A Clinical Reasoning-Encoded Case Library Developed through Natural Language Processing.

IMPORTANCE: Case reports that externalize expert diagnostic reasoning are utilized for clinical reasoning instruction but are difficult to search based on symptoms, final diagnosis, or differential diagnosis construction. Computational approaches that uncover how experienced diagnosticians analyze the medical information in a case as they formulate a differential diagnosis can guide educational uses of case reports. OBJECTIVE: To develop a "reasoning-encoded" case database for advanced clinical reasoning instruction by applying natural language processing (NLP), a sub-field of artificial intelligence, to a large case report library. DESIGN: We collected 2525 cases from the New England Journal of Medicine (NEJM) Clinical Pathological Conference (CPC) from 1965 to 2020 and used NLP to analyze the medical terminology in each case to derive unbiased (not prespecified) categories of analysis used by the clinical discussant. We then analyzed and mapped the degree of category overlap between cases. RESULTS: Our NLP algorithms identified clinically relevant categories that reflected the relationships between medical terms (which included symptoms, signs, test results, pathophysiology, and diagnoses). NLP extracted 43,291 symptoms across 2525 cases and physician-annotated 6532 diagnoses (both primary and related diagnoses). Our unsupervised learning computational approach identified 12 categories of medical terms that characterized the differential diagnosis discussions within individual cases. We used these categories to derive a measure of differential diagnosis similarity between cases and developed a website ( universeofcpc.com ) to allow visualization and exploration of 55 years of NEJM CPC case series. CONCLUSIONS: Applying NLP to curated instances of diagnostic reasoning can provide insight into how expert clinicians correlate and coordinate disease categories and processes when creating a differential diagnosis. Our reasoning-encoded CPC case database can be used by clinician-educators to design a case-based curriculum and by physicians to direct their lifelong learning efforts.

Race, Ethnicity, and Disparities in Rheumatology Educational Materials.

OBJECTIVE: To characterize the representation of dark skin color in clinical images across 4 major rheumatology training resources. METHODS: We gathered images of patients with rheumatic diseases from the American College of Rheumatology Image Library, UpToDate, the New England Journal of Medicine Images in Clinical Medicine and Clinical Cases filtered by "Rheumatology," and the 9th edition of Kelley's Textbook of Rheumatology. Investigators used Fitzpatrick's skin phototypes to independently code images depicting visible skin as "light" (skin types I to IV), "dark" (skin types V to VI), or "indeterminate." The representation of dark skin in clinical images was compared to the representation of Asian, Native American, and Black individuals within the US Census population and within lupus cases nationally. RESULTS: Of the 1,043 patient images included in the study, 13.4% had dark skin, 84.0% light skin, and 2.6% indeterminate skin color. Dark skin was underrepresented significantly in rheumatology educational materials and lupus images when compared with the representation of Asian, Native American, and Black individuals within the US Census population (13.4% versus 20.6%; χ2  = 32.8, P < 0.001) and in published studies of patients with systemic lupus erythematous (22.6% versus 44.2%; χ2  = 20.0, P < 0.001). CONCLUSION: Darker skin tones are significantly underrepresented in major rheumatology clinical image banks. Improving representation of racial and ethnic minorities in rheumatology education materials can better equip trainees to recognize and diagnose cutaneous manifestations of rheumatic diseases in these groups.

Using Process Improvement and Systems Redesign to Improve Rheumatology Care Quality in a Safety Net Clinic.

