Explore the vast range of titles available.

Much has been written about real-world evidence (RWE), a concept that offers an understanding of the effects of healthcare interventions using routine clinical data. The reflection of diverse real-world practices is a double-edged sword that makes RWE attractive but also opens doors to several biases that need to be minimised both in the design and analytical phases of non-experimental studies. Additionally, it is critical to ensure that researchers who conduct these studies possess adequate methodological expertise and ability to accurately implement these methods. Critical design elements to be considered should include a clearly defined research question using a causal inference framework, choice of a fit-for-purpose data source, inclusion of new users of a treatment with comparators that are as similar as possible to that group, accurately classifying person-time and deciding censoring approaches. Having taken measures to minimise bias ‘by design’, the next step is to implement appropriate analytical techniques (for example propensity scores) to minimise the remnant potential biases. A clear protocol should be provided at the beginning of the study and a report of the results after, including caveats to consider. We also point the readers to readings on some novel analytical methods as well as newer areas of application of RWE. While there is no one-size-fits-all solution to evaluating RWE studies, we have focused our discussion on key methods and issues commonly encountered in comparative observational cohort studies with the hope that readers are better equipped to evaluate non-experimental studies that they encounter in the future.

Cold Start problems in recommender systems pose various challenges in the adoption and use of recommender systems, especially for new item uptake and new user engagement. This restricts organizations to realize the business value of recommender systems as they have to incur marketing and operations costs to engage new users and promote new items. Owing to this, several studies have been done by recommender systems researchers to address the cold start problems. However, there has been very limited recent research done on collating these approaches and algorithms. To address this gap, the paper conducts a systematic literature review of various strategies and approaches proposed by researchers in the last decade, from January 2010 to December 2021, and synthesizes the same into two categories: data-driven strategies and approach-driven strategies. Furthermore, the approach-driven strategies are categorized into five main clusters based on deep learning, matrix factorization, hybrid approaches, or other novel approaches in collaborative filtering and content-based algorithms. The scope of this study is limited to a systematic literature review and it does not include an experimental study to benchmark and recommend the best approaches and their context of use in cold start scenarios.

Background Direct oral anticoagulants (DOACs) are preferred for stroke prevention in nonvalvular atrial fibrillation (NVAF); however, warfarin is still used. This study examined why physicians may choose warfarin over DOACs and the associated safety outcomes in patients with NVAF. Methods We conducted a new‐user, active‐comparator cohort study in newly diagnosed patients with NVAF to assess safety outcomes after the introduction of DOACs in Japan. Results The median observation period was 1120 days; 1428 patients started anticoagulation therapy with warfarin and 1551 with DOACs. Warfarin was chosen for patients with lower creatinine clearance and left ventricular ejection fractions and those using aspirin and verapamil. The unadjusted risk of major bleeding was considerably higher in the warfarin group but was nonsignificant after adjusting for variables associated with the choice of warfarin, in addition to age and sex. The risk of death was higher in the warfarin group, even after adjustments for relevant variables. However, high‐risk subgroups, including those with older ages and multiple comorbidities, such as renal impairment, for whom warfarin was more likely to be selected, had severely compromised prognoses with either anticoagulant. The risk of stroke/systemic embolism was not significantly different between the two groups. Conclusions Warfarin is often chosen for older patients with multiple comorbidities characterized by reduced renal function, which is associated with a higher risk of major bleeding and mortality. These high‐risk patients seem to have a poor prognosis regardless of the type of anticoagulant used. Thus, safe anticoagulant therapy remains a challenge for such patients. This new‐user cohort study explored the reasons physicians opt for warfarin over DOACs in nonvalvular atrial fibrillation (NVAF) patients and associated safety outcomes. It found that older, high‐risk patients who warfarin was often prescribed were linked to increased bleeding and mortality risks, regardless of anticoagulant type.

