Identifying the association between serum urate levels and gout flares in patients taking urate-lowering therapy: a post hoc cohort analysis of the CARES trial with consideration of dropout.

OBJECTIVE: To investigate gout flare rates based on repeated serum urate (SU) measurements in a randomised controlled trial of urate-lowering therapy (ULT), accounting for dropout and death. METHODS: We performed a secondary analysis using data from Cardiovascular Safety of Febuxostat or Allopurinol in Patients with Gout, which randomised participants to febuxostat or allopurinol, titrated to target SU <6 mg/dL with flare prophylaxis for 6 months. SU was categorised as ≤3.9, 4.0-5.9, 6.0-7.9, 8.0-9.9 or ≥ 10 mg/dL at each 3-6 month follow-up. The primary outcome was gout flare. Poisson regression models, adjusted for covariates and factors related to participant retention versus dropout, estimated gout flare incidence rate ratios by time-varying SU category. RESULTS: Among 6183 participants, the median age was 65 years and 84% were male. Peak gout flare rates for all SU categories were observed in months 0-6, coinciding with the initiation of ULT and months 6-12 after stopping prophylaxis. Flare rates were similar across SU groups in the initial year of ULT. During months 36-72, a dose-response relationship was observed between the SU category and flare rate. Lower flare rates were observed when SU ≤3.9 mg/dL and greater rates when SU ≥10 mg/dL, compared with SU 4.0-5.9 mg/dL (p for trend <0.01). CONCLUSION: Gout flare rates were persistently higher when SU ≥6 mg/dL after the first year of ULT after accounting for censoring. The spike in flares in all categories after stopping prophylaxis suggests a longer duration of prophylaxis may be warranted.

Changes in alcohol intake and serum urate changes: longitudinal analyses of annual medical examination database.

INTRODUCTION: Despite the established cross-sectional association between alcohol intake and serum urate (SU), its longitudinal association remains unknown. This study aimed to determine whether changes in alcohol intake have a clinically relevant association with SU change. METHOD: We conducted retrospective analyses using systematically collected annual medical examination data from October 2012 to October 2022 in a Japanese preventive medicine centre. The exposure was changes in alcohol intake between two consecutive visits. The association of SU changes with alcohol intake changes was estimated by mixed-effect linear regression with adjustment for relevant covariates. RESULTS: We analysed 63 486 participants (median age, 47.0 years; 55% women; 58.6% regular alcohol drinkers with a median of 1.4 drinks/day) with 370 572 visits. The median SU level was 5.3 mg/dL, and 506 (0.8%) participants had diagnoses of gout or hyperuricemia without medication use during the study period. Decreasing one daily alcohol intake had a clinically small association with SU changes (-0.019 (95% CI: -0.021 to -0.017) mg/dL). Beer had the largest association with SU (-0.036 (95% CI: -0.039 to -0.032) mg/dL for one beer decrease). Complete discontinuation of any alcohol from a mean of 0.8 drinks/day was associated with -0.056 mg/dL (95% CI: -0.068 to -0.043) decrease in SU; the association became larger in hyperuricemic participants (-0.110 mg/dL (95% CI: -0.154 to -0.066) for alcohol discontinuation from a mean of 1.0 drinks/day). CONCLUSIONS: This study revealed changes in alcohol intake had small associations with SU change at the general Japanese population level. Complete discontinuation of alcohol in hyperuricemic participants had only modest improvement in SU.

Fractures in Patients With Acute Calcium Pyrophosphate Crystal Arthritis Versus Matched Comparators in a Large Cohort Study.

