The Edmonds Blossom algorithm is implemented here using depth-first search, which is intrinsically serial. By streamlining the code, our serial implementation is consistently three to five times faster than the previously fastest general graph matching code. By extracting parallelism across iterations of the algorithm, with coarse-grain locking, we are able to further reduce the run time on random regular graphs four-fold and obtain a two-fold reduction of run time on real-world graphs with similar topology. Solving very sparse graphs (average degree less than four) exhibiting community structure with eight threads led to a slow down of three-fold, but this slow down is replaced by marginal speed up once the average degree is greater than four. We conclude that our parallel coarse-grain locking implementation performs well when extracting parallelism from this augmenting-path-based algorithm and may work well for similar algorithms.

The matching problem formulated as Maximum Cardinality Matching in General Graphs (MCMGG) finds the largest matching on graphs without restrictions. The Micali-Vazirani algorithm has the best asymptotic complexity for solving MCMGG when the graphs are sparse. Parallelizing matching in general graphs on the GPU is difficult for multiple reasons. First, the augmenting path procedure is highly recursive, and NVIDIA GPUs use registers to store kernel arguments, which eventually spill into cached device memory, with a performance penalty. Second, extracting parallelism from the matching process requires partitioning the graph to avoid any overlapping augmenting paths. We propose an implementation of the Micali-Vazirani algorithm which identifies bridge edges using thread-parallel breadth-first search, followed by block-parallel path augmentation and blossom contraction. Augmenting path and Union-find methods were implemented as stack-based iterative methods, with a stack allocated in shared memory. Our experimentation shows that compared to the serial implementation, our approach results in up to 15-fold speed-up for very sparse regular graphs, up to 5-fold slowdown for denser regular graphs, and finally a 50-fold slowdown for power-law distributed Kronecker graphs. This implementation has been open-sourced for further research on developing combinatorial graph algorithms on GPUs.

Observational studies are frequently used in clinical research to estimate the effects of treatments or exposures on outcomes. To reduce the effects of confounding when estimating treatment effects, covariate balancing methods are frequently implemented. This study evaluated, using extensive Monte Carlo simulation, several methods of covariate balancing, and two methods for propensity score estimation, for estimating the average treatment effect on the treated using a hazard ratio from a Cox proportional hazards model. With respect to minimizing bias and maximizing accuracy (as measured by the mean square error) of the treatment effect, the average treatment effect on the treated weighting, fine stratification, and optimal full matching with a conventional logistic regression model for the propensity score performed best across all simulated conditions. Other methods performed well in specific circumstances, such as pair matching when sample sizes were large (n = 5000) and the proportion treated was < 0.25. Statistical power was generally higher for weighting methods than matching methods, and Type I error rates were at or below the nominal level for balancing methods with unbiased treatment effect estimates. There was also a decreasing effective sample size with an increasing number of strata, therefore for stratification-based weighting methods, it may be important to consider fewer strata. Generally, we recommend methods that performed well in our simulations, although the identification of methods that performed well is necessarily limited by the specific features of our simulation. The methods are illustrated using a real-world example comparing beta blockers and angiotensin-converting enzyme inhibitors among hypertensive patients at risk for incident stroke.

