Pragmatic and Observational Research最新文献

Background: Epidemiological studies rely on valid classifications of patients' disease status. However, in the absence of perfect information on every patients' health status, researchers use proxy information with variable and often unknown validity.

Methods: To investigate the validity of using prescription records for disease classification, we conducted a retrospective observational study on a UK-wide database of medical prescriptions and clinical records (Optimum Patient Care Research Database). We used electronic health records of 25,000 randomly selected patients for each year between 2004 and 2020 (total N=425,000) and compared disease classification of 18 different chronic conditions based on clinical records for a period of three years (gold standard) with disease classification based on prescription records for the same period. We then used logistic regression to analyse if positive and negative predicted values (PPV and NPV) were associated with known predictors of disease.

Results: Results showed large variations in PPV ranging from 8% (heart failure) to 94% (all type diabetes) and smaller variations in NPV ranging from 96% (anxiety) to 100% (Type 1 diabetes). Age, sex, ethnicity, and year but not socio-economic status were associated with variations in validity, especially in classifying dementia, diabetes, and depression.

Discussion: Varying validity can partly be explained by different (stratum-specific) prevalence of disease. Additionally, conditions like heart failure can be treated with medication that can also be prescribed for other conditions or can be treated without medication. However, varying validity can also be attributed to imperfect clinical records, which we used as gold standard. As a consequence of low validity, the apparent prevalence based on using prescription records was between 1.3 times lower (all-type diabetes) and up-to 11 times higher (heart failure) than the true prevalence based on the clinical records.

Conclusion: Studies using prescription data to classify disease status run a substantial risk of misclassification bias.

Objective: This study aims to identify predictive factors for the relapse of idiopathic retroperitoneal fibrosis (IRF) and provide instructions for the optimization of the maintenance therapy.

Methods: All patients with a clinical diagnosis of IRF were enrolled and followed up every 3-6 months. Their clinical characteristics, laboratory data and treatment strategies were recorded at each visit.

Results: 96 IRF patients (77 males and 19 females) with a median age of 55 years (interquartile range [IQR], 50-61) were enrolled. The median follow-up time was 2.50 (IQR, 1.75-4.13) years. During the follow-up, 21 patients experienced at least one relapse, with cumulative relapse rates of 10.6%, 32.3%, and 62.4% at 2.5, 5, and 7.5 years, respectively. Initial hydronephrosis was an independent predictor of relapse (Hazard ratio [HR], 5.35; p=0.001). Discontinuation of maintenance therapy (HR, 3.41; 95% CI, 1.4-8.314; p=0.007) was closely associated with relapse. Use (HR, 0.12; p<0.001) and dose (HR, 0.73; p=0.01) of glucocorticoids (GC) in maintenance period were protective factors against relapse. Among patients with hydronephrosis, those who discontinued GC had a higher relapse rate (p=0.009). GC monotherapy or combined immunosuppressants (IM) therapy were more effective in preventing relapse than IM alone (p<0.001).

Conclusion: Our study reveals that initial hydronephrosis and GC withdrawal during maintenance therapy are significant predictors of IRF relapse. Long-term, low-dose GC therapy is benefit for maintaining remission and preventing relapse, especially in patients with initial hydronephrosis.

Purpose: Asthma is a heterogenous disease, and up to 10% of asthma patients have severe disease. While critical, clinical trials in these patients are generally not representative of the broader ambulatory patient population, warranting the need for reliable and rich real-world data to enable clinical, payer, and regulatory decision-making. Currently, observational registries of consented patients rely on longitudinal, clinically rich data to support real-world evidence generation, but are resource intensive. This manuscript describes an asthma pragmatic registry that incorporates rigorous data reliability standards while offering scalability for research.

Patients and methods: The INSIGHTS Asthma Pragmatic Registry is a real-world, retrospective cohort of adult and adolescent patients with moderate-to-severe asthma geographically distributed throughout the United States. The data set is purposely designed to meet data reliability standards of accuracy, completeness, and traceability. Efforts include continuous improvements in the capture of key clinical features (including disease severity) from electronic health records using advanced technology, linkage across electronic health records, medical and pharmacy claims, and death registry, and process to trace data elements back to a source of truth.