OBJECTIVE: Calcium pyrophosphate deposition (CPPD) disease was associated with osteopenia in two cross-sectional studies. We compared fracture risks in patients with acute calcium pyrophosphate (CPP) crystal arthritis versus matched comparators. METHODS: We performed a longitudinal cohort study using electronic health record data from a single large academic health system, with data from 1991 to 2023. Patients with one or more episodes of acute CPP crystal arthritis were matched to comparators on the index date (first documentation of "pseudogout" or synovial fluid CPP crystals or matched encounter) and first encounter in the health system. The primary outcome was first fracture at the humerus, wrist, hip, or pelvis. We excluded patients with fracture before the index date. Covariates included demographics, body mass index, smoking, comorbidities, health care use, glucocorticoids, and osteoporosis treatments. We estimated incidence rates and adjusted hazard ratios for fracture. Sensitivity analyses excluded patients prescribed glucocorticoids, patients prescribed osteoporosis treatments, or patients with rheumatoid arthritis and additionally adjusted for chronic kidney disease. RESULTS: We identified 1,148 patients with acute CPP crystal arthritis matched to 3,730 comparators, with a mean age of 73 years. Glucocorticoids and osteoporosis treatments were more frequent in the acute CPP crystal arthritis cohort. Fracture incidence rates were twice as high in the acute CPP crystal arthritis cohort (11.7 per 1,000 person-years) versus comparators (5.5 per 1,000 person-years). After multivariable adjustment, fracture relative risk was twice as high in the acute CPP crystal arthritis cohort (hazard ratio 1.8 [95% confidence interval 1.3-2.3]); results were similar in sensitivity analyses. CONCLUSION: In this first published study of fractures and CPPD, fracture risk was nearly doubled in patients with acute CPP crystal arthritis.

The 2023 ACR/EULAR Classification Criteria for Calcium Pyrophosphate Deposition Disease.

OBJECTIVE: Calcium pyrophosphate deposition (CPPD) disease is prevalent and has diverse presentations, but there are no validated classification criteria for this symptomatic arthritis. The American College of Rheumatology (ACR) and EULAR have developed the first-ever validated classification criteria for symptomatic CPPD disease. METHODS: Supported by the ACR and EULAR, a multinational group of investigators followed established methodology to develop these disease classification criteria. The group generated lists of candidate items and refined their definitions, collected de-identified patient profiles, evaluated strengths of associations between candidate items and CPPD disease, developed a classification criteria framework, and used multi-criterion decision analysis to define criteria weights and a classification threshold score. The criteria were validated in an independent cohort. RESULTS: Among patients with joint pain, swelling, or tenderness (entry criterion) whose symptoms are not fully explained by an alternative disease (exclusion criterion), the presence of crowned dens syndrome or calcium pyrophosphate crystals in synovial fluid are sufficient to classify a patient as having CPPD disease. In the absence of these findings, a score >56 points using weighted criteria, comprising clinical features, associated metabolic disorders, and results of laboratory and imaging investigations, can be used to classify as CPPD disease. These criteria had a sensitivity of 92.2% and specificity of 87.9% in the derivation cohort (190 CPPD cases, 148 mimickers), whereas sensitivity was 99.2% and specificity was 92.5% in the validation cohort (251 CPPD cases, 162 mimickers). CONCLUSION: The 2023 ACR/EULAR CPPD disease classification criteria have excellent performance characteristics and will facilitate research in this field.

The 2023 ACR/EULAR classification criteria for calcium pyrophosphate deposition disease.

OBJECTIVE: Calcium pyrophosphate deposition (CPPD) disease is prevalent and has diverse presentations, but there are no validated classification criteria for this symptomatic arthritis. The American College of Rheumatology (ACR) and EULAR have developed the first-ever validated classification criteria for symptomatic CPPD disease. METHODS: Supported by the ACR and EULAR, a multinational group of investigators followed established methodology to develop these disease classification criteria. The group generated lists of candidate items and refined their definitions, collected de-identified patient profiles, evaluated strengths of associations between candidate items and CPPD disease, developed a classification criteria framework, and used multi-criterion decision analysis to define criteria weights and a classification threshold score. The criteria were validated in an independent cohort. RESULTS: Among patients with joint pain, swelling, or tenderness (entry criterion) whose symptoms are not fully explained by an alternative disease (exclusion criterion), the presence of crowned dens syndrome or calcium pyrophosphate crystals in synovial fluid are sufficient to classify a patient as having CPPD disease. In the absence of these findings, a score>56 points using weighted criteria, comprising clinical features, associated metabolic disorders, and results of laboratory and imaging investigations, can be used to classify as CPPD disease. These criteria had a sensitivity of 92.2% and specificity of 87.9% in the derivation cohort (190 CPPD cases, 148 mimickers), whereas sensitivity was 99.2% and specificity was 92.5% in the validation cohort (251 CPPD cases, 162 mimickers). CONCLUSION: The 2023 ACR/EULAR CPPD disease classification criteria have excellent performance characteristics and will facilitate research in this field.