UNASSIGNED: Up to 30% of patients with severe aortic stenosis (SAS) (indexed aortic valve area [AVAi] <0.6 cm2/m2) exhibit low-transvalvular gradient despite normal ejection fraction. There is intense debate regarding the prognostic significance of this entity.
UNASSIGNED: The purpose of this study was to compare the outcome of patients with discordant low-gradient SAS (DLG-SAS) vs moderate aortic stenosis (MAS) and high-gradient SAS (HG-SAS).
UNASSIGNED: We used the BEL-F-ASt (Belgium-France-Aortic Stenosis) registry including consecutive patients with AS. Survival was compared overall and after matching (inverse probability weighting and propensity-score matching) for clinical and imaging variables. The analysis was first performed in the overall population (n = 2,582) and then in the population of unoperated patients (n = 1,812).
UNASSIGNED: After-inverse probability weighting-matching, the 3 groups were balanced. Five-year survival was better in MAS than in DLG-SAS and HG-SAS-patients (58.9% vs 47% vs 41.2%, P < 0.001). Similar results were obtained in unoperated patients (54.1% vs 37.9% vs 28.1%, P < 0.001). To explore the impact of MG (≤40 vs >40 mmHg) and AVAi (<0.6 vs ≥0.6 cm2/m2) on outcomes, survival of propensity score-matched cohorts of HG-vs DLG-SAS and MAS vs DLG-SAS were compared. After matching for MG, survival was better in MAS than in DLG-SAS (52% vs 40%, P < 0.001). After matching for AVAi, survival was better in DLG-SAS than in HG-SAS patients (45% vs 33%, P < 0.001).
UNASSIGNED: Survival of DLG-SAS is better than that of HG-SAS and worse than that of MAS patients. At comparable MG, the lower the AVAi, the worse the prognosis, whereas at comparable AVAi, the higher the MG, the worse the prognosis. These data argue that DLG-SAS is an intermediate form in the disease continuum.

Protected areas conserve biodiversity and ecosystem functions but might impede local economic growth. Understanding the global patterns and predictors of different relationships between protected area effectiveness and neighboring community economic growth can inform better implementation of the Kunming-Montreal Global Biodiversity Framework. We assessed 10,143 protected areas globally with matched samples to address the non-random location of protected areas. Our results show that protected areas resist human-induced land cover changes and do not limit nightlight increases in neighboring settlements. This result is robust, using different matching techniques, parameter settings, and selection of covariates. We identify four types of relationships between land cover changes and nightlight changes for each protected area: \"synergy,\" \"retreat,\" and two tradeoff relationships. About half of the protected areas (47.5%) retain their natural land cover and do so despite an increase of nightlights in the neighboring communities. This synergy relationship is the most common globally but varies between biomes and continents. Synergy is less frequent in the Amazon, Southeast Asia, and some developing areas, where most biodiversity resides and which suffer more from poverty. Smaller protected areas and those with better access to cities, moderate road density, and better baseline economic conditions have a higher probability of reaching synergy. Our results are promising, as the expansion of protected areas and increased species protection will rely more on conserving the human-modified landscape with smaller protected areas. Future interventions should address local development and biodiversity conservation together to achieve more co-benefits.

Recently, there has been considerable progress in developing new technologies and equipment for the medical field, including minimally invasive surgeries. Evaluating the effectiveness of these treatments requires study designs like randomized controlled trials. However, due to the nature of certain treatments, randomization is not always feasible, leading to the use of observational studies. The effect size estimated from observational studies is subject to selection bias caused by confounders. One method to reduce this bias is propensity scoring. This study aimed to introduce a propensity score matching process between two groups using a practical example with R. Additionally, Rex, an Excel add-in graphical user interface statistical program, is provided for researchers unfamiliar with R programming. Further techniques, such as matching with three or more groups, propensity score weighting and stratification, and imputation of missing values, are summarized to offer approaches for more complex studies not covered in this tutorial.

This study assesses the impact of a voucher project that targeted vulnerable and poor pregnant women in Uganda. Highly subsidised vouchers gave access to a package of safe delivery services consisting of four antenatal visits, safe delivery, one postnatal visit, the treatment and management of selected pregnancy-related medical conditions and complications, and emergency transport. Vouchers were sold during the project\'s operational period from 2016 to 2019. This study covers 8 out of 25 project-benefiting districts in Uganda and a total of 1,881 pregnancies, including both beneficiary and non-beneficiary mothers. Using a matching design, the results show a positive effect on the survival of new-born babies. The difference in the survival rate between the control group and the treatment group is 5.4% points, indicating that the voucher project reduced infant mortality by more than 65 per cent.