Results: The pragmatic registry implementation includes patients meeting moderate-to-severe asthma eligibility criteria since Jan 1, 2014, with routine data refreshes, resulting in 9,185 patients meeting eligibility criteria through August 1, 2024. The median age of patients is 53 years. At least two-thirds of patients are female, of white race, and non-Hispanic ethnicity. On average, patients were followed for more than five years.

Conclusion: The INSIGHTS Asthma Pragmatic Registry is a new paradigm in observational research, which blends the data reliability and richness of traditional, consented registries while incorporating the flexibility and scale of utilizing routinely collected data. Ongoing efforts include maintaining longitudinality of patient data and improvements to sustain high data reliability per regulatory standards to support this unique data set for asthma research.

Background: Standardized variables and their definitions are essential for robust delivery and reporting of clinical studies and quality improvement in chronic obstructive pulmonary disease (COPD). This protocol describes the rationale and methodology for the derivation and definition of clinical study variables for patients with multimorbidity or at high risk of multimorbidity in COPD.

Methods and analysis: We will follow a four-step process. This will include the formation of an Executive Committee, a Steering Committee and an International Consensus Group. We will conduct a systematic review of the literature from which potential clinical study variables will be extracted and their definitions proposed. Using a modified Delphi process, the Steering Committee will select candidate clinical study variables and as necessary refine their definitions. All three groups will then vote and give feedback on the candidate clinical study variables in a modified Delphi process (with rounds until consensus is achieved) to reach a final suite of internationally agreed hierarchically classified clinical study variables.

Ethics and dissemination: It is anticipated that the results will be published in a peer-reviewed journal and presented in a variety of forums. Ethical approval was not required for this study.

Introduction: Multiple sclerosis (MS) is a major demyelinating disease causing disability in young adults. Scarce data exists on the clinical benefits and persistence of disease modifying therapies (DMTs) amongst our local population. The primary aim of this study was to investigate the effectiveness, tolerability and discontinuation pattern of various DMTs amongst Hong Kong patients.

Methods: This was a retrospective study of adult MS patients treated in a tertiary hospital in Hong Kong from 1st December 2012 to 30th November 2022. Demographic, clinical and radiological characteristics were retrieved.

Results: A total of 103 patients were analysed. Interferons (annualised relapse rate (ARR) reduction 0.492, p <0.001) and fingolimod (ARR reduction 0.557, p = 0.038) had significant ARR reductions in our cohort. Overall treatment persistence were comparable, but teriflunomide had more discontinuation due to side effects and/or intolerance (hazard ratio (HR) 7.50, p = 0.029). Baseline Expanded Disability Status Score (EDSS) more than or equal to 4.0 (odds ratio (OR) 4.56, p = 0.045), and presence of brainstem lesions on magnetic resonance imaging (OR 3.15, p = 0.039) were associated with greater disability progression, while visual onset symptoms at diagnosis (OR 0.14, p = 0.007) were associated with lower risk.

Discussion: This study illustrated the complexity of managing MS patients. Patients' clinical characteristics, disease activity, risk appetite and treatment side effects should be taken in consideration to reach an informed decision on the use of DMTs in our local MS patients.

Background: Cardiovascular disease (CVD) continues to be a significant health threat to humans globally, and a significant burden on healthcare systems. Cardiovascular risk prediction utilizing machine learning (ML) models in patients with asthma remains vastly underexplored.

Methods: In this cohort study consisting of 641,042 participants, we used routinely collected electronic healthcare record data to explore various ML algorithms including logistic regression, penalized logistic regression, decision trees, random forest and gradient boost to develop a model with high specificity.