Denosumab and incidence of type 2 diabetes among adults with osteoporosis: population based cohort study.

OBJECTIVE: To estimate the effect of denosumab compared with oral bisphosphonates on reducing the risk of type 2 diabetes in adults with osteoporosis. DESIGN: Population based study involving emulation of a randomized target trial using electronic health records. SETTING: IQVIA Medical Research Data primary care database in the United Kingdom, 1995-2021. PARTICIPANTS: Adults aged 45 years or older who used denosumab or an oral bisphosphonate for osteoporosis. MAIN OUTCOME MEASURES: The primary outcome was incident type 2 diabetes, as defined by diagnostic codes. Cox proportional hazards models were used to estimate adjusted hazard ratios and 95% confidence intervals, comparing denosumab with oral bisphosphonates using an as treated approach. RESULTS: 4301 new users of denosumab were matched on propensity score to 21 038 users of an oral bisphosphonate and followed for a mean of 2.2 years. The incidence rate of type 2 diabetes in denosumab users was 5.7 (95% confidence interval 4.3 to 7.3) per 1000 person years and in oral bisphosphonate users was 8.3 (7.4 to 9.2) per 1000 person years. Initiation of denosumab was associated with a reduced risk of type 2 diabetes (hazard ratio 0.68, 95% confidence interval 0.52 to 0.89). Participants with prediabetes appeared to benefit more from denosumab compared with an oral bisphosphonate (hazard ratio 0.54, 0.35 to 0.82), as did those with a body mass index ≥30 (0.65, 0.40 to 1.06). CONCLUSIONS: In this population based study, denosumab use was associated with a lower risk of incident type 2 diabetes compared with oral bisphosphonate use in adults with osteoporosis. This study provides evidence at a population level that denosumab may have added benefits for glucose metabolism compared with oral bisphosphonates.

Patient-Reported Outcomes in Calcium Pyrophosphate Deposition Disease Compared to Gout and Osteoarthritis.

OBJECTIVE: Calcium pyrophosphate deposition (CPPD) disease prevalence is similar to that of gout and osteoarthritis (OA), yet CPPD outcomes research greatly lags behind research in these other forms of arthritis. We compared validated patient-reported outcome measures in patients with CPPD vs gout and OA. METHODS: Patients with CPPD were recruited from Brigham and Women's Hospital from 2018 to 2022. Presence of CPPD manifestations (acute calcium pyrophosphate [CPP] crystal arthritis, chronic CPP inflammatory arthritis, and/or OA with CPPD) was confirmed by medical record review. Baseline surveys included the Gout Assessment Questionnaire version 2.0, modified to ask about "pseudogout" rather than "gout"; Routine Assessment of Patient Index Data 3 (RAPID-3); and Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC). We compared responses in patients with CPPD against published gout and OA cohort studies. RESULTS: Among 47 patients with CPPD, the mean age was 71.9 years and 51% were female. Sixty-eight percent had at least 1 episode of acute CPP crystal arthritis, 40% had chronic CPP inflammatory arthritis, and 62% had OA with CPPD. Pain visual analog scale scores during a flare were similar in CPPD (mean 6.8 [SD 1.9]) and gout (mean 6.7 [SD 2.6]; P = 0.78). Patients with CPPD reported significantly greater unmet treatment need than patients with gout (P = 0.04). RAPID-3 scores in CPPD (mean 8.1 [SD 5.6]) were lower than in gout (mean 12.1 [SD 6.2]; P < 0.01) and similar in OA (mean 6.8 [SD 6.1]; P = 0.30). Patients with CPPD had significantly worse WOMAC stiffness scores than patients with mild OA, and significantly better WOMAC function scores than patients with severe OA. CONCLUSION: Patients with CPPD may experience pain comparable to that in gout and OA and reported substantial unmet treatment needs.