Research on the problem of new user cold-start recommendation generally leverages user side information to suggest items to new users. This approach, however, is impractical due to privacy concerns. In this paper, we propose ColdGAN, an end-to-end GAN-based recommendation system that makes no use of side information to resolve the new user cold-start recommendation problem. The proposed ColdGAN explores the merit of GAN that enables precise data generation given imprecise data. Our generative network learns to predict item ratings that cold-start users would make in the future given their limited rating behavior data. The predicted ratings are evaluated by the discriminative network trained for determining whether the ratings are precise enough. Moreover, a novel rejuvenation function and relevant item loss are incorporated into ColdGAN to enhance the predictions made by the learned generative network. Experiments based on three real-world datasets demonstrate that ColdGAN significantly outperforms many state-of-the-art recommendation systems. Also, our designed rejuvenation function and relevant item loss are effective in guiding our generative network to infer item ratings of cold-start new users.

The benefit of secondary prevention in hypertensive patients with peripheral arterial disease (PAD) is based on continual simultaneous taking of statins, antiplatelet agents and antihypertensive agents, preferably angiotensin-converting enzyme inhibitors (ACEIs) or angiotensin receptor blockers (ARBs). Our study was aimed at a) the analysis of the extent of non-persistence with multiple medication classes, and b) identifying factors associated with the likelihood of non-persistence. In our cohort study, 3,401 hypertensive patients (1,853 females and 1,548 males) aged ≥65 years treated simultaneously with statins, antiplatelet agents and ACEIs/ARBs and in whom PAD was newly diagnosed during 2012 were analysed. A patient was classified as non-persistent when he/she was non-persistent with at least one of the three analysed medication classes. The most important characteristics associated with the probability of non-persistence were identified using the Cox regression. At the end of the follow-up period (mean length 1.8 years), 1,869 (55.0%) patients (including 1,090 females and 779 males) were classified as non-persistent. In the whole study cohort, factors associated with non-persistence were female sex, atrial fibrillation, and being a new user of at least one of the analysed medication classes; in males, they were university education, atrial fibrillation, and epilepsy, and, in females, being a new user. Identification of sex differences in factors associated with non-persistence makes it possible to determine the groups of patients in whom special attention should be paid to improving their persistence with a combination of medicines in order to ensure successful secondary prevention of PAD.

Recommender systems are tools that help users in the decision-making process of choosing items that may be relevant for them among a vast amount of other items. One of the main problems of recommender systems is the cold start problem, which occurs when either new items or new users are added to the system and, therefore, there is no previous information about them. This article presents a multi-source dataset optimized for the study and the alleviation of the cold start problem. This dataset contains info about the users, the items (movies), and ratings with some contextual information. The article also presents an example user behavior-driven algorithm using the introduced dataset for creating recommendations under the cold start situation. In order to create these recommendations, a mixed method using collaborative filtering and user-item classification has been proposed. The results show recommendations with high accuracy and prove the dataset to be a very good asset for future research in the field of recommender systems in general and with the cold start problem in particular.