Results: The penalized logistic regression model was identified to be the best and simplest classification model in terms of discriminatory power (AUC = 0.85). The gradient boost model was found to be the best predictive model in terms of calibration where the predicted and observed probabilities at risk of CVD match or are closely aligned. In all models, the number of previous cardiovascular events was the most influential predictor, followed by age and prescriptions related to cardiovascular medications. The top predictor alone produced a reasonable level of predictive power (AUC = 0.66).

Conclusion: We have created a novel prediction model for predicting CVD within a year of asthma diagnosis for patients with asthma at least 50 years old. Using penalized logistic regression, we achieved a high level of accuracy. By implementing this model, it would be possible to screen out patients with low risk of CVD with high specificity and acceptable sensitivity. Penalized logistic regression and gradient boost models have similar accuracy in screening out individuals at low risk of CVD. For this objective, penalized logistic regression may be more suitable than gradient boost models for implementation as it is simpler to use and more transparent. At the probability threshold of 8% (outcome prevalence), both models' effectiveness in reducing unnecessary treatments was by approximately 52%. These ML models performed better compared with traditional statistical-based risk prediction models. The unique contribution of the study is the construction of prediction models for CVD disease within 12 months from asthma diagnosis based on regression and machine learning models and the comparison of their accuracy to identify the best model based on suitable statistical measures such as AUC and calibration. Further prospective studies using different populations and external validation are required to assess and validate the ML risk prediction models.

Purpose: This real-world analysis described the characteristics and therapeutic management of patients with metastatic castration-resistant prostate cancer (mCRPC) in Italy before and after 2015, when androgen receptor signalling inhibitors (ARPI) entered clinical practice.

Patients and methods: An observational retrospective analysis was conducted using administrative healthcare databases from a pool of Italian Local Health Units, covering ~6.2 million residents. Adult men with ≥1 prescription of androgen deprivation therapy (ADT) from January 2011 to June 2022 were identified. mCRPC was proxied by treatment patterns-addition of docetaxel/cabazitaxel or ARPI to ADT-and confirmed by hospital discharge for metastasis. Patients were stratified according to their inclusion (date of the last inclusion criterion met): pre-2015 (2011-2014) and post-2015 (2015-2020).

Results: Among 1890 mCRPC patients identified, 551 (29%) received ≥2 treatment lines. Chemotherapy (CHT) was the predominant first-line (1L) therapy in both cohorts [97.2% vs 82.9%], followed by ARPI [2.8% vs 17.1%]. In a 2020 sensitivity analysis (n=406), 1L therapy was ARPI in 76% and CHT in 24%. Among pre- and post-2015 patients, 29.4% and 31.6% received second-line (2L) therapy, mostly ARPI. Median OS from ADT initiation was 46.4 months (pre-2015) and 53.9 months (post-2015). In post-2015 patients, median OS from mCRPC index-date-reflecting the proxy-based diagnosis-was 14.2 months.

Conclusion: In Italian clinical practice, CHT remains the most common 1L therapy though ARPI use has increased since 2015. While OS from ADT initiation has improved, survival from mCRPC diagnosis-based on proxies in administrative data-remains poor, underscoring an unmet clinical need. Differences in OS estimates depending on the starting point (ADT initiation vs mCRPC diagnosis) should be considered when interpreting results.

Purpose: Recent developments in electroanatomical mapping (EAM), intracardiac echocardiography, and sheath technology have allowed for a pronounced reduction or complete elimination of fluoroscopy during catheter ablation of atrial fibrillation (AF). This real-world study evaluates the procedural efficiency, clinical effectiveness, and safety of a zero- to minimal-fluoroscopy workflow in an exclusively persistent AF (PsAF) population.

Methods: Data on consecutive PsAF catheter ablations performed with a fluoroscopy minimization workflow by three operators at a single high-volume center in the United States between January 2017 and December 2018 were collected and analyzed. All procedures were performed with EAM guidance and a contact force ablation catheter. Patients were followed for a year post-ablation for safety, arrhythmia recurrences, and reablation. Outcomes of interest included procedural efficiency measures, single-procedure success (freedom from reablation at any time or recurrence after 90-day blanking), and procedure-related complications.