Humoral and cellular immune responses in persons with rheumatoid arthritis after a third dose of mRNA COVID-19 vaccine.

OBJECTIVE: Disease-modifying anti-rheumatic drugs (DMARDs) that treat rheumatoid arthritis (RA) may reduce immune responses to COVID-19 vaccination. We compared humoral and cell-mediated immunity before and after a 3rd dose of mRNA COVID vaccine in RA subjects. METHODS: RA patients that received 2 doses of mRNA vaccine enrolled in an observational study in 2021 before receiving a 3rd dose. Subjects self-reported holding or continuing DMARDs. Blood samples were collected pre- and 4 weeks after the 3rd dose. 50 healthy controls provided blood samples. Humoral response was measured with in-house ELISA assays for anti-Spike IgG (anti-S) and anti-receptor binding domain IgG (anti-RBD). T cell activation was measured after stimulation with SARS-CoV-2 peptide. Spearman's correlations assessed the relationship between anti-S, anti-RBD, and frequencies of activated T cells. RESULTS: Among 60 subjects, mean age was 63 years and 88% were female. 57% of subjects held at least 1 DMARD around the 3rd dose. 43% (anti-S) and 62% (anti-RBD) had a normal humoral response at week 4, defined as ELISA within 1 standard deviation of the healthy control mean. No differences in antibody levels were observed based on holding DMARDs. Median frequency of activated CD4 T cells was significantly greater post- vs. pre-3rd dose. Changes in antibody levels did not correlate with change in frequency of activated CD4 T cells. CONCLUSION: Virus-specific IgG levels significantly increased in RA subjects using DMARDs after completing the primary vaccine series, though fewer than two-thirds achieved a humoral response like healthy controls. Humoral and cellular changes were not correlated.

Predictors of low spike antibody response in patients with systemic rheumatic disease after an initial course of COVID-19 vaccination.

BACKGROUND: Patients with rheumatic disease may mount a suboptimal serologic response to COVID-19 vaccination. We evaluated predictors of low antibody response in a clinic-based cohort. METHODS: We conducted a cross-sectional study using electronic health record (EHR) data at Brigham and Women's Hospital, Boston, MA. Patients with systemic rheumatic disease that had SARS-CoV-2 spike antibody (Ab) tested using the Roche Elecsys immunoassay, February-August 2021, after 2 doses of mRNA vaccine or 1 dose of adenovirus vector vaccine were identified. Demographics, systemic rheumatic disease, vaccination dates, and disease-modifying antirheumatic drugs (DMARDs) were extracted. The primary outcome was low spike Ab (≤ 200 U/mL). Logistic regression models estimated predictors of low spike Ab. RESULTS: Among 382 patients, the mean age was 57 years, 77% were female, and 37% had low spike Ab. Older age (OR 1.03, 95% CI [1.02, 1.05]), SLE (OR 4.81 [2.08, 8.43], reference: inflammatory arthritis), prednisone (OR 1.67 [1.03, 2.74]), and rituximab (OR 22.91 [9.85, 53.29]) were significantly associated with higher odds of low spike Ab. Use of csDMARD monotherapy (OR 0.12 [0.04, 0.33]) and JAK inhibitors (OR 0.41 [0.18, 0.92]) were associated with significantly lower odds for low spike Ab. After adjusting for systemic rheumatic disease and DMARDs, SLE and rituximab remained significantly associated with low spike Ab. CONCLUSIONS: Over a third of patients with systemic rheumatic disease with spike Ab tested in routine care had low spike Ab after 2 doses of mRNA or 1 dose of adenovirus vector COVID-19 vaccine. SLE and rituximab were significant risk factors for low spike Ab. KEY POINTS: • More than one-third of patients with systemic rheumatic disease that had spike Ab tested in routine care had low spike Ab after 2 doses of mRNA or 1 dose of adenovirus vector COVID-19 vaccine. • Diagnosis of SLE, use of prednisone, and use of rituximab were significantly associated with greater odds of low spike antibodies. • These data underscore the importance of additional doses of COVID-19 vaccine and prophylactic Evusheld in immunosuppressed patients with systemic rheumatic disease as recommended by the US Centers for Disease Control.

