Natural language processing to identify lupus nephritis phenotype in electronic health records.

BACKGROUND: Systemic lupus erythematosus (SLE) is a rare autoimmune disorder characterized by an unpredictable course of flares and remission with diverse manifestations. Lupus nephritis, one of the major disease manifestations of SLE for organ damage and mortality, is a key component of lupus classification criteria. Accurately identifying lupus nephritis in electronic health records (EHRs) would therefore benefit large cohort observational studies and clinical trials where characterization of the patient population is critical for recruitment, study design, and analysis. Lupus nephritis can be recognized through procedure codes and structured data, such as laboratory tests. However, other critical information documenting lupus nephritis, such as histologic reports from kidney biopsies and prior medical history narratives, require sophisticated text processing to mine information from pathology reports and clinical notes. In this study, we developed algorithms to identify lupus nephritis with and without natural language processing (NLP) using EHR data from the Northwestern Medicine Enterprise Data Warehouse (NMEDW). METHODS: We developed five algorithms: a rule-based algorithm using only structured data (baseline algorithm) and four algorithms using different NLP models. The first NLP model applied simple regular expression for keywords search combined with structured data. The other three NLP models were based on regularized logistic regression and used different sets of features including positive mention of concept unique identifiers (CUIs), number of appearances of CUIs, and a mixture of three components (i.e. a curated list of CUIs, regular expression concepts, structured data) respectively. The baseline algorithm and the best performing NLP algorithm were externally validated on a dataset from Vanderbilt University Medical Center (VUMC). RESULTS: Our best performing NLP model incorporated features from both structured data, regular expression concepts, and mapped concept unique identifiers (CUIs) and showed improved F measure in both the NMEDW (0.41 vs 0.79) and VUMC (0.52 vs 0.93) datasets compared to the baseline lupus nephritis algorithm. CONCLUSION: Our NLP MetaMap mixed model improved the F-measure greatly compared to the structured data only algorithm in both internal and external validation datasets. The NLP algorithms can serve as powerful tools to accurately identify lupus nephritis phenotype in EHR for clinical research and better targeted therapies.

HV1 mtDNA Reveals the High Genetic Diversity and the Ancient Origin of Vietnamese Dogs.

In this study, samples from 429 dog individuals across three main regions of Vietnam (Southern Vietnam (SVN), Central Vietnam (CVN), and Northern Vietnam (NVN)) were collected to analyze the 582 bp region mtDNA HVI, so as to study the genetic diversity and to screen the rare haplotype E in the Vietnamese village dog population. Nine new haplotypes A, two new haplotypes B, and three haplotypes C were unique to Vietnam dogs, in which the new haplotypes An3, An7, Cn1, and Cn3 concerned mutations at new polymorphism sites (15,517, 15,505, 15,479, and 15,933, respectively) which have not been previously reported. The detection of haplotypes A9 and A29, and the appearance of haplotype A200 in the two individual dogs sampled support that the Southeast Asian dog is the ancestor of today's Australian dingo and Polynesian dog. The two rare haplotypes E (E1 and E4) were reconfirmed in Vietnamese dogs and discussed. This study also contributes to strengthening the theory of domestication of dogs to the south of the Yangtze River and the Southeast Asian origin of the dingo.

Association of the Systemic Lupus International Collaborating Clinics Frailty Index and Damage Accrual in Longstanding Systemic Lupus Erythematosus.