OBJECTIVE: To develop and evaluate interventions to improve quality of care in 4 priority areas in an urban safety net adult rheumatology clinic serving a racially/ethnically and socioeconomically diverse patient population. METHODS: The Institute for Healthcare Improvement's Model for Improvement was used to redesign clinical processes to achieve prespecified benchmarks in the following areas from 2015 to 2017: 13-valent pneumococcal conjugate vaccine (PCV13) administration among immunocompromised patients; disease activity monitoring with the Clinical Disease Activity Index (CDAI) for patients with rheumatoid arthritis; latent tuberculosis infection (LTBI) screening for new biologic users with RA; and reproductive health counseling among women receiving potentially teratogenic medications. We measured performance for each using standardized metrics, defined as the proportion of eligible patients receiving recommended care. RESULTS: There were 1205 patients seen in the clinic between 2015 and 2017. Regarding demographics, 71% were women, 88% identified as racial/ethnic minorities, and 45% were eligible for at least 1 of the quality measures. Shewart charts for the PCV13 and CDAI measures showed evidence of improved healthcare delivery over time. Benchmarks were achieved for the CDAI and LTBI measures with 93% and 91% performance, respectively. Performance for the PCV13 and reproductive health counseling measures was 78% and 46%, respectively, but did not meet prespecified improvement targets. CONCLUSION: Through an interprofessional approach, we were able to achieve durable improvements in key rheumatology quality measures largely by enhancing workflow, engaging nonphysician providers, and managing practice variation.

Physical Activity to Reduce Fatigue in Rheumatoid Arthritis: A Randomized Controlled Trial.

OBJECTIVE: Effective treatments for rheumatoid arthritis (RA) fatigue are limited. We tested the effect of a pedometer-based intervention on increasing physical activity and decreasing fatigue among individuals with RA. METHODS: Participants completed baseline questionnaires; had 1 week of activity monitoring; were randomized to control (education [EDUC]), pedometer and step-monitoring diary (PED), or pedometer and diary plus step targets (PED+) groups, and were followed for 21 weeks. At week 10, questionnaires were administered by phone to all participants. During the final week, all participants again had 1 week of activity monitoring. Primary outcomes were changes in average weekly steps and fatigue (Patient-Reported Outcomes Measurement Information System 7-item questionnaire) from baseline to week 21. Secondary outcomes were self-reported disease activity, physical function, pain interference, and depressive symptoms. Changes in steps were tested using a linear mixed model. Changes in fatigue were tested with repeated-measures models, including baseline, week-10, and week-21 scores. RESULTS: A total of 96 individuals participated. Eight did not complete the 21-week assessments. Both intervention groups significantly increased steps (+1,441 [P = 0.004] for PED and +1,656 [P = 0.001] for PED+), and the EDUC group decreased steps (-747 [P = 0.14]) (group-by-time interaction P = 0.0025). Between-group changes in fatigue were not significantly different (interaction P = 0.21). Mean changes in fatigue scores from baseline to week 21 were -1.6 (with-group P = 0.26), -3.2 (P = 0.02), and -4.8 (P = 0.0002) for EDUC, PED, and PED+ groups, respectively. Function and self-reported disease activity also improved in the PED and PED+ groups. CONCLUSION: Provision of pedometers, with and without providing step targets, was successful in increasing activity levels and decreasing fatigue in this sample of individuals with RA.

Neurologic Manifestations of Primary Sjögren Syndrome.

The neurologic manifestations of primary Sjögren syndrome are varied and can be divided anatomically into 2 categories: peripheral neuropathies and central nervous system (CNS) conditions. Distal sensory and sensorimotor neuropathies are the most common manifestations of peripheral nerve disease in primary Sjögren syndrome. CNS manifestations associated with primary Sjögren syndrome include focal central lesions, conditions that mimic multiple sclerosis, encephalitis, aseptic meningitis, cerebellar syndromes, movement disorders, and problems with memory, cognition, and depression. The heterogeneity of neurologic manifestations in primary Sjögren syndrome complicates the approach to treatment, which should be directed toward the underlying neuropathologic mechanism.

Role of Sleep Disturbance, Depression, Obesity, and Physical Inactivity in Fatigue in Rheumatoid Arthritis.