Association of Ultraviolet B Radiation and Risk of Systemic Lupus Erythematosus Among Women in the Nurses' Health Studies.

OBJECTIVE: Ultraviolet (UV) radiation exposure is associated with photosensitivity, rashes, and flares in systemic lupus erythematosus (SLE). However, it is not known whether UV exposure increases risk of developing SLE. We examined UV exposure and SLE risk in a large prospective cohort. METHODS: The Nurses' Health Study (NHS) enrolled 121,700 US female nurses in 1976; in 1989, 116,429 nurses were enrolled in NHS II. Biennial questionnaires collected lifestyle and medical data. Self-reported incident SLE by American College of Rheumatology classification criteria was confirmed by medical record review. Ambient UV exposure was estimated by linking geocoded residential addresses with a spatiotemporal UV exposure model. Cox models estimated hazard ratios (HRs) and 95% confidence intervals (95% CIs) across tertiles of time-varying cumulative average UV. We examined SLE risk overall and stratified by anti-Ro/La antibodies and by cutaneous manifestations from 1976 through 2014 (NHS)/2015 (NHS II), adjusting for confounders. RESULTS: With 6,054,665 person-years of exposure, we identified 297 incident SLE cases; the mean ± SD age at diagnosis was 49.8 ± 10.6 years. At diagnosis, 16.8% of women had +anti-Ro/La, and 80% had either +anti-Ro/La or ≥1 cutaneous manifestation. Compared with the lowest UV exposure tertile, risk of overall SLE was increased, but not significantly (HR 1.28 [95%CI 0.96-1.70]). Women in the highest tertile had increased risk of malar rash (HR 1.62 [95% CI 1.04-2.52]). CONCLUSION: Cumulative UV exposure was not associated with SLE risk. Higher UV exposure, however, was associated with increased risk of malar rash at presentation. UV exposure may trigger SLE onset with malar rash among susceptible women.

Cardiovascular Disease Risk in Calcium Pyrophosphate Deposition Disease: A Nationwide Study of Veterans.

OBJECTIVE: Calcium pyrophosphate deposition (CPPD) disease represents a common crystalline arthritis with a range of manifestations. Our goal was to investigate risks for cardiovascular events in patients with CPPD. METHODS: We performed a retrospective matched cohort analysis in the Veterans Health Administration Corporate Data Warehouse, 2010-2014. CPPD was defined by ≥1 International Classification of Diseases, Ninth Revision codes for chondrocalcinosis or calcium metabolism disorder. CPPD patients were age- and sex-matched to approximately 4 patients without codes for CPPD; we excluded patients with a cardiovascular event during the 365 days prior to the index date. Demographic information, traditional cardiovascular risk factors, medications, and health care utilization were assessed at baseline. The primary outcome was a major adverse cardiovascular event (MACE: myocardial infarction, acute coronary syndrome, coronary revascularization, stroke, or death). Secondary outcomes included individual components of MACE. Cox proportional hazards models estimated fully adjusted hazard ratios (HRs) and 95% confidence intervals (95% CIs). RESULTS: We identified 23,124 CPPD patients matched to 86,629 non-CPPD patients with >250,000 person-years of follow-up. The study population was 96% male, mean age was 78 years, and 75% were White. The frequency of traditional cardiovascular risk factors was similar between the 2 cohorts. CPPD was not significantly associated with risk for MACE (HR 0.98 [95% CI 0.94-1.02]) in fully adjusted models, though risks of myocardial infarction, acute coronary syndrome, and stroke were significantly higher in the CPPD cohort compared to the non-CPPD cohort. CONCLUSION: CPPD did not confer an increased risk for MACE, a composite end point including all-cause mortality. Our results propose CPPD as a novel risk factor for MACE components, including myocardial infarction, acute coronary syndrome, and stroke.