OBJECTIVE: To externally validate the Systemic Lupus International Collaborating Clinics Frailty Index (SLICC-FI) in a prevalent systemic lupus erythematosus (SLE) cohort and to assess the ability of the SLICC-FI to predict organ damage accrual among individuals with longstanding SLE. METHODS: This was a secondary analysis of data from the Study of Lupus Vascular and Bone Long-Term Endpoints (SOLVABLE) cohort, which consists of adult women from the Chicago Lupus Database who met the 1997 revised American College of Rheumatology (ACR) classification criteria for SLE. There were 185 patients with SLE enrolled, of whom 149 patients were included in a 5-year follow-up analysis. The SLICC-FI and SLICC/ACR Damage Index (SDI) scores were calculated at baseline and 5-year follow-up. Unadjusted and adjusted logistic regression models estimated the association of baseline SLICC-FI scores (per 0.05 increase) with damage accrual at 5-year follow-up. RESULTS: At enrollment the mean ± SD age of the 149 patients was 43.30 ± 10.15 years, the mean ± SD disease duration was 11.93 ± 8.46 years, and the mean ± SD SDI score was 1.64 ± 1.83. At baseline, the mean ± SD SLICC-FI score was 0.18 ± 0.08, and 36% of participants were categorized as frail (SLICC-FI score >0.21). In a model adjusted for age, race, and disease duration, each 0.05-unit increase in the baseline SLICC-FI score was associated with 28% higher odds of subsequent damage accrual (odds ratio 1.28, 95% confidence interval 1.01-1.63). CONCLUSION: In a prevalent cohort of women with established SLE, higher baseline SLICC-FI scores were associated with a higher risk of subsequent damage accrual at 5-year follow-up.

Use of Popular Opinion Leader Models to Disseminate Information About Clinical Trials to Black Individuals With Lupus in Two US Cities.

OBJECTIVE: Clinical trials for systemic lupus erythematosus ("lupus") under enroll Black individuals despite higher disease prevalence, morbidity, and mortality among Black compared to White individuals. To begin to address this disparity, we leveraged community-academic partnerships in 2 US cities (Boston and Chicago) to train popular opinion leaders (POLs) to disseminate information about clinical trials in predominantly Black communities. METHODS: The team of community and academic partners collaboratively developed a 5-module curriculum about clinical trials, barriers, facilitators, and structural racism in research. We enrolled POLs in Boston and Chicago to participate virtually in the curriculum and assessed knowledge gained by comparing pre- and post-test scores. We described the POLs' ability to disseminate information about clinical trials through their communities. RESULTS: We enrolled 19 POLs in Boston and 16 in Chicago; overall, 71% reported a lupus diagnosis, 94% were female, and 80% self-identified as Black or African American. The program was adapted to virtual formats due to the COVID-19 pandemic. POLs demonstrated significant improvement comparing pre/post scores for the conduct of clinical trials and history of racism in clinical research. Fifteen POLs (43%) reported their dissemination of information about clinical trials. Information reached 425 community members in Boston (90% virtually) and 1,887 in Chicago (95% virtually). CONCLUSION: By leveraging community-academic partnerships, we developed and implemented a curriculum to promote familiarity with clinical trials, leading to information dissemination by POLs in predominantly Black communities that are underrepresented in lupus clinical trials. The program successfully transitioned to a virtual model during the COVID-19 pandemic.

Three Distinct Transcriptional Profiles of Monocytes Associate with Disease Activity in Scleroderma Patients.

OBJECTIVE: Patients with diffuse cutaneous systemic sclerosis (dcSSc) display a complex clinical phenotype. Transcriptional profiling of whole blood or tissue from patients are affected by changes in cellular composition that drive gene expression and an inability to detect minority cell populations. We undertook this study to focus on the 2 main subtypes of circulating monocytes, classical monocytes (CMs) and nonclassical monocytes (NCMs) as a biomarker of SSc disease severity. METHODS: SSc patients were recruited from the Prospective Registry for Early Systemic Sclerosis. Clinical data were collected, as well as peripheral blood for isolation of CMs and NCMs. Age-, sex-, and race-matched healthy volunteers were recruited as controls. Bulk macrophages were isolated from the skin in a separate cohort. All samples were assayed by RNA sequencing (RNA-seq). RESULTS: We used an unbiased approach to cluster patients into 3 groups (groups A-C) based on the transcriptional signatures of CMs relative to controls. Each group maintained their characteristic transcriptional signature in NCMs. Genes up-regulated in group C demonstrated the highest expression compared to the other groups in SSc skin macrophages, relative to controls. Patients from groups B and C exhibited worse lung function than group A, although there was no difference in SSc skin disease at baseline, relative to controls. We validated our approach by applying our group classifications to published bulk monocyte RNA-seq data from SSc patients, and we found that patients without skin disease were most likely to be classified as group A. CONCLUSION: We are the first to show that transcriptional signatures of CMs and NCMs can be used to unbiasedly stratify SSc patients and correlate with disease activity outcome measures.

