Stefan Kleinert

Referral delays and errors in inflammatory rheumatic diseases (IRDs) are common. RhePort is a German online self-referral tool that estimates IRD probability based on a structured questionnaire and expert-derived algorithm. This study evaluated whether machine learning could improve IRD discrimination of RhePort. In this retrospective internal validation study, 1,333 unique RhePort questionnaires linked to rheumatologist-confirmed diagnoses were analyzed. A shared stratified five-fold cross-validation design was used for Logistic Regression, Neural Network, XGBoost, and LightGBM models. For each algorithm, feature selection and Bayesian hyperparameter optimization were performed, followed by evaluation of ensemble strategies. Discrimination, calibration, clinically relevant operating points, and SHAP-based interpretability were assessed against the original RhePort score. IRD prevalence was 35.6% (475/1,333). LightGBM achieved the strongest individual performance, with an area under the receiver operating characteristic curve of 0.791 using 30 features. A weighted ensemble of LightGBM and Logistic Regression further improved performance to 0.815 and achieved the lowest Brier score (0.166). At 95% sensitivity, specificity increased from 9% for RhePort to 32%; at 90% sensitivity, specificity increased from 15% to 48%. The most influential predictors included C-reactive protein, painful small joints, dactylitis pattern, sex, and erythrocyte sedimentation rate. Ensemble AUC-ROC was 0.828 among 746 patients with CRP and/or ESR information and 0.786 among 587 patients without these laboratory data. Machine learning improved IRD classification over the current expert-derived RhePort score and identified questionnaire items that may be omitted without loss of performance, demonstrating the potential of machine learning to enhance rheumatology referral.

Zusammenfassung Patienten mit rheumatoider Arthritis (RA), Psoriasisarthritis (PsA) und axialer Spondyloarthritis (axSpA) haben ein persistierend erhöhtes kardiovaskuläres (KV-)Risiko und eine höhere Mortalität, unabhängig von klassischen KV-Risikofaktoren. Eine konsequente Entzündungskontrolle reduziert KV-Ereignisse. Glukokortikoide erhöhen das Risiko dosis- und zeitabhängig bereits bei ≤ 5 mg Prednisolon pro Tag. „Disease-modifying antirheumatic drugs“, insbesondere Tumornekrosefaktor(TNF)-Inhibitoren, wirken über die Reduktion systemischer Inflammation protektiv. Bei der Therapie mit Januskinase(JAK)-Inhibitoren und Langzeittherapie mit Glukokortikoiden sind ein strukturiertes KV-Screening sowie die leitliniengerechte Behandlung von Risikofaktoren (inkl. Lipidmanagement/Statintherapie) besonders relevant. Die Primärprävention folgt den kardiologischen-Empfehlungen der European Society of Cardiology (ESC). Impfungen (Influenza, COVID-19, Pneumokokken, respiratorisches Synzytialvirus, Zoster) sind bei kardiovaskulären Risikogruppen eine wirksame Säule der KV-Prävention; Evidenz bei entzündlich-rheumatischen Erkrankungen fehlt bislang. Das Hauptproblem der KV-Prävention bleibt die Umsetzung: Digitale klinische Erinnerungen bzw. Entscheidungsunterstützungssysteme und Multikomponentenstrategien können hier die Implementierung von Empfehlungen verbessern.

Psoriatic arthritis (PsA) is a complex disease with heterogeneous disease manifestations. Composite scores including patient reported parameters are the recommended tools to measure these different aspects. Patient and physician often show discordance in their disease assessment. To describe the discordance between patient (PtGA) and physician reported global disease activity (PhGA) and to show its association with mental health. Descriptive cross-sectional analysis of three independent cohorts of PsA patients – RABBIT-SpA register, national database (NDB), RheumaDatenRheport GbR (RHADAR) network – receiving routine rheumatology care in Germany. Association between depressive symptoms and discordance between PtGA and PhGA was analysed using a linear regression model taking age, sex, and depression score into account. 1931 RABBIT-SpA, 1794 NDB, and 1728 RHADAR patients were analysed. Mean age was 55, 56, and 60 years. Mean discordance (PtGA-PhGA) was 1.8 in RABBIT-SPA, 1.9 in NDB, and 2 in RHADAR. Discordance > = 3 was present in 34% (RABBIT-SpA), 35% (NDB), and 37% (RHADAR). Discordance increased with increasing PtGA. In patients with severe depressive symptoms, discordance was larger than in those with no depressive symptoms. Depressive symptoms were significantly associated with discordance also if age and sex were accounted for. In this analysis of three independent PsA cohorts, we found a clinically meaningful discordance between physician and patient assessed disease activity in more than a third of the patients. As patients report higher levels of disease activity, discordant judgments become increasingly prevalent. Depressive symptoms are strongly associated with discordant assessments of global disease activity.