Confirming Prior and Identifying Novel Correlates of Acute Calcium Pyrophosphate Crystal Arthritis.

OBJECTIVE: To investigate previously identified and novel correlates of acute calcium pyrophosphate (CPP) crystal arthritis among well-characterized cases. METHODS: In this case-control study, we identified cases of acute CPP crystal arthritis using a validated algorithm (positive predictive value 81%) applied in the Partners HealthCare electronic health record (EHR). Cases were matched to general patient controls on the year of first EHR encounter and index date. Prespecified potential correlates included sex, race, and comorbidities and medications previously associated with CPP deposition/acute CPP crystal arthritis in the literature. We estimated odds ratios (ORs) and 95% confidence intervals using conditional logistic regression models adjusted for demographic characteristics, comorbidities, medications prescribed in the past 90 days, health care utilization, and multimorbidity score. RESULTS: We identified 1,697 cases matched to 6,503 controls. Mean ± SD age was 73.7 ± 11.8 years, 56.7% were female, 80.8% were White, and 10.3% were Black. All prespecified covariates were more common in cases than controls. Osteoarthritis (OR 3.08), male sex (OR 1.35), rheumatoid arthritis (OR 2.09), gout (OR 2.83), proton pump inhibitors (OR 1.94), loop diuretics (OR 1.60), and thiazides (OR 1.46) were significantly associated with acute CPP crystal arthritis after full adjustment. Black race was associated with lower odds for acute CPP crystal arthritis compared to White race (OR 0.47). CONCLUSION: Using a validated algorithm to identify nearly 1,700 patients with acute CPP crystal arthritis, we confirmed important correlates of this acute manifestation of CPP deposition. This is the first study to report higher odds for acute CPP crystal arthritis among males.

Imaging Features of Calcium Pyrophosphate Deposition Disease: Consensus Definitions From an International Multidisciplinary Working Group.

OBJECTIVE: To develop definitions for imaging features being considered as potential classification criteria for calcium pyrophosphate deposition (CPPD) disease, additional to clinical and laboratory criteria, and to compile example images of CPPD on different imaging modalities. METHODS: The American College of Rheumatology and European Alliance of Associations for Rheumatology CPPD classification criteria Imaging Advisory Group (IAG) and Steering Committee drafted definitions of imaging features that are characteristic of CPPD on conventional radiography (CR), conventional computed tomography (CT), dual-energy CT (DECT), and magnetic resonance imaging (MRI). An anonymous expert survey was undertaken by a 35-member Combined Expert Committee, including all IAG members. The IAG and 5 external musculoskeletal radiologists with expertise in CPPD convened virtually to further refine item definitions and voted on example images illustrating CR, CT, and DECT item definitions, with ≥90% agreement required to deem them acceptable. RESULTS: The Combined Expert Committee survey indicated consensus on all CR definitions. The IAG and external radiologists reached consensus on CT and DECT item definitions, which specify that calcium pyrophosphate deposits appear less dense than cortical bone. The group developed an MRI definition and acknowledged limitations of this modality for CPPD. Ten example images for CPPD were voted acceptable (4 CR, 4 CT, and 2 DECT), and 3 images of basic calcium phosphate deposition were voted acceptable to serve as contrast against imaging features of CPPD. CONCLUSION: An international group of rheumatologists and musculoskeletal radiologists defined imaging features characteristic of CPPD on CR, CT, and DECT and assembled a set of example images as a reference for future clinical research studies.

Reliability and Diagnostic Accuracy of Radiography for the Diagnosis of Calcium Pyrophosphate Deposition: Performance of the Novel Definitions Developed by an International Multidisciplinary Working Group.