Electronic health record data use in the assessment of quality indicators for glucocorticoid osteoporosis screening in systemic lupus erythematosus.

OBJECTIVE: To determine whether electronic health record (EHR) data components could be identified and used to assess bone health quality indicators in patients with systemic lupus erythematosus as a foundation for population health management. METHODS: We identified patients in our EHR system who had diagnosis codes for lupus from 2012 to 2017 and characterized them based on the frequency and dosage of prescribed glucocorticoid medications. The medical records of patients who received repeated high-dose glucocorticoid orders were further reviewed for osteoporosis, osteoporotic fractures, receipt of appropriate preventive screening, and orders for protective medications based on established quality indicators. Descriptive statistics were calculated to summarize results. RESULTS: We identified 617 patients with a lupus diagnosis; 414 received glucocorticoid prescriptions, 189 received chronic, high-dose; and 83 received chronic, low-dose prescription orders. Of those with chronic high-dose glucocorticoid prescriptions, 14% had an osteoporosis diagnosis, 3% had an osteoporotic fracture, 51% received a prescription for calcium/vitamin D, 43% had bone mineral density screening orders, 20% received a spine radiograph order, 29% had a documented T-score, 12% received a prescription for osteoporosis medication, and 6% had a documented osteoporosis screening. We were able to identify data elements in the EHR for all nine components of the osteoporosis management quality indicator. CONCLUSIONS: It is possible to identify data in the EHR for all attributes of the quality indicator for osteoporosis in lupus patients who receive chronic high-dose glucocorticoids. However, missing data and need to extract data from text-based notes may make development of population management tools challenging.

Motivational interviewing intervention for increasing physical activity and improving dietary behaviors: The Lupus Intervention Fatigue Trial protocol.

The Lupus Intervention Fatigue Trial (LIFT) is a prospective, randomized controlled trial to assess the effectiveness of a six-month motivational interviewing intervention program versus an educational control to reduce fatigue in persons with systematic lupus erythematosus (SLE). Participants are randomized using a stratified, 1:1 allocation design to the LIFT intervention or control arm. We plan to enroll 236 participants to achieve the target of 200 persons with six-month follow-up for the primary endpoint. Specific aims of this study are to evaluate the impact of the LIFT intervention on 1) self-reported measures of fatigue and 2) impact on accelerometer-measured physical activity. The primary study outcome is six-month change in fatigue from baseline, assessed by the Fatigue Severity Score (FSS). Additional outcomes include objective measures of physical activity, including non-sedentary behavior and moderate-to-vigorous activity (secondary outcome), and adherence to the LIFT dietary intervention, as assessed by nutrient density (diet quality) and recommended food groups/eating patterns (exploratory outcome) in persons with SLE. Intervention effectiveness will be assessed using an intention-to-treat two-arm comparison of six-month change in FSS, with one interim monitoring analysis. A two-sample independent group t-test will compare the six-month changes in FSS between the study arms. Intervention effect durability will be assessed 12-months after baseline (6 months after completion of the intervention). Enrollment began in June 2019 and is expected to end in June 2023. This study will inform future intervention strategies that promote physical activity and improved diet quality to reduce fatigue in persons with SLE.

Evaluation of structured data from electronic health records to identify clinical classification criteria attributes for systemic lupus erythematosus.

