data extraction, and risk of bias assessment using the prediction model study risk of bias assessment tool were performed by two investigators. Health conditions were categorized according to international classification of diseases (ICD-10). Extracted data were analyzed quantitatively. We identified 106 studies investigating 42 health conditions. These studies included 207 ML prediction models supplied by the PHC data of 24.2 million participants from 19 countries. We found that 92.4% of the studies were retrospective and 77.3% of the studies reported diagnostic predictive ML models. A majority (76.4%) of all the studies were for models' development without conducting external validation. Risk of bias assessment revealed that 90.8% of the studies were of high or unclear risk of bias. The most frequently reported health conditions were diabetes mellitus (19.8%) and Alzheimer's disease (11.3%). Our study provides a summary on the presently available ML prediction models within PHC. We draw the attention of digital health policy makers, ML models developer, and health care professionals for more future interdisciplinary research collaboration in this regard.

Predicting a diagnosis of ankylosing spondylitis using primary care health records-A machine learning approach

Ankylosing spondylitis is the second most common cause of inflammatory arthritis. However, a successful diagnosis can take a decade to confirm from symptom onset (via x-rays). The aim of this study was to use machine learning methods to develop a profile of the characteristics of people who are likely to be given a diagnosis of AS in future. The Secure Anonymised Information Linkage databank was used. Patients with ankylosing spondylitis were identified using their routine data and matched with controls who had no record of a diagnosis of ankylosing spondylitis or axial spondyloarthritis. Data was analysed separately for men and women. The model was developed using feature/variable selection and principal component analysis to develop decision trees. The decision tree with the highest average F value was selected and validated with a test dataset. The model for men indicated that lower back pain, uveitis, and NSAID use under age 20 is associated with AS development. The model for women showed an older age of symptom presentation compared to men with back pain and multiple pain relief medications. The models showed good prediction (positive predictive value 70%-80%) in test data but in the general population where prevalence is very low (0.09% of the population in this dataset) the positive predictive value would be very low (0.33%-0.25%). Machine learning can be used to help profile and understand the characteristics of people who will develop AS, and in test datasets with artificially high prevalence, will perform well. However, when applied to a general population with low prevalence rates, such as that in primary care, the positive predictive value for even the best model would be 1.4%. Multiple models may be needed to narrow down the population over time to improve the predictive value and therefore reduce the time to diagnosis of ankylosing spondylitis.

Predictive factors for degenerative lumbar spinal stenosis: a model obtained from a machine learning algorithm technique

BACKGROUND

Degenerative lumbar spinal stenosis (DLSS) is the most common spine disease in the elderly population. It is usually associated with lumbar spine joints/or ligaments degeneration. Machine learning technique is an exclusive method for handling big data analysis; however, the development of this method for spine pathology is rare. This study aims to detect the essential variables that predict the development of symptomatic DLSS using the random forest of machine learning (ML) algorithms technique.

METHODS

A retrospective study with two groups of individuals. The first included 165 with symptomatic DLSS (sex ratio 80 M/85F), and the second included 180 individuals from the general population (sex ratio: 90 M/90F) without lumbar spinal stenosis symptoms. Lumbar spine measurements such as vertebral or spinal canal diameters from L1 to S1 were conducted on computerized tomography (CT) images. Demographic and health data of all the participants (e.g., body mass index and diabetes mellitus) were also recorded.

RESULTS

The decision tree model of ML demonstrate that the anteroposterior diameter of the bony canal at L5 (males) and L4 (females) levels have the greatest stimulus for symptomatic DLSS (scores of 1 and 0.938). In addition, combination of these variables with other lumbar spine features is mandatory for developing the DLSS.

CONCLUSIONS

Our results indicate that combination of lumbar spine characteristics such as bony canal and vertebral body dimensions rather than the presence of a sole variable is highly associated with symptomatic DLSS onset.

Machine learning identification of thresholds to discriminate osteoarthritis and rheumatoid arthritis synovial inflammation

BACKGROUND

We sought to identify features that distinguish osteoarthritis (OA) and rheumatoid arthritis (RA) hematoxylin and eosin (H&E)-stained synovial tissue samples.

METHODS

We compared fourteen pathologist-scored histology features and computer vision-quantified cell density (147 OA and 60 RA patients) in H&E-stained synovial tissue samples from total knee replacement (TKR) explants. A random forest model was trained using disease state (OA vs RA) as a classifier and histology features and/or computer vision-quantified cell density as inputs.

RESULTS

Synovium from OA patients had increased mast cells and fibrosis (p < 0.001), while synovium from RA patients exhibited increased lymphocytic inflammation, lining hyperplasia, neutrophils, detritus, plasma cells, binucleate plasma cells, sub-lining giant cells, fibrin (all p < 0.001), Russell bodies (p = 0.019), and synovial lining giant cells (p = 0.003). Fourteen pathologist-scored features allowed for discrimination between OA and RA, producing a micro-averaged area under the receiver operating curve (micro-AUC) of 0.85±0.06. This discriminatory ability was comparable to that of computer vision cell density alone (micro-AUC = 0.87±0.04). Combining the pathologist scores with the cell density metric improved the discriminatory power of the model (micro-AUC = 0.92±0.06). The optimal cell density threshold to distinguish OA from RA synovium was 3400 cells/mm2, which yielded a sensitivity of 0.82 and specificity of 0.82.

CONCLUSIONS

H&E-stained images of TKR explant synovium can be correctly classified as OA or RA in 82% of samples. Cell density greater than 3400 cells/mm2 and the presence of mast cells and fibrosis are the most important features for making this distinction.

Diagnosing rheumatoid arthritis disease using fuzzy expert system and machine learning techniques

Rheumatoid Arthritis (RA) is a very common autoimmune disease that causes significant morbidity and mortality, and therefore early diagnosis and treatment are important. Early diagnosis of RA and knowing the severity of the disease are very important for the treatment to be applied. The diagnosis of RA usually requires a physical examination, laboratory tests, and a review of the patient’s medical history. In this study, the diagnosis of RA was made with two different methods using a fuzzy expert system (FES) and machine learning (ML) techniques, which were designed and implemented with the help of a specialist in the field, and the results were compared. For this purpose, blood counts were taken from 286 people, including 91 men and 195 women from various age groups. In the first method, an FES structure that determines the severity of RA disease has been established from blood count using the laboratory test results of CRP, ESR, RF, and ANA. The FES result that determines RA disease severity, the Anti-CCP level that is used to distinguish RA disease, and the patient’s medical history were used to design the Decision Support System (DSS) that diagnoses RA disease. The DSS is web-based and publicly accessible. In the second method, RA disease was diagnosed using kNN, SVM, LR, DT, NB, and MLP algorithms, which are widely used in machine learning. To examine the effect of the patient’s history on RA disease diagnosis, two different models were used in machine learning techniques, one with and one without the patient’s history. The results of the fuzzy-based DSS were also compared with the diagnoses made by the specialist and the diagnoses made according to the 2010 ACR / EULAR RA classification criteria. The performed DSS has achieved a diagnostic success rate of 94.05% on 286 patients. In the study of machine learning techniques, the highest success rate was achieved with the LR model. While the success rate of the model was 91.25 % with only blood count data, the success rate was 97.90% with the addition of the patient’s history. In addition to the high success rate, the results show that the patient’s history is important in diagnosing RA disease.