OBJECTIVE: Fatigue is a major concern for individuals with rheumatoid arthritis (RA). However, in order to treat fatigue adequately, its sources need to be identified. METHODS: Data were collected during a single home visit (number of participants = 158). All participants had physician-diagnosed RA. Assessments of self-reported sleep quality, depression, physical activity, RA disease activity, muscle strength, functional limitations, and body composition were made. Information on demographics, medications, and smoking was collected. The Fatigue Severity Inventory (FSI; measuring average fatigue over the past 7 days) was used as the primary outcome. Analyses were first conducted to evaluate bivariate relationships with fatigue. Correlations among risk factors were examined. Multivariate analyses identified independent predictors of fatigue. RESULTS: The mean ± SD age was 59 ± 11 years, the mean ± SD disease duration was 21 ± 13 years, and 85% of subjects were female. The mean ± SD FSI rating was 3.8 ± 2.0 (range 0-10). In multivariate analyses, self-reported disease activity, poor sleep, depression, and obesity were independently associated with fatigue. Physical inactivity was correlated with poor sleep, depression, and obesity. Mediation analyses indicated that physical inactivity had an indirect association with fatigue, mediated by poor sleep, depression, and obesity. CONCLUSION: This cross-sectional study suggests that fatigue may not be solely a result of RA disease activity, but may result from a constellation of factors that includes RA disease activity or pain, but also includes inactivity, depression, obesity, and poor sleep. The results suggest new avenues for interventions to improve fatigue in individuals with RA, such as increasing physical activity or addressing depression or obesity.

Efficiency Gains for Rheumatology Consultation Using a Novel Electronic Referral System in a Safety-Net Health Setting.

OBJECTIVE: Health information technology (HIT) holds promise in increasing access to rheumatologists by improving the quality and efficiency of referrals, but few studies have examined its use for this purpose. We evaluated the use and impact of a novel electronic referral (eReferral) system in rheumatology in a safety-net health system. METHODS: We examined eReferrals over 4 years. Our primary outcome was use of preconsultation exchange, defined as back-and-forth communication between referring and specialty care providers, facilitating triage of referrals, requests for more information, or resolution of questions without a visit. We calculated the proportion of eReferrals that underwent preconsultation exchange, time to reviewer response, and number of visits scheduled. To increase generalizability, we selected a random sample of eReferrals to undergo additional blinded, adjudicated review to assess agreement on appropriateness for preconsultation exchange. RESULTS: Between 2008 and 2012, 2,383 eReferrals were reviewed and 2,105 were eligible for analysis. One-fourth of eReferrals were resolved without a clinic visit. The proportion of eReferrals undergoing preconsultation exchange increased over time (55% in 2008 versus 74% in 2011), and the volume of referrals also steadily increased over time. Reviewer response time averaged between 1 and 4 days. In the random sample of eReferrals that underwent adjudicated review, agreement between reviewers was high (κ = 0.72). CONCLUSION: HIT-enabled preconsultation exchange was used for a majority of eReferrals and facilitated communication between referring clinicians and rheumatologists. This redesigned system of care allowed for triage of a high number of referrals, with many referrals determined to be appropriate for preconsultation exchange.

Muscle Strength and Changes in Physical Function in Women With Systemic Lupus Erythematosus.

OBJECTIVE: Cross-sectional studies have observed that muscle weakness is associated with worse physical function among women with systemic lupus erythematosus (SLE). The present study examines whether reduced upper and lower extremity muscle strength predict declines in function over time among adult women with SLE. METHODS: One hundred forty-six women from a longitudinal SLE cohort participated in the study. All measures were collected during in-person research visits approximately 2 years apart. Upper extremity muscle strength was assessed by grip strength. Lower extremity muscle strength was assessed by peak knee torque of extension and flexion. Physical function was assessed using the Short Physical Performance Battery (SPPB). Regression analyses modeled associations of baseline upper and lower extremity muscle strength with followup SPPB scores controlling for baseline SPPB, age, SLE duration, SLE disease activity (Systemic Lupus Activity Questionnaire), physical activity level, prednisone use, body composition, and depression. Secondary analyses tested whether associations of baseline muscle strength with followup in SPPB scores differed between intervals of varying baseline muscle strength. RESULTS: Lower extremity muscle strength strongly predicted changes over 2 years in physical function even when controlling for covariates. The association of reduced lower extremity muscle strength with reduced physical function in the future was greatest among the weakest women. CONCLUSION: Reduced lower extremity muscle strength predicted clinically significant declines in physical function, especially among the weakest women. Future studies should test whether therapies that promote preservation of lower extremity muscle strength may prevent declines in function among women with SLE.

