Search

Background: Over the last few decades, the efficacy of biological therapies for COPD has been evaluated by different randomized controlled trials (RCTs). Still, the evaluation of real-world data and patient-reported outcome measures (PROMs) have not been performed in this field before. In the current work, we present a systematic literature review of the real-world data and PROMs of biological treatments for COPD. Methods: Three large databases (MEDLINE/PubMed, Scopus, and ScienceDirect) were utilized for the systematic literature review. Clinical studies (RCT, cohorts, case series/reports) assessing patients with COPD treated by any biological therapy were included. Results: The review resulted in twelve eligible studies (nine randomized controlled trials and three “real-world” case series/reports). The evaluation of PROMs in the included studies was mainly limited to the severity and burden of respiratory symptoms. Most biological therapies were associated with improved PROMs compared to the baseline, although not for the placebo. Dupilumab was the only biologic therapy with proven efficacy in RCT for both objective and subjective measures. One prior study reported patients’ self-perceived drug effects, and none evaluated patients’ perceived disease status. Only 25 patients were assessed in a real-world setting for all biologic therapies combined. Real-world data were retrospective in the form of case reports or series. Conclusions: There are limited data on patients’ experience with biological therapies for COPD. While real-world data and PROMs are missing, biases such as a placebo effect must be considered, requiring their incorporation with objective outcomes from prospective controlled trials.

Introduction: Convolutional Neural Network (CNN) systems in healthcare are influenced by unbalanced datasets and varying sizes. This article delves into the impact of dataset size, class imbalance, and their interplay on CNN systems, focusing on the size of the training set versus imbalance—a unique perspective compared to the prevailing literature. Furthermore, it addresses scenarios with more than two classification groups, often overlooked but prevalent in practical settings. Methods: Initially, a CNN was developed to classify lung diseases using X-ray images, distinguishing between healthy individuals and COVID-19 patients. Later, the model was expanded to include pneumonia patients. To evaluate performance, numerous experiments were conducted with varied data sizes and imbalance ratios for both binary and ternary classifications, measuring various indices to validate the model’s efficacy. Results: The study revealed that increasing dataset size positively impacts CNN performance, but this improvement saturates beyond a certain size. A novel finding is that the data balance ratio influences performance more significantly than dataset size. The behavior of three-class classification mirrored that of binary classification, underscoring the importance of balanced datasets for accurate classification. Conclusions: This study emphasizes the fact that achieving balanced representation in datasets is crucial for optimal CNN performance in healthcare, challenging the conventional focus on dataset size. Balanced datasets improve classification accuracy, both in two-class and three-class scenarios, highlighting the need for data-balancing techniques to improve model reliability and effectiveness. Motivation: Our study is motivated by a scenario with 100 patient samples, offering two options: a balanced dataset with 200 samples and an unbalanced dataset with 500 samples (400 healthy individuals). We aim to provide insights into the optimal choice based on the interplay between dataset size and imbalance, enriching the discourse for stakeholders interested in achieving optimal model performance. Limitations: Recognizing a single model’s generalizability limitations, we assert that further studies on diverse datasets are needed.

Hypersensitivity pneumonitis (HP) is a heterogeneous interstitial lung disease (ILD) that may be difficult to confidently diagnose. Recently, the 2020 ATS/JRS/ALAT HP diagnostic guidelines were published, yet data validating their performance in real-life settings are scarce. We aimed to assess the diagnostic performance of the HP guidelines compared to the gold-standard multidisciplinary discussion (MDD). For this purpose, we included consecutive ILD patients that underwent diagnostic bronchoscopy between 2017 and 2020 in three large medical centers. Four diagnostic factors (antigen exposure history, chest computed tomography pattern, bronchoalveolar lavage lymphocyte count, and histology results) were used to assign guidelines-based HP diagnostic confidence levels for each patient. A sensitivity analysis was performed, with MDD diagnosis as the reference standard. Overall, 213 ILD patients were included, 45 (21%) with an MDD diagnosis of HP. The guidelines’ moderate (≥70%) confidence threshold produced optimal performance with 73% sensitivity for HP, 89% specificity, and a J-index of 0.62. The area under the receiver operating characteristic curve (AUC) for a correct guidelines-based diagnosis was 0.86. The guidelines had better performance for non-fibrotic than fibrotic HP (AUC 0.92 vs. 0.82). All diagnostic factors, except bronchoalveolar lavage lymphocyte count, were independent predictors for MDD diagnosis of HP in a multivariate analysis. In conclusion, the HP guidelines exhibited a good diagnostic performance compared to MDD diagnosis in real-life setting.