OBJECTIVE: To assess the reliability and diagnostic accuracy of new radiographic imaging definitions developed by an international multidisciplinary working group for identification of calcium pyrophosphate deposition (CPPD). METHODS: Patients with knee osteoarthritis scheduled for knee replacement were enrolled. Two radiologists and 2 rheumatologists twice assessed radiographic images for presence or absence of CPPD in menisci, hyaline cartilage, tendons, joint capsule, or synovial membrane, using the new definitions. In case of disagreement, a consensus decision was made and considered for the assessment of diagnostic performance. Histologic examination of postsurgical specimens under compensated polarized light microscopy was the reference standard. Prevalence-adjusted bias-adjusted kappa values were used to assess reliability, and diagnostic performance statistics were calculated. RESULTS: Sixty-seven patients were enrolled for the reliability study. The interobserver reliability was substantial in most of the assessed structures when considering all 4 readers (κ range 0.59-0.90), substantial to almost perfect among radiologists (κ range 0.70-0.91), and moderate to almost perfect among rheumatologists (κ range 0.46-0.88). The intraobserver reliability was substantial to almost perfect for all the observers (κ range 0.70-1). Fifty-one patients were included in the accuracy study. Radiography demonstrated an overall specificity of 92% for CPPD, but sensitivity remained low for all sites and for the overall diagnosis (54%). CONCLUSION: The new radiographic definitions of CPPD are highly specific against the gold standard of histologic diagnosis. When the described radiographic findings are present, these definitions allow for a definitive diagnosis of CPPD, rather than other calcium-containing crystal depositions; however, a negative radiographic finding does not exclude the diagnosis.

Weakly Semi-supervised phenotyping using Electronic Health records.

OBJECTIVE: Electronic Health Record (EHR) based phenotyping is a crucial yet challenging problem in the biomedical field. Though clinicians typically determine patient-level diagnoses via manual chart review, the sheer volume and heterogeneity of EHR data renders such tasks challenging, time-consuming, and prohibitively expensive, thus leading to a scarcity of clinical annotations in EHRs. Weakly supervised learning algorithms have been successfully applied to various EHR phenotyping problems, due to their ability to leverage information from large quantities of unlabeled samples to better inform predictions based on a far smaller number of patients. However, most weakly supervised methods are subject to the challenge to choose the right cutoff value to generate an optimal classifier. Furthermore, since they only utilize the most informative features (i.e., main ICD and NLP counts) they may fail for episodic phenotypes that cannot be consistently detected via ICD and NLP data. In this paper, we propose a label-efficient, weakly semi-supervised deep learning algorithm for EHR phenotyping (WSS-DL), which overcomes the limitations above. MATERIALS AND METHODS: WSS-DL classifies patient-level disease status through a series of learning stages: 1) generating silver standard labels, 2) deriving enhanced-silver-standard labels by fitting a weakly supervised deep learning model to data with silver standard labels as outcomes and high dimensional EHR features as input, and 3) obtaining the final prediction score and classifier by fitting a supervised learning model to data with a minimal number of gold standard labels as the outcome, and the enhanced-silver-standard labels and a minimal set of most informative EHR features as input. To assess the generalizability of WSS-DL across different phenotypes and medical institutions, we apply WSS-DL to classify a total of 17 diseases, including both acute and chronic conditions, using EHR data from three healthcare systems. Additionally, we determine the minimum quantity of training labels required by WSS-DL to outperform existing supervised and semi-supervised phenotyping methods. RESULTS: The proposed method, in combining the strengths of deep learning and weakly semi-supervised learning, successfully leverages the crucial phenotyping information contained in EHR features from unlabeled samples. Indeed, the deep learning model's ability to handle high-dimensional EHR features allows it to generate strong phenotype status predictions from silver standard labels. These predictions, in turn, provide highly effective features in the final logistic regression stage, leading to high phenotyping accuracy in notably small subsets of labeled data (e.g. n = 40 labeled samples). CONCLUSION: Our method's high performance in EHR datasets with very small numbers of labels indicates its potential value in aiding doctors to diagnose rare diseases as well as conditions susceptible to misdiagnosis.