More efficient means of identifying patients with inflammatory rheumatic diseases (IRDs) could allow earlier diagnosis and treatment. The objective of this study was to evaluate the characteristics of a revised version of an online patient questionnaire-based self-referral tool, RhePort 1.3. This prospective study included adult patients with musculoskeletal complaints presenting for a first rheumatology visit at German RheumaDatenRhePort (RHADAR) rheumatology network centers. All patients completed the RhePort 1.3 questionnaire on patient characteristics and symptoms. Data from RhePort 1.3 were compared with historical data from previous versions. Of 614 patients, 225 (36.6%) were diagnosed with an IRD by a rheumatologist and 164/225 IRD patients (72.9%) had a RhePort 1.3 score > 1, the cut-off point used to determine the need for rheumatologic evaluation. A score > 1 was associated with an approximately two-fold higher IRD risk (odds ratio [95% confidence interval] of 1.98 [1.39, 2.83] vs ≤ 1) and had good sensitivity (73%) and moderate specificity (42%). Among patients referred through a standard referral pathway (n = 283), RhePort 1.3 scores > 1 in addition to physician referral were associated with increases in rheumatology-diagnosed IRD rates from 33.2% (physician referral only) to 45.7%. RhePort 1.3 had higher accuracy than earlier versions (54% vs 35%). We conclude that modest changes to the RhePort questionnaire resulted in increased accuracy. A score > 1 was associated with a doubled risk for an IRD and higher IRD rates in physician-referred patients. These data suggest that RhePort has the potential to streamline the rheumatologist’s workload and improve resource use. Further modifications are required to improve specificity.

Numerous biologic and targeted synthetic disease-modifying anti-rheumatic drugs (b/tsDMARDs) have been approved for treating rheumatoid arthritis (RA), including Janus kinase inhibitors (JAKi), rituximab, abatacept, interleukin-6 inhibitors (IL-6i) and tumor necrosis factor inhibitors (TNFi). Real-world data regarding treatment persistence and drug survival in the time period following the approval of JAKI are limited. To investigate the persistence and drug survival of different (b/tsDMARDs) in patients with RA based on real-world data from German outpatients. We performed a retrospective analysis of RA patients in the German RHADAR database who received a newly prescribed therapy with a b/tsDMARD between January 15, 2015 and October 17, 2023. To compare drug survival, we used Cox regression analyses adjusted for age, sex and disease duration. We included 4678 patients (79.9% female, mean age 58.7 years, mean disease duration 11.7 years). TNFi (47.8%) and JAKi (29.4%) were most frequently prescribed. The five-year drug survival rate was highest for JAKi (68.3%), followed by TNFi (58.6%), IL-6i (58.6%), abatacept (55.0%), and rituximab (53.3%). Compared to JAKi, Cox regression showed a higher discontinuation risk for rituximab (HR 1.36, p = 0.015), abatacept (HR 1.46, p < 0.001), IL-6i (HR 1.20, p = 0.026), and TNFi (HR 1.29, p < 0.001). Rituximab had the highest two-year survival (70.3%) but a sharp decline afterward, with 60% of discontinuations occurring between March 2020 and July 2022. In a real-world setting in German outpatients with RA, JAKi was associated with longer drug survival than bDMARDS. There were no significant differences in treatment persistence among different bDMARDs.