Artificial Intelligence in Rheumatoid Arthritis: Current Status and Future Perspectives: A State-of-the-Art Review

Investigation of the potential applications of artificial intelligence (AI), including machine learning (ML) and deep learning (DL) techniques, is an exponentially growing field in medicine and healthcare. These methods can be critical in providing high-quality care to patients with chronic rheumatological diseases lacking an optimal treatment, like rheumatoid arthritis (RA), which is the second most prevalent autoimmune disease. Herein, following reviewing the basic concepts of AI, we summarize the advances in its applications in RA clinical practice and research. We provide directions for future investigations in this field after reviewing the current knowledge gaps and technical and ethical challenges in applying AI. Automated models have been largely used to improve RA diagnosis since the early 2000s, and they have used a wide variety of techniques, e.g., support vector machine, random forest, and artificial neural networks. AI algorithms can facilitate screening and identification of susceptible groups, diagnosis using omics, imaging, clinical, and sensor data, patient detection within electronic health record (EHR), i.e., phenotyping, treatment response assessment, monitoring disease course, determining prognosis, novel drug discovery, and enhancing basic science research. They can also aid in risk assessment for incidence of comorbidities, e.g., cardiovascular diseases, in patients with RA. However, the proposed models may vary significantly in their performance and reliability. Despite the promising results achieved by AI models in enhancing early diagnosis and management of patients with RA, they are not fully ready to be incorporated into clinical practice. Future investigations are required to ensure development of reliable and generalizable algorithms while they carefully look for any potential source of bias or misconduct. We showed that a growing body of evidence supports the potential role of AI in revolutionizing screening, diagnosis, and management of patients with RA. However, multiple obstacles hinder clinical applications of AI models. Incorporating the machine and/or deep learning algorithms into real-world settings would be a key step in the progress of AI in medicine.

Unsupervised probabilistic models for sequential Electronic Health Records

We develop an unsupervised probabilistic model for heterogeneous Electronic Health Record (EHR) data. Utilizing a mixture model formulation, our approach directly models sequences of arbitrary length, such as medications and laboratory results. This allows for subgrouping and incorporation of the dynamics underlying heterogeneous data types. The model consists of a layered set of latent variables that encode underlying structure in the data. These variables represent subject subgroups at the top layer, and unobserved states for sequences in the second layer. We train this model on episodic data from subjects receiving medical care in the Kaiser Permanente Northern California integrated healthcare delivery system. The resulting properties of the trained model generate novel insight from these complex and multifaceted data. In addition, we show how the model can be used to analyze sequences that contribute to assessment of mortality likelihood.

Novel digital approaches to the assessment of problematic opioid use

The opioid epidemic continues to contribute to loss of life through overdose and significant social and economic burdens. Many individuals who develop problematic opioid use (POU) do so after being exposed to prescribed opioid analgesics. Therefore, it is important to accurately identify and classify risk factors for POU. In this review, we discuss the etiology of POU and highlight novel approaches to identifying its risk factors. These approaches include the application of polygenic risk scores (PRS) and diverse machine learning (ML) algorithms used in tandem with data from electronic health records (EHR), clinical notes, patient demographics, and digital footprints. The implementation and synergy of these types of data and approaches can greatly assist in reducing the incidence of POU and opioid-related mortality by increasing the knowledge base of patient-related risk factors, which can help to improve prescribing practices for opioid analgesics.

Big data analyses and individual health profiling in the arena of rheumatic and musculoskeletal diseases (RMDs)

Health care processes are under constant development and will need to embrace advances in technology and health science aiming to provide optimal care. Considering the perspective of increasing treatment options for people with rheumatic and musculoskeletal diseases, but in many cases not reaching all treatment targets that matter to patients, care systems bare potential to improve on a holistic level. This review provides an overview of systems and technologies under evaluation over the past years that show potential to impact diagnosis and treatment of rheumatic diseases in about 10 years from now. We summarize initiatives and studies from the field of electronic health records, biobanking, remote monitoring, and artificial intelligence. The combination and implementation of these opportunities in daily clinical care will be key for a new era in care of our patients. This aims to inform rheumatologists and healthcare providers concerned with chronic inflammatory musculoskeletal conditions about current important and promising developments in science that might substantially impact the management processes of rheumatic diseases in the 2030s.

Clinical predictors of response to methotrexate in patients with rheumatoid arthritis: a machine learning approach using clinical trial data

BACKGROUND

Methotrexate is the preferred initial disease-modifying antirheumatic drug (DMARD) for rheumatoid arthritis (RA). However, clinically useful tools for individualized prediction of response to methotrexate treatment in patients with RA are lacking. We aimed to identify clinical predictors of response to methotrexate in patients with rheumatoid arthritis (RA) using machine learning methods.

METHODS

Randomized clinical trials (RCT) of patients with RA who were DMARD-naïve and randomized to placebo plus methotrexate were identified and accessed through the Clinical Study Data Request Consortium and Vivli Center for Global Clinical Research Data. Studies with available Disease Activity Score with 28-joint count and erythrocyte sedimentation rate (DAS28-ESR) at baseline and 12 and 24 weeks were included. Latent class modeling of methotrexate response was performed. The least absolute shrinkage and selection operator (LASSO) and random forests methods were used to identify predictors of response.

RESULTS

A total of 775 patients from 4 RCTs were included (mean age 50 years, 80% female). Two distinct classes of patients were identified based on DAS28-ESR change over 24 weeks: "good responders" and "poor responders." Baseline DAS28-ESR, anti-citrullinated protein antibody (ACPA), and Health Assessment Questionnaire (HAQ) score were the top predictors of good response using LASSO (area under the curve [AUC] 0.79) and random forests (AUC 0.68) in the external validation set. DAS28-ESR ≤ 7.4, ACPA positive, and HAQ ≤ 2 provided the highest likelihood of response. Among patients with 12-week DAS28-ESR > 3.2, ≥ 1 point improvement in DAS28-ESR baseline-to-12-week was predictive of achieving DAS28-ESR ≤ 3.2 at 24 weeks.