Using observational study data as an external control group for a clinical trial: an empirical comparison of methods to account for longitudinal missing data.

BACKGROUND: Observational data are increasingly being used to conduct external comparisons to clinical trials. In this study, we empirically examined whether different methodological approaches to longitudinal missing data affected study conclusions in this setting. METHODS: We used data from one clinical trial and one prospective observational study, both Norwegian multicenter studies including patients with recently diagnosed rheumatoid arthritis and implementing similar treatment strategies, but with different stringency. A binary disease remission status was defined at 6, 12, and 24 months in both studies. After identifying patterns of longitudinal missing outcome data, we evaluated the following five approaches to handle missingness: analyses of patients with complete follow-up data, multiple imputation (MI), inverse probability of censoring weighting (IPCW), and two combinations of MI and IPCW. RESULTS: We found a complex non-monotone missing data pattern in the observational study (N = 328), while missing data in the trial (N = 188) was monotone due to drop-out. In the observational study, only 39.0% of patients had complete outcome data, compared to 89.9% in the trial. All approaches to missing data indicated favorable outcomes of the treatment strategy in the trial and resulted in similar study conclusions. Variations in results across approaches were mainly due to variations in estimated outcomes for the observational data. CONCLUSIONS: Five different approaches to handle longitudinal missing data resulted in similar conclusions in our example. However, the extent and complexity of missing observational data affected estimated comparative outcomes across approaches, highlighting the need for careful consideration of methods to account for missingness in this setting. Based on this empirical examination, we recommend using a prespecified advanced missing data approach to account for longitudinal missing data, and to conduct alternative approaches in sensitivity analyses.

Risk of cardiovascular events in patients having had acute calcium pyrophosphate crystal arthritis.

OBJECTIVES: Calcium pyrophosphate deposition (CPPD) disease, broadly defined, has been associated with increased risk of cardiovascular (CV) events. We investigated risk of CV events in patients with acute CPP crystal arthritis, the acute manifestation of CPPD. METHODS: Cohort study using Mass General Brigham electronic health record (EHR) data, 1991-2017. Patients with acute CPP crystal arthritis were identified using a published machine learning algorithm with positive predictive value 81%. Comparators were matched on year of EHR entry and index date of patients with acute CPP crystal arthritis (first positive synovial fluid CPP result or mention of 'pseudogout', or matched encounter). Major adverse cardiovascular event (MACE) was a composite of non-fatal CV event (myocardial infarction, acute coronary syndrome, coronary revascularisation, stroke) and death. We estimated incidence rates (IRs) and adjusted hazard ratios for MACE, non-fatal CV event and death, allowing for differential estimates during years 0-2 and 2-10. Sensitivity analyses included: (1) patients with acute CPP crystal arthritis diagnosed during outpatient visits, (2) patients with linked Medicare data, 2007-2016 and (3)patients matched on number of CV risk factors. RESULTS: We matched 1200 acute CPP crystal arthritis patients to 3810 comparators. IR for MACE in years 0-2 was 91/1000 person-years (p-y) in acute CPP crystal arthritis and 59/1000 p-y in comparators. In years 2-10, IR for MACE was 58/1000 p-y in acute CPP crystal arthritis and 53/1000 p-y in comparators. Acute CPP crystal arthritis was significantly associated with increased risk for MACE in years 0-2 (HR 1.32, 95% CI 1.01 to 1.73) and non-fatal CV event in years 0-2 (HR 1.92, 95% CI 1.12 to 3.28) and years 2-10 (HR 2.18, 95% CI 1.27 to 3.75), but not death. Results of sensitivity analyses were similar to the primary analysis; in the outpatient-only analysis, risk of non-fatal CVE was significantly elevated in years 2-10 but not in years 0-2. CONCLUSIONS: Acute CPP crystal arthritis was significantly associated with elevated short and long-term risk for non-fatal CV event.