Aim/Purpose: This study aims to develop a solution for personalized tourism recommendations that addresses information overload, data sparsity, and the cold-start problem. It focuses on enabling tourists to choose the most suitable tourism-related facilities, such as restaurants and hotels, that match their individual needs and preferences. Background: The tourism industry is experiencing a significant shift towards digitalization due to the increasing use of online platforms and the abundance of user data. Travelers now heavily rely on online resources to explore destinations and associated options like hotels, restaurants, attractions, transportation, and events. In this dynamic landscape, personalized recommendation systems play a crucial role in enhancing user experience and ensuring customer satisfaction. However, existing recommendation systems encounter major challenges in precisely understanding the complexities of user preferences within the tourism domain. Traditional approaches often rely solely on user ratings, neglecting the complex nature of travel choices. Data sparsity further complicates the issue, as users might have limited interactions with the system or incomplete preference profiles. This sparsity can hinder the effectiveness of these systems, leading to inaccurate or irrelevant recommendations. The cold-start problem presents another challenge, particularly with new users who lack a substantial interaction history within the system, thereby complicating the task of recommending relevant options. These limitations can greatly hinder the performance of recommendation systems and ultimately reduce user satisfaction with the overall experience. Methodology: The proposed User-based Multi-Criteria Trust-aware Collaborative Filtering (UMCTCF) approach exploits two key aspects to enhance both the accuracy and coverage of recommendations within tourism recommender systems: multi-criteria user preferences and implicit trust networks. Multi-criteria ratings capture the various factors that influence user preferences for specific tourism items, such as restaurants or hotels. These factors surpass a simple one-star rating and take into account the complex nature of travel choices. Implicit trust relationships refer to connections between users that are established through shared interests and past interactions without the need for explicit trust declarations. By integrating these elements, UMCTCF aims to provide more accurate and reliable recommendations, especially when data sparsity limits the ability to accurately predict user preferences, particularly for new users. Furthermore, the approach employs a switch hybridization scheme, which combines predictions from different components within UMCTCF. This scheme leads to a more robust recommendation strategy by leveraging diverse sources of information. Extensive experiments were conducted using real-world tourism datasets encompassing restaurants and hotels to evaluate the effectiveness of UMCTCF. The performance of UMCTCF was then compared against baseline methods to assess its prediction accuracy and coverage. Contribution: This study introduces a novel and effective recommendation approach, UMCTCF, which addresses the limitations of existing methods in personalized tourism recommendations by offering several key contributions. First, it transcends simple item preferences by incorporating multi-criteria user preferences. This allows UMCTCF to consider the various factors that users prioritize when making tourism decisions, leading to a more comprehensive understanding of user choices and, ultimately, more accurate recommendations. Second, UMCTCF leverages the collective wisdom of users by incorporating an implicit trust network into the recommendation process. By incorporating these trust relationships into the recommendation process, UMCTCF enhances its effectiveness, particularly in scenarios with data sparsity or new users with limited interaction history. Finally, UMCTCF demonstrates robustness towards data sparsity and the cold-start problem. This resilience in situations with limited data or incomplete user profiles makes UMCTCF particularly suitable for real-world applications in the tourism domain. Findings: The results consistently demonstrated UMCTCF’s superiority in key metrics, effectively addressing the challenges of data sparsity and new users while enhancing both prediction accuracy and coverage. In terms of prediction accuracy, UMCTCF yielded significantly more accurate predictions of user preferences for tourism items compared to baseline methods. Furthermore, UMCTCF achieved superior coverage compared to baseline methods, signifying its ability to recommend a wider range of tourism items, particularly for new users who might have limited interaction history within the system. This increased coverage has the potential to enhance user satisfaction by offering a more diverse and enriching set of recommendations. These findings collectively highlight the effectiveness of UMCTCF in addressing the challenges of personalized tourism recommendations, paving the way for improved user satisfaction and decision-making within the tourism domain. Recommendations for Practitioners: The proposed UMCTCF approach offers a potential opportunity for tourism recommendation systems, enabling practitioners to create solutions that prioritize the needs and preferences of users. By incorporating UMCTCF into online tourism platforms, tourists can utilize its capabilities to make well-informed decisions when selecting tourism-related facilities. Furthermore, UMCTCF’s robust design allows it to function effectively even in scenarios with data sparsity or new users with limited interaction history. This characteristic makes UMCTCF particularly valuable for real-world applications, especially in scenarios where these limitations are common obstacles. Recommendation for Researchers: The success of UMCTCF can open up new avenues in personalized recommendation research. One promising direction lies in exploring the integration of additional contextual information, such as temporal (time-based) or location-based information. By incorporating these elements, the model could be further improved, allowing for even more personalized recommendations. Furthermore, exploring the potential of UMCTCF in domains other than tourism has considerable significance. By exploring its effectiveness in other e-commerce domains, researchers can broaden the impact of UMCTCF and contribute to the advancement of personalized recommendation systems across various industries. Impact on Society: UMCTCF has the potential to make a positive impact on society in various ways. By delivering accurate and diverse recommendations that are tailored to individual user preferences, UMCTCF fosters a more positive and rewarding user experience with tourism recommendation systems. This can lead to increased user engagement with tourism platforms, ultimately enhancing overall satisfaction with travel planning. Furthermore, UMCTCF enables users to make more informed decisions through broader and more accurate recommendations, potentially reducing planning stress and leading to more fulfilling travel experiences. Future Research: Expanding upon the success of UMCTCF, future research activities can explore several promising paths. Enriching UMCTCF with various contextual data, such as spatial or location-based data, to enhance recommendation accuracy and relevance. Leveraging user-generated content, like reviews and social media posts, could provide deeper insights into user preferences and sentiments, improving personalization. Additionally, applying UMCTCF in various e-commerce domains beyond tourism, such as online shopping, entertainment, and healthcare, could yield valuable insights and enhance recommendation systems. Finally, exploring the integration of optimization algorithms could improve both recommendation accuracy and efficiency.