CONCLUSIONS

We have developed and externally validated a prediction model for response to methotrexate within 24 weeks in DMARD-naïve patients with RA, providing variably weighted clinical features and defined cutoffs for clinical decision-making.

Predicting Hospital Readmission for Campylobacteriosis from Electronic Health Records: A Machine Learning and Text Mining Perspective

(1) Background: This study investigates influential risk factors for predicting 30-day readmission to hospital for Campylobacter infections (CI). (2) Methods: We linked general practitioner and hospital admission records of 13,006 patients with CI in Wales (1990–2015). An approach called TF-zR (term frequency-zRelevance) technique was presented to evaluates how relevant a clinical term is to a patient in a cohort characterized by coded health records. The zR is a supervised term-weighting metric to assign weight to a term based on relative frequencies of the term across different classes. Cost-sensitive classifier with swarm optimization and weighted subset learning was integrated to identify influential clinical signals as predictors and optimal model for readmission prediction. (3) Results: From a pool of up to 17,506 variables, 33 most predictive factors were identified, including age, gender, Townsend deprivation quintiles, comorbidities, medications, and procedures. The predictive model predicted readmission with 73% sensitivity and 54% specificity. Variables associated with readmission included male gender, recurrent tonsillitis, non-healing open wounds, operation for in-gown toenails. Cystitis, paracetamol/codeine use, age (21–25), and heliclear triple pack use, were associated with a lower risk of readmission. (4) Conclusions: This study gives a profile of clustered variables that are predictive of readmission associated with campylobacteriosis.

Machine learning approaches to improve disease management of patients with rheumatoid arthritis: review and future directions

INTRODUCTION

Although the management of rheumatoid arthritis (RA) has improved in major way over the last decades, this disease still leads to an important burden for patients and society, and there is a need to develop more personalized approaches. Machine learning (ML) methods are more and more used in health-related studies and can be applied to different sorts of data (clinical, radiological, or 'omics' data). Such approaches may improve the management of patients with RA.

AREAS COVERED

In this paper, we propose a review regarding ML approaches applied to RA. A scoping literature search was performed in PubMed, in September 2021 using the following MeSH terms: 'arthritis, rheumatoid' and 'machine learning'. Based on this search, the usefulness of ML methods for RA diagnosis, monitoring, and prediction of response to treatment and RA outcomes, is discussed.

EXPERT OPINION

ML methods have the potential to revolutionize RA-related research and improve disease management and patient care. Nevertheless, these models are not yet ready to contribute fully to rheumatologists' daily practice. Indeed, these methods raise technical, methodological, and ethical issues, which should be addressed properly to allow their implementation. Collaboration between data scientists, clinical researchers, and physicians is therefore required to move this field forward.

Identification of a Rule to Predict Response to Sarilumab in Patients with Rheumatoid Arthritis Using Machine Learning and Clinical Trial Data

INTRODUCTION

In rheumatoid arthritis, time spent using ineffective medications may lead to irreversible disease progression. Despite availability of targeted treatments, only a minority of patients achieve sustained remission, and little evidence exists to direct the choice of biologic disease-modifying antirheumatic drugs in individual patients. Machine learning was used to identify a rule to predict the response to sarilumab and discriminate between responses to sarilumab versus adalimumab, with a focus on clinically feasible blood biomarkers.

METHODS

The decision tree model GUIDE was trained using a data subset from the sarilumab trial with the most biomarker data, MOBILITY, to identify a rule to predict disease activity after sarilumab 200 mg. The training set comprised 18 categorical and 24 continuous baseline variables; some data were omitted from training and used for validation by the algorithm (cross-validation). The rule was tested using full datasets from four trials (MOBILITY, MONARCH, TARGET, and ASCERTAIN), focusing on the recommended sarilumab dose of 200 mg.

RESULTS

In the training set, the presence of anti-cyclic citrullinated peptide antibodies, combined with C-reactive protein > 12.3 mg/l, was identified as the "rule" that predicts American College of Rheumatology 20% response (ACR20) to sarilumab. In testing, the rule reliably predicted response to sarilumab in MOBILITY, MONARCH, and ASCERTAIN for many efficacy parameters (e.g., ACR70 and the 28-joint disease activity score using CRP [DAS28-CRP] remission). The rule applied less to TARGET, which recruited individuals refractory to tumor necrosis factor inhibitors. The potential clinical benefit of the rule was highlighted in a clinical scenario based on MONARCH data, which found that increased ACR70 rates could be achieved by treating either rule-positive patients with sarilumab or rule-negative patients with adalimumab.

CONCLUSIONS

Well-established and clinically feasible blood biomarkers can guide individual treatment choice. Real-world validation of the rule identified in this post hoc analysis is merited.

CLINICAL TRIAL REGISTRATION

NCT01061736, NCT02332590, NCT01709578, NCT01768572.

Machine Learning in Rheumatic Diseases

With advances in information technology, the demand for using data science to enhance healthcare and disease management is rapidly increasing. Among these technologies, machine learning (ML) has become ubiquitous and indispensable for solving complex problems in many scientific fields, including medical science. ML allows the development of guidelines and framing of the evaluation system for complex diseases based on massive data. In the analysis of rheumatic diseases, which are chronic and remarkably heterogeneous, ML can be anticipated to be extremely helpful in deciphering and revealing the inherent interrelationships in disease development and progression, which can further enhance the overall understanding of the disease, optimize patients' stratification, calibrate therapeutic strategies, and predict prognosis and outcomes. In this review, the basics of ML, its potential clinical applications in rheumatology, together with its strengths and limitations are summarized.

Health Equity in Artificial Intelligence and Primary Care Research: Protocol for a Scoping Review

BACKGROUND

Though artificial intelligence (AI) has the potential to augment the patient-physician relationship in primary care, bias in intelligent health care systems has the potential to differentially impact vulnerable patient populations.

OBJECTIVE

The purpose of this scoping review is to summarize the extent to which AI systems in primary care examine the inherent bias toward or against vulnerable populations and appraise how these systems have mitigated the impact of such biases during their development.

METHODS

We will conduct a search update from an existing scoping review to identify studies on AI and primary care in the following databases: Medline-OVID, Embase, CINAHL, Cochrane Library, Web of Science, Scopus, IEEE Xplore, ACM Digital Library, MathSciNet, AAAI, and arXiv. Two screeners will independently review all abstracts, titles, and full-text articles. The team will extract data using a structured data extraction form and synthesize the results in accordance with PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses Extension for Scoping Reviews) guidelines.