Elevated risk for autoimmune disorders in iraq and afghanistan veterans with posttraumatic stress disorder.

BACKGROUND: Posttraumatic stress disorder (PTSD) is associated with endocrine and immune abnormalities that could increase risk for autoimmune disorders. However, little is known about the risk for autoimmune disorders among individuals with PTSD. METHODS: We conducted a retrospective cohort study of 666,269 Iraq and Afghanistan veterans under age 55 who were enrolled in the Department of Veterans Affairs health care system between October 7, 2001, and March 31, 2011. Generalized linear models were used to examine if PTSD, other psychiatric disorders, and military sexual trauma exposure increased risk for autoimmune disorders, including thyroiditis, inflammatory bowel disease, rheumatoid arthritis, multiple sclerosis, and lupus erythematosus, adjusting for age, gender, race, and primary care visits. RESULTS: PTSD was diagnosed in 203,766 veterans (30.6%), and psychiatric disorders other than PTSD were diagnosed in an additional 129,704 veterans (19.5%). Veterans diagnosed with PTSD had significantly higher adjusted relative risk (ARR) for diagnosis with any of the autoimmune disorders alone or in combination compared with veterans with no psychiatric diagnoses (ARR = 2.00; 95% confidence interval, 1.91-2.09) and compared with veterans diagnosed with psychiatric disorders other than PTSD (ARR = 1.51; 95% confidence interval, 1.43-1.59; p < .001). The magnitude of the PTSD-related increase in risk for autoimmune disorders was similar in women and men, and military sexual trauma exposure was independently associated with increased risk in both women and men. CONCLUSIONS: Trauma exposure and PTSD may increase risk for autoimmune disorders. Altered immune function, lifestyle factors, or shared etiology may underlie this association.

Muscle strength, muscle mass, and physical disability in women with systemic lupus erythematosus.

OBJECTIVE: Data describing relationships between muscle strength, muscle mass, and physical disability among individuals with systemic lupus erythematosus (SLE) are limited. The present study examines the relationship of muscle strength and muscle mass with physical disability among adult women with SLE. METHODS: A total of 146 women from a longitudinal SLE cohort participated in the study. All measures were collected during an in-person research visit. Lower extremity muscle strength was assessed by peak knee torque of extension and flexion and by chair-stand time. Total lean body mass, appendicular lean mass, and fat mass (kg/m(2) ) were measured by whole-body dual x-ray absorptiometry. Self-reported physical disability was assessed using the Short Form 36 health survey (SF-36) physical functioning subscale, and the Valued Life Activities (VLA) disability scale. Spearman's rank correlation coefficients tested the correlations between muscle strength, muscle mass, and disability scores. Regression analyses modeled the effect of lower extremity muscle strength and mass on SF-36 and VLA disability scores controlling for age, SLE duration, SLE disease activity measured with the Systemic Lupus Activity Questionnaire, physical activity level, prednisone use, body composition, and depression. RESULTS: On all measures, reduced lower extremity muscle strength was associated with poorer SF-36 and VLA disability scores. Trends persisted after adjustment for covariates. Muscle mass was moderately correlated with muscle strength, but did not contribute significantly to adjusted regression models. CONCLUSION: Lower extremity muscle strength, but not muscle mass, was strongly associated with physical disability scores. While further studies are needed, these findings suggest that improving muscle strength may reduce physical disability among women with SLE.