Results: The study included 406 PsAF patients (age 67.8 ± 10.1 years, 65.3% male, and CHA₂DS₂-VASc score 2.9 ± 1.5). Over 85% of ablations were performed without fluoroscopy, and ablations utilizing fluoroscopy averaged only 0.6 ± 1.5 minutes for a mean fluoroscopy time of 0.1 ± 0.6 minutes overall. Mean procedure time was 89.5 minutes, with 96.3% of the procedures including ablation beyond the pulmonary veins. Single-procedure success was 73.6% (95% confidence interval: [68.7%, 77.8%]). Eight patients (2.0%) experienced a procedure-related complication.

Conclusion: Minimal-fluoroscopy ablation was performed safely and without compromise to procedural efficiency or clinical effectiveness in a real-world population of PsAF patients, despite more extensive ablation than a typical paroxysmal AF population.

Integrating real-world evidence (RWE) into evidence-based medicine (EBM) enhances healthcare decision-making. RWE provides insights into the real-world effectiveness and safety of therapies and health technologies, filling gaps that clinical trials may leave. EBM, which concentrates on therapeutic issues, depends on rigorous evaluation of evidence, including data from randomized controlled trials (RCTs) and RWE. Combining evidence from RCTs and RWE when forming recommendations offers a comprehensive understanding of benefits and risks by considering their strengths, limitations, and standardized methods. The 2nd European Academy of Allergy & Clinical Immunology/Respiratory Effectiveness Group (EAACI/REG) Workshop, held in Rome, Italy, on October 4th, 2023, discussed integrating RWE and EBM. The goals were to develop recommendations for high-quality RWE and its inclusion in evidence syntheses, with a particular focus on airway diseases. During the discussion, key topics emerged. An "action plan" is needed to share these topics in various formats. RCTs are currently seen as providing the strongest evidence, so how to incorporate Non-Randomized Studies of Interventions (NRSI) requires careful consideration. An educational plan and collaboration with patients' organizations are also very important. A collaborative approach involving patients, clinicians, and regulators is essential for achieving meaningful results and can be adapted as needed for cultural differences. A "glossary" of terms used in this context will be created to improve understanding. Setting benchmarks for data quality and reliability, such as quality thresholds, in disease-specific studies requires collaboration with research method experts. Managing and recording registries according to standardized protocols and quality standards from well-designed registries will ensure the data is valid and accurate.

Purpose: The New User Design can be applied if the target drug has not been administered for a specified period. Therefore, comparisons between drugs administered alone are easier to undertake than comparisons of drugs used in combination. Thus, assessing concomitant medications may be associated with several challenges, including limitations to the New User Design. One such limitation is performing analyses that consider the history of administration of drugs of the same class. In the present study, we considered the limitations of the New User Design and proposed solutions based on the potential of the Prevalent New User Design.

Patients and methods: Using the Japan Medical Data Center database (JMDC), patients diagnosed with diabetes mellitus who received sulfonylureas (SUs) between December 2009 and December 2010 with subsequent addition or switch to dipeptidyl peptidase-4 inhibitors (DPP4Is) were categorized into the SU+DPP4I group. The odds ratio (OR) was estimated using conditional logistic regression analysis. Using the "elapsed time" and "number of prescriptions" axes of the Prevalent New User Design, records from 1,426 and 1,342 individuals, respectively, were analyzed.

Results: The hypoglycemia risk ORs were 1.50 (95% confidence interval [CI] 0.25-9.00) for the "elapsed time" axis and 1.67 (95% CI 0.40-7.00) for the "number of prescriptions" axis. These findings are consistent with the results of a meta-analysis of previous randomized controlled trials.

Conclusion: Our findings suggest that the Prevalent New User Design can be effectively applied for real-world risk assessment scenarios; this design constitutes a potential alternative design to the New User Design. We adopted a Prevalent New User Design considering the patients' treatment history. However, there was a limitation in that we could not obtain information regarding the patients' perceptions of treatment prior to initiating therapy.