RESULTS

This review will provide an assessment of the current state of health care equity within AI for primary care. Specifically, we will identify the degree to which vulnerable patients have been included, assess how bias is interpreted and documented, and understand the extent to which harmful biases are addressed. As of October 2020, the scoping review is in the title- and abstract-screening stage. The results are expected to be submitted for publication in fall 2021.

CONCLUSIONS

AI applications in primary care are becoming an increasingly common tool in health care delivery and in preventative care efforts for underserved populations. This scoping review would potentially show the extent to which studies on AI in primary care employ a health equity lens and take steps to mitigate bias.

INTERNATIONAL REGISTERED REPORT IDENTIFIER (IRRID)

PRR1-10.2196/27799.

Application of Artificial Intelligence in Community-Based Primary Health Care: Systematic Scoping Review and Critical Appraisal

BACKGROUND

Research on the integration of artificial intelligence (AI) into community-based primary health care (CBPHC) has highlighted several advantages and disadvantages in practice regarding, for example, facilitating diagnosis and disease management, as well as doubts concerning the unintended harmful effects of this integration. However, there is a lack of evidence about a comprehensive knowledge synthesis that could shed light on AI systems tested or implemented in CBPHC.

OBJECTIVE

We intended to identify and evaluate published studies that have tested or implemented AI in CBPHC settings.

METHODS

We conducted a systematic scoping review informed by an earlier study and the Joanna Briggs Institute (JBI) scoping review framework and reported the findings according to PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analysis-Scoping Reviews) reporting guidelines. An information specialist performed a comprehensive search from the date of inception until February 2020, in seven bibliographic databases: Cochrane Library, MEDLINE, EMBASE, Web of Science, Cumulative Index to Nursing and Allied Health Literature (CINAHL), ScienceDirect, and IEEE Xplore. The selected studies considered all populations who provide and receive care in CBPHC settings, AI interventions that had been implemented, tested, or both, and assessed outcomes related to patients, health care providers, or CBPHC systems. Risk of bias was assessed using the Prediction Model Risk of Bias Assessment Tool (PROBAST). Two authors independently screened the titles and abstracts of the identified records, read the selected full texts, and extracted data from the included studies using a validated extraction form. Disagreements were resolved by consensus, and if this was not possible, the opinion of a third reviewer was sought. A third reviewer also validated all the extracted data.

RESULTS

We retrieved 22,113 documents. After the removal of duplicates, 16,870 documents were screened, and 90 peer-reviewed publications met our inclusion criteria. Machine learning (ML) (41/90, 45%), natural language processing (NLP) (24/90, 27%), and expert systems (17/90, 19%) were the most commonly studied AI interventions. These were primarily implemented for diagnosis, detection, or surveillance purposes. Neural networks (ie, convolutional neural networks and abductive networks) demonstrated the highest accuracy, considering the given database for the given clinical task. The risk of bias in diagnosis or prognosis studies was the lowest in the participant category (4/49, 4%) and the highest in the outcome category (22/49, 45%).

CONCLUSIONS

We observed variabilities in reporting the participants, types of AI methods, analyses, and outcomes, and highlighted the large gap in the effective development and implementation of AI in CBPHC. Further studies are needed to efficiently guide the development and implementation of AI interventions in CBPHC settings.

Electronic Medical Record-Based Case Phenotyping for the Charlson Conditions: Scoping Review

Background

Electronic medical records (EMRs) contain large amounts of rich clinical information. Developing EMR-based case definitions, also known as EMR phenotyping, is an active area of research that has implications for epidemiology, clinical care, and health services research.

Objective

This review aims to describe and assess the present landscape of EMR-based case phenotyping for the Charlson conditions.

Methods

A scoping review of EMR-based algorithms for defining the Charlson comorbidity index conditions was completed. This study covered articles published between January 2000 and April 2020, both inclusive. Embase (Excerpta Medica database) and MEDLINE (Medical Literature Analysis and Retrieval System Online) were searched using keywords developed in the following 3 domains: terms related to EMR, terms related to case finding, and disease-specific terms. The manuscript follows the Preferred Reporting Items for Systematic reviews and Meta-analyses extension for Scoping Reviews (PRISMA) guidelines.

Results

A total of 274 articles representing 299 algorithms were assessed and summarized. Most studies were undertaken in the United States (181/299, 60.5%), followed by the United Kingdom (42/299, 14.0%) and Canada (15/299, 5.0%). These algorithms were mostly developed either in primary care (103/299, 34.4%) or inpatient (168/299, 56.2%) settings. Diabetes, congestive heart failure, myocardial infarction, and rheumatology had the highest number of developed algorithms. Data-driven and clinical rule–based approaches have been identified. EMR-based phenotype and algorithm development reflect the data access allowed by respective health systems, and algorithms vary in their performance.

Conclusions

Recognizing similarities and differences in health systems, data collection strategies, extraction, data release protocols, and existing clinical pathways is critical to algorithm development strategies. Several strategies to assist with phenotype-based case definitions have been proposed.

Identifying chronic disease patients using predictive algorithms in pharmacy administrative claims: an application in rheumatoid arthritis

OBJECTIVE

To evaluate the predictive performance of logistic and linear regression versus machine learning (ML) algorithms to identify patients with rheumatoid arthritis (RA) treated with target immunomodulators (TIMs) using only pharmacy administrative claims.

METHODS

Adults aged 18-64 years with ≥1 TIM claim in the IBM MarketScan commercial database were included in this retrospective analysis. The predictive ability of logistic regression to identify RA patients was compared with supervised ML classification algorithms including random forest (RF), decision trees, linear support vector machines (SVMs), neural networks, naïve Bayes classifier, linear discriminant analysis (LDA), quadratic discriminant analysis (QDA), and K-nearest neighbors (k-NN). Model performance was evaluated using F1 score, accuracy, precision, sensitivity, area under the receiver operating characteristic curve (AUROC), and Matthews correlation coefficient (MCC). Analyses were conducted in all-patient and etanercept-only samples.

RESULTS

In the all-patients sample, ML approaches did not outperform logistic regression. RF showed small improvements versus logistic regression that were not considered remarkable, respectively: F1 score (84.55% vs 83.96%), accuracy (84.05% vs 83.79%), sensitivity (84.53% vs 82.20%), AUROC (84.04% vs 83.85%), and MCC (68.07% vs 67.66%). Findings were similar in the etanercept samples.