Objectives Although patients with immune-mediated inflammatory diseases (IMID) are thought to be more susceptible to viral infections, it is unclear whether their presentation differs between patients with IMID and healthy controls. This study aimed to investigate the symptom pattern of common viral infections in patients with IMID and compare it with controls without IMIDs. Design A cross-sectional study conducted between 1 February and 30 April 2020, using a questionnaire. Setting Seven tertiary regional care centers in Germany, which specialised in the care of patients with IMID (namely, in gastroenterology, dermatology, rheumatology and immunology clinical care). Participants One thousand nine hundred nine participants completed the survey (757 patients with IMID; 1152 non-IMID controls). Primary outcome measure The occurrence of 11 common viral illness symptoms within the preceding 3 months in patients with IMID and non-IMID controls. Results Symptom data were clustered, based on number and co-occurrance, into 3 major clusters and 2 subclusters ranked by the average number of symptoms. Patients with inflammatory bowel disease and psoriasis were significantly overrepresented in the lower-frequency subcluster of the polysymptomatic cluster. Patients with rheumatoid arthritis were overrepresented in the lower-frequency subclusters of the intermediate and oligo-/asymptomatic clusters. Controls were over-represented only in the higher-frequency subclusters of each major cluster where none of the IMIDs were over-represented. Spondyloarthritis and other IMIDs were also overrepresented in the low-frequency subcluster, but the results were not significant. Overall, patients with rheumatoid arthritis patients reported fewer symptoms (rate ratio=0.68, 95% CI, 0.59 to 0.80) than non-IMID controls. Conclusion Patients with IMID are over-represented in low-frequency subclusters, even among individuals who have reported a broad range of viral infection symptoms. This pattern suggests that the manifestations of viral infections are different between patients with IMID and controls, thus challenging the accurate and early diagnosis of infections.

In recent years Janus kinase inhibitors (JAKi) have joined tumor necrosis factor inhibitors (TNFi) and interleukin (IL)-17 inhibitors (IL-17i) as approved disease modifying anti-rheumatic drugs (DMARD) for moderate to severe forms of axial spondyloarthritis (axSpA). Drug survival in axSpA patients has not been well studied in a real-world outpatient scenario since the approval of JAKi. We aimed to analyze the three drug classes based on modes of actions (MoA) for their persistence rates among German axSpA outpatients. A retrospective analysis of the RHADAR database for axSpA patients with a new initiation of TNFi, IL-17i, or JAKi treatment between January 2015 and October 2023 was conducted. Analyses included Kaplan-Meier curves and adjusted Cox regressions for drug discontinuation. 1222 new biological DMARD (TNFi [n = 954], IL-17i [n = 190]) or JAKi (n = 78) treatments were reported. The median drug survival was 31 months for TNFi, 25 for IL-17i, and 18 for JAKi. The corresponding 2-year drug survival rate was 79.6%, 72.6%, and 62.8% for TNFi, IL-17i, and JAKi, respectively. The probability for discontinuation for JAKi was significantly higher compared with TNFi (HR 1.91 [95% CI 1.22–2.99]) as well as for IL-17i compared with TNFi (HR 1.43 [95% CI 1.02–2.01]), possibly related to more frequent use of TNFis as first-line therapy. IL-17i and JAKi discontinuation probabilities were similar. Primary non-response was the reason for drug discontinuation in most cases across all MoA. TNFi treatment might persist longer than JAKi and IL-17i in German axSpA outpatients, possibly related to more severe or refractory disease in patients with JAKi-treated or IL-17i-treated axSpA.

Background: The diagnosis of inflammatory rheumatic diseases (IRDs) is often delayed due to unspecific symptoms and a shortage of rheumatologists. Digital diagnostic decision support systems (DDSSs) have the potential to expedite diagnosis and help patients navigate the health care system more efficiently. Objective: The aim of this study was to assess the diagnostic accuracy of a mobile artificial intelligence (AI)–based symptom checker (Ada) and a web-based self-referral tool (Rheport) regarding IRDs. Methods: A prospective, multicenter, open-label, crossover randomized controlled trial was conducted with patients newly presenting to 3 rheumatology centers. Participants were randomly assigned to complete a symptom assessment using either Ada or Rheport. The primary outcome was the correct identification of IRDs by the DDSSs, defined as the presence of any IRD in the list of suggested diagnoses by Ada or achieving a prespecified threshold score with Rheport. The gold standard was the diagnosis made by rheumatologists. Results: A total of 600 patients were included, among whom 214 (35.7%) were diagnosed with an IRD. Most frequent IRD was rheumatoid arthritis with 69 (11.5%) patients. Rheport’s disease suggestion and Ada’s top 1 (D1) and top 5 (D5) disease suggestions demonstrated overall diagnostic accuracies of 52%, 63%, and 58%, respectively, for IRDs. Rheport showed a sensitivity of 62% and a specificity of 47% for IRDs. Ada’s D1 and D5 disease suggestions showed a sensitivity of 52% and 66%, respectively, and a specificity of 68% and 54%, respectively, concerning IRDs. Ada’s diagnostic accuracy regarding individual diagnoses was heterogenous, and Ada performed considerably better in identifying rheumatoid arthritis in comparison to other diagnoses (D1: 42%; D5: 64%). The Cohen κ statistic of Rheport for agreement on any rheumatic disease diagnosis with Ada D1 was 0.15 (95% CI 0.08-0.18) and with Ada D5 was 0.08 (95% CI 0.00-0.16), indicating poor agreement for the presence of any rheumatic disease between the 2 DDSSs. Conclusions: To our knowledge, this is the largest comparative DDSS trial with actual use of DDSSs by patients. The diagnostic accuracies of both DDSSs for IRDs were not promising in this high-prevalence patient population. DDSSs may lead to a misuse of scarce health care resources. Our results underscore the need for stringent regulation and drastic improvements to ensure the safety and efficacy of DDSSs. Trial Registration: German Register of Clinical Trials DRKS00017642; https://drks.de/search/en/trial/DRKS00017642