OBJECTIVE: Our objective was to develop algorithms to identify lupus clinical classification criteria attributes using structured data found in the electronic health record (EHR) and determine whether they could be used to describe a cohort of people with lupus and discriminate them from a defined healthy control cohort. METHODS: We created gold standard lupus and healthy patient cohorts that were fully adjudicated for the American College of Rheumatology (ACR), Systemic Lupus International Collaborating Clinics (SLICC) and European League Against Rheumatism/ACR (EULAR/ACR) classification criteria and had matched EHR data. We implemented rule-based algorithms using structured data within the EHR system for each attribute of the three classification criteria. Individual criteria attribute and classification criteria algorithms as a whole were assessed over our combined cohorts and the overall performance of the algorithms was measured through sensitivity and specificity. RESULTS: Individual classification criteria attributes had a wide range of sensitivities, 7% (oral ulcers) to 97% (haematological disorders) and specificities, 56% (haematological disorders) to 98% (photosensitivity), but all could be identified in EHR data. In general, algorithms based on laboratory results performed better than those primarily based on diagnosis codes. All three classification criteria systems effectively distinguished members of our case and control cohorts, but the SLICC criteria-based algorithm had the highest overall performance (76% sensitivity, 99% specificity). CONCLUSIONS: It is possible to characterise disease manifestations in people with lupus using classification criteria-based algorithms that assess structured EHR data. These algorithms may reduce chart review burden and are a foundation for identifying subpopulations of patients with lupus based on disease presentation to support precision medicine applications.

Truck Platooning Under Real Traffic Conditions: First Insights on Behavioral Adaptations and Gap Preference of Professional Drivers.

OBJECTIVE: The aim of the study was to investigate (1) how different gap sizes are perceived by professional truck drivers under real traffic conditions and (2) whether semi-automated platoon driving leads to changes in driving behavior of subsequent manual driving. BACKGROUND: Platoon driving is a current branch in the development of automated driving in which two or more vehicles build a convoy. The lead vehicle is controlled manually while following vehicles are electronically coupled to it and drive semi-automated with small gaps in order to achieve a better traffic flow and potential fuel savings. METHOD: In a real road experiment, 10 trained professional truck drivers completed a total of 33 test drives with a two-truck platoon on the German highway A9 with a gap size of either 15 or 21 m, in the leading and the following vehicle. RESULTS: (1) The drivers experienced both gap sizes as comfortable and preferred the smaller gap size of 15 m. (2) Both gap sizes led to significantly higher standard deviation of lane position in post- compared to pre-platoon driving. No significant difference in distance keepings in post- compared to pre-platoon driving occurred. Qualitative data give hints on difficulties, when switching back to regular truck driving. CONCLUSION: The results implicate that small gap sizes are perceived as comfortable by drivers and that platoon driving has an influence on subsequent manual driving. APPLICATION: Countermeasures to behavioral adaptations should be considered in order to ensure a safe conduction of platoon driving.

Assessing the Concordance of Clinical Classification Criteria for Lupus Between Electronic Health Records and a Physician Curated Registry.

We developed a computable phenotype for systemic lupus erythematosus (SLE) based on the Systemic Lupus International Collaborative Clinics clinical classification criteria set for SLE. We evaluated the phenotype over registry and EHR data for the same patient population to determine concordance of criteria detected in both datasets and to assess which types of structured data detected individual classification criteria. We identified a concordance of 68% between registry and EHR data relying solely on structured data.

Addressing Lupus Health Disparities: The MONARCAS Community and Academic Collaborative Program.

Purpose: The Centers for Disease Control (CDC) Popular Opinion Leader (POL) model was implemented in a lupus education program (MONARCAS) for the Latino community. The program aim was to increase lupus awareness by training high school students, community health workers, and parents. Methods: A curriculum was developed training POLs to disseminate concepts about lupus signs and symptoms. Pre- and post-program questions assessed lupus knowledge and message dissemination. Results: POL groups represented distinct demographic characteristics with Spanish or English language dominance. POLs reported increased lupus knowledge and program satisfaction. Conclusions: Future program goals should aim to increase understanding and improving access to care for Latino communities affected by lupus.