CONCLUSION

Logistic regression and ML approaches successfully identified patients with RA in a large pharmacy administrative claims database. The ML algorithms were no better than logistic regression at prediction. RF, SVMs, LDA, and ridge classifier showed comparable performance, while neural networks, decision trees, naïve Bayes classifier, and QDA underperformed compared with logistic regression in identifying patients with RA.

The 2019 National Natural language processing (NLP) Clinical Challenges (n2c2)/Open Health NLP (OHNLP) shared task on clinical concept normalization for clinical records

OBJECTIVE

The 2019 National Natural language processing (NLP) Clinical Challenges (n2c2)/Open Health NLP (OHNLP) shared task track 3, focused on medical concept normalization (MCN) in clinical records. This track aimed to assess the state of the art in identifying and matching salient medical concepts to a controlled vocabulary. In this paper, we describe the task, describe the data set used, compare the participating systems, present results, identify the strengths and limitations of the current state of the art, and identify directions for future research.

MATERIALS AND METHODS

Participating teams were provided with narrative discharge summaries in which text spans corresponding to medical concepts were identified. This paper refers to these text spans as mentions. Teams were tasked with normalizing these mentions to concepts, represented by concept unique identifiers, within the Unified Medical Language System. Submitted systems represented 4 broad categories of approaches: cascading dictionary matching, cosine distance, deep learning, and retrieve-and-rank systems. Disambiguation modules were common across all approaches.

RESULTS

A total of 33 teams participated in the MCN task. The best-performing team achieved an accuracy of 0.8526. The median and mean performances among all teams were 0.7733 and 0.7426, respectively.

CONCLUSIONS

Overall performance among the top 10 teams was high. However, several mention types were challenging for all teams. These included mentions requiring disambiguation of misspelled words, acronyms, abbreviations, and mentions with more than 1 possible semantic type. Also challenging were complex mentions of long, multi-word terms that may require new ways of extracting and representing mention meaning, the use of domain knowledge, parse trees, or hand-crafted rules.

Cardiovascular risk assessment in patients with rheumatoid arthritis using carotid ultrasound B-mode imaging

Rheumatoid arthritis (RA) is a systemic chronic inflammatory disease that affects synovial joints and has various extra-articular manifestations, including atherosclerotic cardiovascular disease (CVD). Patients with RA experience a higher risk of CVD, leading to increased morbidity and mortality. Inflammation is a common phenomenon in RA and CVD. The pathophysiological association between these diseases is still not clear, and, thus, the risk assessment and detection of CVD in such patients is of clinical importance. Recently, artificial intelligence (AI) has gained prominence in advancing healthcare and, therefore, may further help to investigate the RA-CVD association. There are three aims of this review: (1) to summarize the three pathophysiological pathways that link RA to CVD; (2) to identify several traditional and carotid ultrasound image-based CVD risk calculators useful for RA patients, and (3) to understand the role of artificial intelligence in CVD risk assessment in RA patients. Our search strategy involves extensively searches in PubMed and Web of Science databases using search terms associated with CVD risk assessment in RA patients. A total of 120 peer-reviewed articles were screened for this review. We conclude that (a) two of the three pathways directly affect the atherosclerotic process, leading to heart injury, (b) carotid ultrasound image-based calculators have shown superior performance compared with conventional calculators, and (c) AI-based technologies in CVD risk assessment in RA patients are aggressively being adapted for routine practice of RA patients.

Charting the life course: Emerging opportunities to advance scientific approaches using life course research

Life course research embraces the complexity of health and disease development, tackling the extensive interactions between genetics and environment. This interdisciplinary blueprint, or theoretical framework, offers a structure for research ideas and specifies relationships between related factors. Traditionally, methodological approaches attempt to reduce the complexity of these dynamic interactions and decompose health into component parts, ignoring the complex reciprocal interaction of factors that shape health over time. New methods that match the epistemological foundation of the life course framework are needed to fully explore adaptive, multilevel, and reciprocal interactions between individuals and their environment. The focus of this article is to (1) delineate the differences between lifespan and life course research, (2) articulate the importance of complex systems science as a methodological framework in the life course research toolbox to guide our research questions, (3) raise key questions that can be asked within the clinical and translational science domain utilizing this framework, and (4) provide recommendations for life course research implementation, charting the way forward. Recent advances in computational analytics, computer science, and data collection could be used to approximate, measure, and analyze the intertwining and dynamic nature of genetic and environmental factors involved in health development.

A national initiative in data science for health: an evaluation of the UK Farr Institute

OBJECTIVE

To evaluate the extent to which the inter-institutional, inter-disciplinary mobilisation of data and skills in the Farr Institute contributed to establishing the emerging field of data science for health in the UK.

DESIGN AND OUTCOME MEASURES

We evaluated evidence of six domains characterising a new field of science:defining central scientific challenges,demonstrating how the central challenges might be solved,creating novel interactions among groups of scientists,training new types of experts,re-organising universities,demonstrating impacts in society.We carried out citation, network and time trend analyses of publications, and a narrative review of infrastructure, methods and tools.

SETTING

Four UK centres in London, North England, Scotland and Wales (23 university partners), 2013-2018.

RESULTS

1. The Farr Institute helped define a central scientific challenge publishing a research corpus, demonstrating insights from electronic health record (EHR) and administrative data at each stage of the translational cycle in 593 papers with at least one Farr Institute author affiliation on PubMed. 2. The Farr Institute offered some demonstrations of how these scientific challenges might be solved: it established the first four ISO27001 certified trusted research environments in the UK, and approved more than 1000 research users, published on 102 unique EHR and administrative data sources, although there was no clear evidence of an increase in novel, sustained record linkages. The Farr Institute established open platforms for the EHR phenotyping algorithms and validations (>70 diseases, CALIBER). Sample sizes showed some evidence of increase but remained less than 10% of the UK population in primary care-hospital care linked studies. 3.The Farr Institute created novel interactions among researchers: the co-author publication network expanded from 944 unique co-authors (based on 67 publications in the first 30 months) to 3839 unique co-authors (545 papers in the final 30 months). 4. Training expanded substantially with 3 new masters courses, training >400 people at masters, short-course and leadership level and 48 PhD students. 5. Universities reorganised with 4/5 Centres established 27 new faculty (tenured) positions, 3 new university institutes. 6. Emerging evidence of impacts included: > 3200 citations for the 10 most cited papers and Farr research informed eight practice-changing clinical guidelines and policies relevant to the health of millions of UK citizens.