The categorization of axial spondyloarthritis (axSpA) into radiographic (r-axSpA) and non-radiographic (nr-axSpA) subtypes is important in clinical trials but may be of less value in clinical practice. This exploratory cross-sectional, multi-center study evaluated patients with axSpA under routine care at German clinical rheumatology sites (RHADAR real-world database), with a focus on imaging data used for diagnostic classifications. Our analyses included 371 patients with axSpA. The mean (standard deviation [SD]) age was 50.9 (14.0) years, disease duration was 16.4 (13.5) years, and 39.6% were female. Based on the rheumatologist’s final assessment, almost half of patients had definite r-axSpA (n = 179; 48.2%), 53 (14.3%) had suspected r-axSpA, 112 (30.2%) had non-radiographic-axSpA (nr-axSpA), and 27 (7.3%) had undefined axSpA. Patients assessed with definite or suspected r-axSpA were more likely to be treated with disease-modifying antirheumatic drugs (DMARDs) (62.0% and 64.2%, respectively) compared with nr-axSpA or undefined axSpA patients (37.5% and 48.1%, respectively). Almost all patients (348/371; 93.8%) had sacroiliac joint imaging data (radiographs or magnetic resonance imaging) documented in their charts, but only 216 (58.2%) had conventional radiographs required for formal diagnosis of r-axSpA by modified New York criteria. Follow-up radiographic imaging in nr-axSpA patients was uncommon (23/216 [25.0%]) but confirmed r-axSpA in 9/23 patients (39.1%). In conclusion, radiographs were available for slightly more than half of axSpA patients. Follow-up imaging was infrequent during rheumatology care in Germany but confirmed r-axSpA in ~ 40% of patients originally considered to have nr-axSpA. The distinction between r-axSpA and nr-axSpA may be ill-defined in routine clinical practice.

Several studies have shown that tapering or stopping disease-modifying anti-rheumatic drugs (DMARDs) in rheumatoid arthritis (RA) patients in sustained remission is feasible. However, tapering/stopping bears the risk of decline in physical function as some patients may relapse and face increased disease activity. Here, we analyzed the impact of tapering or stopping DMARD treatment on the physical function of RA patients. The study was a post hoc analysis of physical functional worsening for 282 patients with RA in sustained remission tapering and stopping DMARD treatment in the prospective randomized RETRO study. HAQ and DAS-28 scores were determined in baseline samples of patients continuing DMARD (arm 1), tapering their dose by 50% (arm 2), or stopping after tapering (arm 3). Patients were followed over 1 year, and HAQ and DAS-28 scores were evaluated every 3 months. The effect of treatment reduction strategy on functional worsening was assessed in a recurrent-event Cox regression model with a study-group (control, taper, and taper/stop) as the predictor. Two-hundred and eighty-two patients were analyzed. In 58 patients, functional worsening was observed. The incidences suggest a higher probability of functional worsening in patients tapering and/or stopping DMARDs, which is likely due to higher relapse rates in these individuals. At the end of the study, however, functional worsening was similar among the groups. Point estimates and survival curves show that the decline in functionality according to HAQ after tapering or discontinuation of DMARDs in RA patients with stable remission is associated with recurrence, but not with an overall functional decline.