A machine learning-based prediction model for gout in hyperuricemics: a nationwide cohort study.

OBJECTIVE: To develop a machine learning-based prediction model for identifying hyperuricemic participants at risk of developing gout. METHODS: A retrospective nationwide Israeli cohort study used the Clalit Health Insurance database of 473 124 individuals to identify adults 18 years or older with at least two serum urate measurements exceeding 6.8 mg/dl between January 2007 and December 2022. Patients with a prior gout diagnosis or on gout medications were excluded. Patients' demographic characteristics, community and hospital diagnoses, routine medication prescriptions and laboratory results were used to train a risk prediction model. A machine learning model, XGBoost, was developed to predict the risk of gout. Feature selection methods were used to identify relevant variables. The model's performance was evaluated using the receiver operating characteristic area under the curve (ROC AUC) and precision-recall AUC. The primary outcome was the diagnosis of gout among hyperuricemic patients. RESULTS: Among the 301 385 participants with hyperuricemia included in the analysis, 15 055 (5%) were diagnosed with gout. The XGBoost model had a ROC-AUC of 0.781 (95% CI 0.78-0.784) and precision-recall AUC of 0.208 (95% CI 0.195-0.22). The most significant variables associated with gout diagnosis were serum uric acid levels, age, hyperlipidemia, non-steroidal anti-inflammatory drugs and diuretic purchases. A compact model using only these five variables yielded a ROC-AUC of 0.714 (95% CI 0.706-0.723) and a negative predictive value (NPV) of 95%. CONCLUSIONS: The findings of this cohort study suggest that a machine learning-based prediction model had relatively good performance and high NPV for identifying hyperuricemic participants at risk of developing gout.

Obstetric antiphospholipid syndrome carries an increased lifetime risk for obstetric and thrombotic complications-a population-based study.

Background: Antiphospholipid syndrome (APS) can present with either a thromboembolic event (thrombotic APS, TAPS) or an obstetric complication (obstetric APS, OAPS). Data on long-term complications in the different APS phenotypes are limited. Objectives: We aimed to compare obstetric history, antiphospholipid antibody profiles, obstetric and thromboembolic complications, and pregnancy outcomes between TAPS and OAPS. Methods: This retrospective cohort study included women who delivered singleton pregnancies between 1998 and 2020. One hundred sixteen thousand four hundred nine women were included, resulting in 320,455 deliveries. Among the included patients, 71 were diagnosed with APS, 49 were classified as OAPS, and 22 as TAPS. The demographics, obstetric, neonatal, and thrombotic outcomes were compared among TAPS, OAPS, and the general obstetric population. Results: OAPS patients had an increased risk of thrombotic events compared with the general obstetric population (odds ratio [OR] 18.0; 95% CI, 8.7-37.2). In pregnancies following the diagnosis of APS, despite standard antithrombotic treatment, OAPS patients exhibited an elevated risk of placenta-related and neonatal complications compared with the general obstetric population (late fetal loss [adjusted OR {aOR}, 15.3; 95% CI, 0.5-27.5], stillbirth [aOR, 5.9; 95% CI, 2.2-15.4], placental abruption [aOR, 4.8; 95% CI, 1.5-15.3], preeclampsia [aOR, 4.4; 95% CI, 2.5-7.7], fetal growth restriction [aOR, 4.3; 95% CI, 8.5-27.5], small for gestational age neonate [aOR, 4.0; 95% CI, 2.4-6.6], and low Apgar scores [Apgar'1: aOR, 2.6; 95% CI, 1.3-10.4; Apgar'5: aOR, 3.7; 95% CI, 1.3-10.4]). TAPS patients exhibited increased risk of preeclampsia (aOR, 3.1; 95% CI, 1.2-8). Conclusion: OAPS patients exhibit a heightened risk of thrombotic events compared with the general obstetric population. Despite treatment, OAPS and TAPS still presented obstetric complications. These findings, after confirmation in prospective studies, need to be taken into consideration when planning the treatment approach for these patients.

An international multicenter study comparing COVID-19 omicron outcomes in patients with hematological malignancies treated with obinutuzumab versus rituximab.

OBJECTIVES: Hematological malignancy (HM) patients treated with anti-CD20 monoclonal antibodies are at higher risk for severe COVID-19. A previous single-center study showed worse outcomes in patients treated with obinutuzumab compared to rituximab. We examined this hypothesis in a large international multicenter cohort. METHODS: We included HM patients from 15 centers, from five countries treated with anti-CD20, comparing those treated with obinutuzumab (O-G) to rituximab (R-G) between December 2021 and June 2022, when Omicron lineage was dominant. RESULTS: We collected data on 1048 patients. Within the R-G (n = 762, 73%), 191 (25%) contracted COVID-19 compared to 103 (36%) in the O-G. COVID-19 patients in the O-G were younger (61 ± 11.7 vs. 64 ± 14.5, p = 0.039), had more indolent HM diagnosis (aggressive lymphoma: 3.9% vs. 67.0%, p < 0.001), and most were on maintenance therapy at COVID-19 diagnosis (63.0% vs. 16.8%, p < 0.001). Severe-critical COVID-19 occurred in 31.1% of patients in the O-G and 22.5% in the R-G. In multivariable analysis, O-G had a 2.08-fold increased risk for severe-critical COVID-19 compared to R-G (95% CI 1.13-3.84), adjusted for Charlson comorbidity index, sex, and tixagevimab/cilgavimab (T-C) prophylaxis. Further analysis comparing O-G to R-G demonstrated increased hospitalizations (51.5% vs. 35.6% p = 0.008), ICU admissions (12.6% vs. 5.8%, p = 0.042), but the nonsignificant difference in COVID-19-related mortality (n = 10, 9.7% vs. n = 12, 6.3%, p = 0.293). CONCLUSIONS: Despite younger age and a more indolent HM diagnosis, patients receiving obinutuzumab had more severe COVID-19 outcomes than those receiving rituximab. Our findings underscore the need to evaluate the risk-benefit balance when considering obinutuzumab therapy for HM patients during respiratory viral outbreaks.

Glucose variability is a marker for COVID-19 severity and mortality

ABSTRACT Objective: We aimed to investigate the association between glucose coefficient of variation (CV) and mortality and disease severity in hospitalized patients with coronavirus disease-19 (COVID-19). Subjects and methods: Retrospective cohort study in a tertiary center of patients with COVID-19 admitted to designated departments between March 11th, 2020, and November 2nd, 2020. We divided patients based on quartiles of glucose CV after stratification to those with and without diabetes mellitus (DM). Main outcomes were length of stay and in-hospital mortality. Results: The cohort included 565 patients with a mean age of 67.71 ± 15.45 years, and 62.3% were male. Of the entire cohort, 44.4% had DM. The median glucose CV was 32.8% and 20.5% in patients with and without DM, respectively. In patients with DM, higher glucose CV was associated with a longer hospitalization in the unadjusted model (OR = 2.7, 95% CI [1.3,5.6] for Q4), and when adjusted for age, sex, comorbidities, and laboratory markers, this association was no longer statistically significant (OR = 1.3, 95% CI [0.4,4.5] for Q4). In patients with and without DM, higher glucose CV was associated with higher rates of in-hospital mortality in the unadjusted model, but adjustment for comorbidities and laboratory markers eliminated the association (OR = 0.5, 95% CI [0.1,3.4] for Q4 in patients with DM). Conclusion: Higher glucose CV was associated with increased in-hospital mortality and length of stay, but this association disappeared when the adjustment included laboratory result data. Glucose CV can serve as a simple and cheap marker for mortality and severity of disease in patients with COVID-19.