CONCLUSION

The Farr Institute played a major role in establishing and growing the field of data science for health in the UK, with some initial evidence of benefits for health and healthcare. The Farr Institute has now expanded into Health Data Research (HDR) UK but key challenges remain including, how to network such activities internationally.

Electronic health records and polygenic risk scores for predicting disease risk

Accurate prediction of disease risk based on the genetic make-up of an individual is essential for effective prevention and personalized treatment. Nevertheless, to date, individual genetic variants from genome-wide association studies have achieved only moderate prediction of disease risk. The aggregation of genetic variants under a polygenic model shows promising improvements in prediction accuracies. Increasingly, electronic health records (EHRs) are being linked to patient genetic data in biobanks, which provides new opportunities for developing and applying polygenic risk scores in the clinic, to systematically examine and evaluate patient susceptibilities to disease. However, the heterogeneous nature of EHR data brings forth many practical challenges along every step of designing and implementing risk prediction strategies. In this Review, we present the unique considerations for using genotype and phenotype data from biobank-linked EHRs for polygenic risk prediction.

A systematic review of the applications of artificial intelligence and machine learning in autoimmune diseases

Autoimmune diseases are chronic, multifactorial conditions. Through machine learning (ML), a branch of the wider field of artificial intelligence, it is possible to extract patterns within patient data, and exploit these patterns to predict patient outcomes for improved clinical management. Here, we surveyed the use of ML methods to address clinical problems in autoimmune disease. A systematic review was conducted using MEDLINE, embase and computers and applied sciences complete databases. Relevant papers included "machine learning" or "artificial intelligence" and the autoimmune diseases search term(s) in their title, abstract or key words. Exclusion criteria: studies not written in English, no real human patient data included, publication prior to 2001, studies that were not peer reviewed, non-autoimmune disease comorbidity research and review papers. 169 (of 702) studies met the criteria for inclusion. Support vector machines and random forests were the most popular ML methods used. ML models using data on multiple sclerosis, rheumatoid arthritis and inflammatory bowel disease were most common. A small proportion of studies (7.7% or 13/169) combined different data types in the modelling process. Cross-validation, combined with a separate testing set for more robust model evaluation occurred in 8.3% of papers (14/169). The field may benefit from adopting a best practice of validation, cross-validation and independent testing of ML models. Many models achieved good predictive results in simple scenarios (e.g. classification of cases and controls). Progression to more complex predictive models may be achievable in future through integration of multiple data types.

Applied machine learning and artificial intelligence in rheumatology

Machine learning as a field of artificial intelligence is increasingly applied in medicine to assist patients and physicians. Growing datasets provide a sound basis with which to apply machine learning methods that learn from previous experiences. This review explains the basics of machine learning and its subfields of supervised learning, unsupervised learning, reinforcement learning and deep learning. We provide an overview of current machine learning applications in rheumatology, mainly supervised learning methods for e-diagnosis, disease detection and medical image analysis. In the future, machine learning will be likely to assist rheumatologists in predicting the course of the disease and identifying important disease factors. Even more interestingly, machine learning will probably be able to make treatment propositions and estimate their expected benefit (e.g. by reinforcement learning). Thus, in future, shared decision-making will not only include the patient’s opinion and the rheumatologist’s empirical and evidence-based experience, but it will also be influenced by machine-learned evidence.

Preliminary Results on Pulmonary Tuberculosis Detection in Chest X-Ray Using Convolutional Neural Networks

Tuberculosis (TB), is an ancient disease that probably affects humans since pre-hominids. This disease is caused by bacteria belonging to the mycobacterium tuberculosis complex and usually affects the lungs in up to 67% of cases. In 2019, there were estimated to be over 10 million tuberculosis cases in the world, in the same year TB was between the ten leading causes of death, and the deadliest from a single infectious agent. Chest X-ray (CXR) has recently been promoted by the WHO as a tool possibly placed early in screening and triaging algorithms for TB detection. Numerous TB prevalence surveys have demonstrated that CXR is the most sensitive screening tool for pulmonary TB and that a significant proportion of people with TB are asymptomatic in the early stages of the disease. This study presents experimentation of classic convolutional neural network architectures on public CRX databases in order to create a tool applied to the diagnostic aid of TB in chest X-ray images. As result the study has an AUC ranging from 0.78 to 0.84, sensitivity from 0.76 to 0.86 and specificity from 0.58 to 0.74 depending on the network architecture. The observed performance by these metrics alone are within the range of metrics found in the literature, although there is much room for metrics improvement and bias avoiding. Also, the usage of the model in a triage use-case could be used to validate the efficiency of the model in the future.

Identification of validated case definitions for medical conditions used in primary care electronic medical record databases: a systematic review

Objectives

Data derived from primary care electronic medical records (EMRs) are being used for research and surveillance. Case definitions are required to identify patients with specific conditions in EMR data with a degree of accuracy. The purpose of this study is to identify and provide a summary of case definitions that have been validated in primary care EMR data.

Materials and Methods

We searched MEDLINE and Embase (from inception to June 2016) to identify studies that describe case definitions for clinical conditions in EMR data and report on the performance metrics of these definitions.

Results

We identified 40 studies reporting on case definitions for 47 unique clinical conditions. The studies used combinations of International Classification of Disease version 9 (ICD-9) codes, Read codes, laboratory values, and medications in their algorithms. The most common validation metric reported was positive predictive value, with inconsistent reporting of sensitivity and specificity.

Discussion

This review describes validated case definitions derived in primary care EMR data, which can be used to understand disease patterns and prevalence among primary care populations. Limitations include incomplete reporting of performance metrics and uncertainty regarding performance of case definitions across different EMR databases and countries.

Conclusion

Our review found a significant number of validated case definitions with good performance for use in primary care EMR data. These could be applied to other EMR databases in similar contexts and may enable better disease surveillance when using clinical EMR data. Consistent reporting across validation studies using EMR data would facilitate comparison across studies.

Systematic review registration

PROSPERO CRD42016040020 (submitted June 8, 2016, and last revised June 14, 2016).

UK phenomics platform for developing and validating electronic health record phenotypes: CALIBER

OBJECTIVE

Electronic health records (EHRs) are a rich source of information on human diseases, but the information is variably structured, fragmented, curated using different coding systems, and collected for purposes other than medical research. We describe an approach for developing, validating, and sharing reproducible phenotypes from national structured EHR in the United Kingdom with applications for translational research.