BACKGROUND: Patients with initial stroke or transient ischemic attack (TIA) are at high risk for further strokes, death or cardiovascular events. Even the first-ever stroke is associated with a high chance of disability and need for assistance. The risk of long-term health care demands increases with each subsequent event. Although the inpatient sector already provides a high standard of care in Germany, it can be difficult to obtain cross-sectoral aftercare. Thus, the study investigated whether a structured case management program can avoid stroke recurrences.
METHODS: The study was conducted with a quasi-experimental study design in three regions in North Rhine-Westphalia. Patients with first-ever stroke or TIA were eligible to participate. The intervention group was prospectively recruited and supported by a case manager during a one-year follow-up. Optimal Full Matching was used to generate a control group based on statutory claims data. The primary outcome was the stroke recurrence. Recurrence and mortality were analysed by using Cox regression; other secondary outcomes were examined with test-based procedures and with logistic regressions. Additionally, subgroup analyses were performed.
RESULTS: From June 2018 to March 2020, 1,512 patients were enrolled in the intervention group. Claims data from 19,104 patients have been transmitted for establishing the control group. After the matching process, 1,167 patients of each group were included in the analysis. 70 recurrences (6.0%) occurred in the intervention group and 67 recurrences (5.7%) in the control group. With a hazard ratio of 1.06 (95% CI: [1.42-0.69]; p=0.69), no significant effect was found for the primary outcome. With regard to the secondary outcome mortality, 36 patients in the intervention group and 46 in the control group died (3.1% vs. 3.9%). Again, there was no significant effect (HR: 0.86; 95% CI: [0.58-1.28], p=0.46).
CONCLUSIONS: Based on the present findings, the case management approach for stroke patients evaluated here was unable to demonstrate an improvement in health care. Potential effects of case management might not be adequately depicted in short observation periods. Thus, future studies should consider longer observation periods.
CONCLUSIONS: A panel of experts should discuss whether the core approach of case management to support cost-intensive individual cases is contrary to a broad implementation with a one-size-fits-all intervention for stroke patients. In this case, further research should focus on more specific study populations.

A complete science of human behavior requires a comprehensive account of the verbal behavior those humans exhibit. Existing behavioral theories of such verbal behavior have produced compelling insight into language\'s underlying function, but the expansive program of research those theories deserve has unfortunately been slow to develop. We argue that the status quo\'s manually implemented and study-specific coding systems are too resource intensive to be worthwhile for most behavior analysts. These high input costs in turn discourage research on verbal behavior overall. We propose lexicon-based sentiment analysis as a more modern and efficient approach to the study of human verbal products, especially naturally occurring ones (e.g., psychotherapy transcripts, social media posts). In the present discussion, we introduce the reader to principles of sentiment analysis, highlighting its usefulness as a behavior analytic tool for the study of verbal behavior. We conclude with an outline of approaches for handling some of the more complex forms of speech, like negation, sarcasm, and speculation. The appendix also provides a worked example of how sentiment analysis could be applied to existing questions in behavior analysis, complete with code that readers can incorporate into their own work.

In large-scale observational data with a hierarchical structure, both clusters and interventions often have more than two levels. Popular methods in the binary treatment literature do not naturally extend to the hierarchical multilevel treatment case. For example, most K-12 and universities have moved to an unprecedented hybrid learning module during the COVID-19 pandemic where learning modes include hybrid and fully remote learning, while students were clustered within a class and school region. It is challenging to evaluate the effectiveness of the learning outcomes of the multilevel treatments in a hierarchically data structured. In this paper, we study a covariates matching method and develop a generalized propensity score matching method to reduce the bias of estimation in the intervention effect. We also propose simple algorithms to assess the covariates balance for each approach. We examine the finite sample performance of the methods via simulation studies and apply the proposed methods to analyze the effectiveness of learning modes during the COVID-19 pandemic.