Recommender systems (RSs) have become key components driving the success of e-commerce and other platforms where revenue and customer satisfaction is dependent on the user’s ability to discover desirable items in large catalogues. As the number of users and items on a platform grows, the computational complexity and the sparsity problem constitute important challenges for any recommendation algorithm. In addition, the most widely studied filtering-based RSs, while effective in providing suggestions for established users and items, are known for their poor performance for the new user and new item (cold-start) problems. Stereotypical modelling of users and items is a promising approach to solving these problems. A stereotype represents an aggregation of the characteristics of the items or users which can be used to create general user or item classes. We propose a set of methodologies for the automatic generation of stereotypes to address the cold-start problem. The novelty of the proposed approach rests on the findings that stereotypes built independently of the user-to-item ratings improve both recommendation metrics and computational performance during cold-start phases. The resulting RS can be used with any machine learning algorithm as a solver, and the improved performance gains due to rate-agnostic stereotypes are orthogonal to the gains obtained using more sophisticated solvers. The paper describes how such item-based stereotypes can be evaluated via a series of statistical tests prior to being used for recommendation. The proposed approach improves recommendation quality under a variety of metrics and significantly reduces the dimension of the recommendation model.

Introduction: As in other chronic conditions, medication adherence is important in the treatment of peripheral arterial disease (PAD). Our study aimed at a) analysing non-adherence to angiotensin-converting enzyme inhibitors (ACEIs) and angiotensin receptor blockers (ARBs) in groups of older ACEI and ARB users with PAD, and b) identifying characteristics associated with non-adherence. Methods: We focused on the implementation phase of adherence (i.e., after treatment initiation and before possible discontinuation of treatment). The study cohort included ACEI/ARB users aged ≥65 years in whom PAD was newly diagnosed during 2012. Non-adherence was defined as Proportion of Days Covered (PDC) < 80%. Results: Among 7,080 ACEI/ARB users (6,578 ACEI and 502 ARB users), there was no significant difference in the overall proportion of non-adherent patients between ACEI and ARB users (13.9% and 15.3%, respectively). There were differences in factors associated with non-adherence between the groups of persistent and non-persistent (i.e., discontinued treatment at some point during follow-up) ACEI and ARB users. Increasing age, dementia and bronchial asthma were associated with non-adherence in persistent ACEI users. General practitioner as index prescriber was associated with adherence in the groups of non-persistent ACEI users and persistent ARB users. Conclusion: Identified factors associated with non-adherence may help in determining the groups of patients who require increased attention.

BackgroundLimited studies have investigated the association between statin therapy and poor glycemic control, especially in the Chinese diabetic population.MethodsTwo prospective diabetes cohorts were drawn from the Kailuan Cohort. In Cohort 1, linear regression models were used to evaluate the association between statin therapy and glycated hemoglobin (HbA1c) level change. In Cohort 2, new user design and conditional logistic models were used to assess associations between statin initiation and poor glycemic control which was a composite outcome comprised of hypoglycemic agent escalation and new-onset hyperglycemia.ResultsAmong 11,755 diabetic patients with medication information, 1400 statin users and 1767 statin nonusers with repeated HbA1c measurements were included in Cohort 1 (mean age: 64.6 ± 10.0 years). After a median follow-up of 3.02 (1.44, 5.00) years, statin therapy was associated with higher HbA1c levels (β: 0.20%; 95%CI: 0.05% to 0.34%). In Cohort 2, 1319 pairs of matched cases/controls were included (mean age: 61.6 ± 9.75 years). After a median follow-up of 4.87 (2.51, 8.42) years, poor glycemic control occurred in 43.0% of statin new users and 31.8% of statin nonusers (OR: 1.69; 95% CI: 1.32 to 2.17; P < 0.001). The statin-associated poor glycemic control risk was significantly higher among patients with lower body mass index (Pint = 0.089). Furthermore, a nonlinear association was observed between statin therapy duration and poor glycemic control (P = 0.003).ConclusionsAmong Chinese diabetic adults, statin therapy was associated with a higher level of HbA1c, and a higher risk of hypoglycemic agent escalation and new-onset hyperglycemia, especially among those who had lower body mass index levels and longer statin therapy duration.