MATERIALS AND METHODS

We implemented a rule-based phenotyping framework, with up to 6 approaches of validation. We applied our framework to a sample of 15 million individuals in a national EHR data source (population-based primary care, all ages) linked to hospitalization and death records in England. Data comprised continuous measurements (for example, blood pressure; medication information; coded diagnoses, symptoms, procedures, and referrals), recorded using 5 controlled clinical terminologies: (1) read (primary care, subset of SNOMED-CT [Systematized Nomenclature of Medicine Clinical Terms]), (2) International Classification of Diseases-Ninth Revision and Tenth Revision (secondary care diagnoses and cause of mortality), (3) Office of Population Censuses and Surveys Classification of Surgical Operations and Procedures, Fourth Revision (hospital surgical procedures), and (4) DM+D prescription codes.

RESULTS

Using the CALIBER phenotyping framework, we created algorithms for 51 diseases, syndromes, biomarkers, and lifestyle risk factors and provide up to 6 validation approaches. The EHR phenotypes are curated in the open-access CALIBER Portal (https://www.caliberresearch.org/portal) and have been used by 40 national and international research groups in 60 peer-reviewed publications.

CONCLUSIONS

We describe a UK EHR phenomics approach within the CALIBER EHR data platform with initial evidence of validity and use, as an important step toward international use of UK EHR data for health research.

A Review of Automatic Phenotyping Approaches using Electronic Health Records

Electronic Health Records (EHR) are a rich repository of valuable clinical information that exist in primary and secondary care databases. In order to utilize EHRs for medical observational research a range of algorithms for automatically identifying individuals with a specific phenotype have been developed. This review summarizes and offers a critical evaluation of the literature relating to studies conducted into the development of EHR phenotyping systems. This review describes phenotyping systems and techniques based on structured and unstructured EHR data. Articles published on PubMed and Google scholar between 2013 and 2017 have been reviewed, using search terms derived from Medical Subject Headings (MeSH). The popularity of using Natural Language Processing (NLP) techniques in extracting features from narrative text has increased. This increased attention is due to the availability of open source NLP algorithms, combined with accuracy improvement. In this review, Concept extraction is the most popular NLP technique since it has been used by more than 50% of the reviewed papers to extract features from EHR. High-throughput phenotyping systems using unsupervised machine learning techniques have gained more popularity due to their ability to efficiently and automatically extract a phenotype with minimal human effort.

Sustainable Health Informatics: Health Informaticians as Alchemists

The digital transformation of health care delivery remains an elusive work in progress. Contextual variation continues to be a significant barrier to the development of sustainable health information systems. In this paper we characterize health informaticians as modern alchemists and use this characterization to describe informatics progress in addressing four key healthcare challenges. We highlight the need for informaticians to be diligent and loyal to basic methodological principles while also appreciating the role that contextual variation plays in informatics research. We also emphasize that meaningful health systems transformation takes time. The insight presented in this paper helps informaticians in our quest to develop sustainable health information systems.

Application of machine learning in the diagnosis of axial spondyloarthritis

PURPOSE OF REVIEW

In this review article, we describe the development and application of machine-learning models in the field of rheumatology to improve the detection and diagnosis rates of underdiagnosed rheumatologic conditions, such as ankylosing spondylitis and axial spondyloarthritis (axSpA).

RECENT FINDINGS

In an attempt to aid in the earlier diagnosis of axSpA, we developed machine-learning models to predict a diagnosis of ankylosing spondylitis and axSpA using administrative claims and electronic medical record data. Machine-learning algorithms based on medical claims data predicted the diagnosis of ankylosing spondylitis better than a model developed based on clinical characteristics of ankylosing spondylitis. With additional clinical data, machine-learning algorithms developed using electronic medical records identified patients with axSpA with 82.6-91.8% accuracy. These two algorithms have helped us understand potential opportunities and challenges associated with each data set and with different analytic approaches. Efforts to refine and validate these machine-learning models are ongoing.

SUMMARY

We discuss the challenges and benefits of machine-learning models in healthcare, along with potential opportunities for its application in the field of rheumatology, particularly in the early diagnosis of axSpA and ankylosing spondylitis.

Period of Measurement in Time-Series Predictions of Disease Counts from 2007 to 2017 in Northern Nevada: Analytics Experiment

BACKGROUND

The literature in statistics presents methods by which autocorrelation can identify the best period of measurement to improve the performance of a time-series prediction. The period of measurement plays an important role in improving the performance of disease-count predictions. However, from the operational perspective in public health surveillance, there is a limitation to the length of the measurement period that can offer meaningful and valuable predictions.

OBJECTIVE

This study aimed to establish a method that identifies the shortest period of measurement without significantly decreasing the prediction performance for time-series analysis of disease counts.

METHODS

The data used in this evaluation include disease counts from 2007 to 2017 in northern Nevada. The disease counts for chlamydia, salmonella, respiratory syncytial virus, gonorrhea, viral meningitis, and influenza A were predicted.

RESULTS

Our results showed that autocorrelation could not guarantee the best performance for prediction of disease counts. However, the proposed method with the change-point analysis suggests a period of measurement that is operationally acceptable and performance that is not significantly different from the best prediction.

CONCLUSIONS

The use of change-point analysis with autocorrelation provides the best and most practical period of measurement.

Primary Care Patient Records in the United Kingdom: Past, Present, and Future Research Priorities

This paper briefly outlines the history of the medical record and the factors contributing to the adoption of computerized records in primary care in the United Kingdom. It discusses how both paper-based and electronic health records have traditionally been used in the past and goes on to examine how enabling patients to access their own primary care record online is changing the form and function of the patient record. In addition, it looks at the evidence for the benefits of Web-based access and discusses some of the challenges faced in this transition. Finally, some suggestions are made regarding the future of the patient record and research questions that need to be addressed to help deepen our understanding of how they can be used more beneficially by both patients and clinicians.

Predicting Hospital Readmission via Cost-Sensitive Deep Learning

With increased use of electronic medical records (EMRs), data mining on medical data has great potential to improve the quality of hospital treatment and increase the survival rate of patients. Early readmission prediction enables early intervention, which is essential to preventing serious or life-threatening events, and act as a substantial contributor to reduce healthcare costs. Existing works on predicting readmission often focus on certain vital signs and diseases by extracting statistical features. They also fail to consider skewness of class labels in medical data and different costs of misclassification errors. In this paper, we recur to the merits of convolutional neural networks (CNN) to automatically learn features from time series of vital sign, and categorical feature embedding to effectively encode feature vectors with heterogeneous clinical features, such as demographics, hospitalization history, vital signs, and laboratory tests. Then, both learnt features via CNN and statistical features via feature embedding are fed into a multilayer perceptron (MLP) for prediction. We use a cost-sensitive formulation to train MLP during prediction to tackle the imbalance and skewness challenge. We validate the proposed approach on two real medical datasets from Barnes-Jewish Hospital, and all data is taken from historical EMR databases and reflects the kinds of data that would realistically be available at the clinical prediction system in hospitals. We find that early prediction of readmission is possible and when compared with state-of-the-art existing methods used by hospitals, our methods perform significantly better. For example, using the general hospital wards data for 30-day readmission prediction, the area under the curve (AUC) for the proposed model was 0.70, significantly higher than all the baseline methods. Based on these results, a system is being deployed in hospital settings with the proposed forecasting algorithms to support treatment.