Disease Outcomes and Care Fragmentation Among Patients With Systemic Lupus Erythematosus.

OBJECTIVE: To examine the impact of care fragmentation across multiple health care institutions on disease outcomes in patients with systemic lupus erythematosus (SLE). METHODS: Using the Chicago HealthLNK Data Repository, an assembly of electronic health records from 6 institutions, we identified patients with SLE, using International Classification of Diseases, Ninth Revision (ICD-9) codes, whose care was delivered at more than 1 organization. We examined whether patients had severe infections or comorbidities (ICD-9 code defined) that indicated SLE-induced damage. T-tests and chi-square tests were used to examine differences between fragmentation groups. Logistic regression was used to assess factors contributing to the occurrence of disease outcomes. RESULTS: We identified 4,276 patients with SLE. A total of 856 (20%) received care from more than 1 health care institution. African American patients and patients with public insurance were more likely to experience care fragmentation compared to white and private insurance patients (odds ratio [OR] 1.66, 95% confidence interval [95% CI] 1.44-1.97 and OR 1.63, 95% CI 1.42-1.95). We identified increased risk of infections (OR 1.57, 95% CI 1.30-1.88), cardiovascular disease (OR 1.51, 95% CI 1.23-1.86), end-stage renal disease (OR 1.34, 95% CI 1.05-1.70), nephritis (OR 1.28, 95% CI 1.07-1.54), and stroke (OR 1.28, 95% CI 1.01-1.62) among patients with fragmented care, adjusted for age, sex, race, insurance status, length of followup time, and total visit count. CONCLUSION: In this cross-site cohort of SLE patients, care fragmentation is associated with increased risk of severe infection and comorbidities. These results suggest that improved health information exchange could positively impact outcomes for SLE patients.

Validity of two new patient-reported outcome measures in systemic sclerosis: Patient-Reported Outcomes Measurement Information System 29-item Health Profile and Functional Assessment of Chronic Illness Therapy-Dyspnea short form.

OBJECTIVE: Many patient-reported outcome (PRO) instruments used in systemic sclerosis (SSc) trials are limited by lack of validation, licensing fees, and complicated scoring systems. We assessed the construct validity for discriminative purposes of 2 new PRO instruments, the Patient-Reported Outcomes Measurement Information System 29-item Health Profile (PROMIS-29) and the Functional Assessment of Chronic Illness Therapy-Dyspnea short form (FACIT-Dyspnea), measuring health status and dyspnea in SSc patients. METHODS: Seventy-three patients participated in a cross-sectional study at a tertiary SSc program. PROMIS-29, FACIT-Dyspnea, and legacy PRO instruments used in clinical trials (Medical Research Council Dyspnea Score, St. George's Respiratory Questionnaire, Health Assessment Questionnaire disability index, and Short Form 36) were administered. Composite severity scores using an adaptation of the Medsger Disease Severity Index were generated using clinical, diagnostic, and laboratory information. PROMIS-29 and FACIT-Dyspnea scores were compared with legacy PRO measures and composite severity scores. RESULTS: The mean patient age (84% women) was 51 years (range 22-72 years). The mean SSc disease duration from the onset of the first non-Raynaud's phenomenon symptom was 7.2 years (range 0-45 years). Spearman's correlation coefficients across FACIT-Dyspnea and PROMIS physical functioning scores with legacy PRO instruments were generally high (range 0.50-0.86); those between PROMIS and FACIT-Dyspnea with composite disease severity scores were more modest, but statistically significant (range 0.33-0.48, P < 0.01). CONCLUSION: PROMIS-29 and FACIT-Dyspnea are valid instruments to measure the health status of SSc patients. PROMIS-29 and FACIT-Dyspnea may be preferable to legacy instruments because they are freely available in multiple languages and simple to administer, score, and interpret.