Large-Scale Analysis of Genetic and Clinical Patient Data

Biomedical data science has experienced an explosion of new data over the past decade. Abundant genetic and genomic data are increasingly available in large, diverse data sets due to the maturation of modern molecular technologies. Along with these molecular data, dense, rich phenotypic data are also available on comprehensive clinical data sets from health care provider organizations, clinical trials, population health registries, and epidemiologic studies. The methods and approaches for interrogating these large genetic/genomic and clinical data sets continue to evolve rapidly, as our understanding of the questions and challenges continue to emerge. In this review, the state-of-the-art methodologies for genetic/genomic analysis along with complex phenomics will be discussed. This field is changing and adapting to the novel data types made available, as well as technological advances in computation and machine learning. Thus, I will also discuss the future challenges in this exciting and innovative space. The promises of precision medicine rely heavily on the ability to marry complex genetic/genomic data with clinical phenotypes in meaningful ways.

Artificial Intelligence: Bayesian versus Heuristic Method for Diagnostic Decision Support

Evoking strength is one of the important contributions of the field of Biomedical Informatics to the discipline of Artificial Intelligence. The University at Buffalo's Orthopedics Department wanted to create an expert system to assist patients with self-diagnosis of knee problems and to thereby facilitate referral to the right orthopedic subspecialist. They had two independent sports medicine physicians review 469 cases. A board-certified orthopedic sports medicine practitioner, L.B., reviewed any disagreements until a gold standard diagnosis was reached. For each case, the patients entered 126 potential answers to 26 questions into a Web interface. These were modeled by an expert sports medicine physician and the answers were reviewed by L.B. For each finding, the clinician specified the sensitivity (term frequency) and both specificity (Sp) and the heuristic evoking strength (ES). Heuristics are methods of reasoning with only partial evidence. An expert system was constructed that reflected the posttest odds of disease-ranked list for each case. We compare the accuracy of using Sp to that of using ES (original model, p < 0.0008; term importance * disease importance [DItimesTI] model, p < 0.0001: Wilcoxon ranked sum test). For patient referral assignment, Sp in the DItimesTI model was superior to the use of ES. By the fifth diagnosis, the advantage was lost and so there is no difference between the techniques when serving as a reminder system.

Machine Learning Based Localization and Classification with Atomic Magnetometers

We demonstrate identification of position, material, orientation, and shape of objects imaged by a ^{85}Rb atomic magnetometer performing electromagnetic induction imaging supported by machine learning. Machine learning maximizes the information extracted from the images created by the magnetometer, demonstrating the use of hidden data. Localization 2.6 times better than the spatial resolution of the imaging system and successful classification up to 97% are obtained. This circumvents the need of solving the inverse problem and demonstrates the extension of machine learning to diffusive systems, such as low-frequency electrodynamics in media. Automated collection of task-relevant information from quantum-based electromagnetic imaging will have a relevant impact from biomedicine to security.

Clinical code set engineering for reusing EHR data for research: A review

INTRODUCTION

The construction of reliable, reusable clinical code sets is essential when re-using Electronic Health Record (EHR) data for research. Yet code set definitions are rarely transparent and their sharing is almost non-existent. There is a lack of methodological standards for the management (construction, sharing, revision and reuse) of clinical code sets which needs to be addressed to ensure the reliability and credibility of studies which use code sets.

OBJECTIVE

To review methodological literature on the management of sets of clinical codes used in research on clinical databases and to provide a list of best practice recommendations for future studies and software tools.

METHODS

We performed an exhaustive search for methodological papers about clinical code set engineering for re-using EHR data in research. This was supplemented with papers identified by snowball sampling. In addition, a list of e-phenotyping systems was constructed by merging references from several systematic reviews on this topic, and the processes adopted by those systems for code set management was reviewed.

RESULTS

Thirty methodological papers were reviewed. Common approaches included: creating an initial list of synonyms for the condition of interest (n=20); making use of the hierarchical nature of coding terminologies during searching (n=23); reviewing sets with clinician input (n=20); and reusing and updating an existing code set (n=20). Several open source software tools (n=3) were discovered.

DISCUSSION

There is a need for software tools that enable users to easily and quickly create, revise, extend, review and share code sets and we provide a list of recommendations for their design and implementation.

CONCLUSION

Research re-using EHR data could be improved through the further development, more widespread use and routine reporting of the methods by which clinical codes were selected.

Contribution of temporal data to predictive performance in 30-day readmission of morbidly obese patients

Background

Reduction of readmissions after discharge represents an important challenge for many hospitals and has attracted the interest of many researchers in the past few years. Most of the studies in this field focus on building cross-sectional predictive models that aim to predict the occurrence of readmission within 30-days based on information from the current hospitalization. The aim of this study is demonstration of predictive performance gain obtained by inclusion of information from historical hospitalization records among morbidly obese patients.

Methods

The California Statewide inpatient database was used to build regularized logistic regression models for prediction of readmission in morbidly obese patients (n = 18,881). Temporal features were extracted from historical patient hospitalization records in a one-year timeframe. Five different datasets of patients were prepared based on the number of available hospitalizations per patient. Sample size of the five datasets ranged from 4,787 patients with more than five hospitalizations to 20,521 patients with at least two hospitalization records in one year. A 10-fold cross validation was repeted 100 times to assess the variability of the results. Additionally, random forest and extreme gradient boosting were used to confirm the results.

Results

Area under the ROC curve increased significantly when including information from up to three historical records on all datasets. The inclusion of more than three historical records was not efficient. Similar results can be observed for Brier score and PPV value. The number of selected predictors corresponded to the complexity of the dataset ranging from an average of 29.50 selected features on the smallest dataset to 184.96 on the largest dataset based on 100 repetitions of 10-fold cross-validation.

Discussion

The results show positive influence of adding information from historical hospitalization records on predictive performance using all predictive modeling techniques used in this study. We can conclude that it is advantageous to build separate readmission prediction models in subgroups of patients with more hospital admissions by aggregating information from up to three previous hospitalizations.