PP133 Developing A Novel Multifaceted Graphical Visualization For Treatment Ranking Within An Interactive Network Meta-Analysis Application

IntroductionNetwork meta-analysis (NMA) is a key methodology for comparing the effectiveness of multiple interventions or treatments simultaneously. This project aimed to ascertain current methods and visualizations for treatment ranking within an NMA framework and to subsequently develop a novel graphic within MetaInsight (an interactive NMA web application), to aid clinicians and stakeholders when making decisions regarding the “best” intervention(s) for their patient(s).MethodsCurrent literature on the methodology or visualization of treatment ranking published in the last 10 years was collated and studied. Based on the literature, a novel graphical visualization was developed using RShiny (RStudio, PBC) and integrated within MetaInsight, which is currently hosted on shinyapps.io.ResultsBayesian analyses produce rank probabilities from which mean or median rank and surface under the cumulative ranking curve can be calculated. For frequentist analyses the p-value is available. The simpler methods may be easier to interpret, but they are often more unstable and do not encompass the whole analysis (and vice versa). To aid interpretation and facilitate sensitivity analysis, an interactive graphic was developed that presents rankings alongside treatment effect and study quality results.ConclusionsTreatment ranking is useful, but the results should be interpreted cautiously, and the visualization should be transparent and all-encompassing. A ‘living’ version of MetaInsight, with treatment ranking, would allow interested parties to follow the evidence base as it grows.

Using Normalized Entropy to Measure Uncertainty of Rankings for Network Meta-analyses

Ranking of treatments offers a straightforward interpretation of results derived from network meta-analysis. However, some published network meta-analyses have overemphasized treatment ranking without paying attention to its uncertainty. According to a review of 91 network meta-analyses, 52 reported treatment ranking, but 43 of them did not report the uncertainty of ranking. Without reporting the uncertainty, small differences in the ranking of treatments may be overinterpreted. Rankograms, cumulative rankograms, the credible/confidence interval of mean rank, the surface under the cumulative ranking curve (SUCRA), and the interquartile range of median rank have been used to show the uncertainty of rankings. However, it is not always straightforward to compare the differences in the distribution of probabilities by inspecting rankograms or to compare the intervals or ranges of treatment ranks. We therefore proposed normalized entropy, which transforms the distribution of ranking probabilities into a single quantitative measure, to facilitate a refined interpretation of uncertainty of treatment ranking. We used 4 real examples to demonstrate the uncertainty of ranking quantified by ranking probabilities, 95% confidence interval of SUCRA, and normalized entropy. We showed that as normalized entropy ranges from 0 to 1 and is independent of the number of treatments, it can be used to compare the uncertainty of treatment ranking within a network meta-analysis (NMA) and between different NMAs. Normalized entropy is an alternative tool for measuring the uncertainty of treatment ranking by improving the translation of results from NMAs to clinical practice and avoiding naïve interpretation of treatment ranking. We therefore recommend normalized entropy to be included in the presentation and interpretation of results from NMAs.

Identifying disease-associated signaling pathways through a novel effector gene analysis

Background Signaling pathway analysis methods are commonly used to explain biological behaviors of disease cells. Effector genes typically decide functional attributes (associated with biological behaviors of disease cells) by abnormal signals they received. The signals that the effector genes receive can be quite different in normal vs. disease conditions. However, most of current signaling pathway analysis methods do not take these signal variations into consideration. Methods In this study, we developed a novel signaling pathway analysis method called signaling pathway functional attributes analysis (SPFA) method. This method analyzes the signal variations that effector genes received between two conditions (normal and disease) in different signaling pathways. Results We compared the SPFA method to seven other methods across 33 Gene Expression Omnibus datasets using three measurements: the median rank of target pathways, the median p-value of target pathways, and the percentages of significant pathways. The results confirmed that SPFA was the top-ranking method in terms of median rank of target pathways and the fourth best method in terms of median p-value of target pathways. SPFA’s percentage of significant pathways was modest, indicating a good false positive rate and false negative rate. Overall, SPFA was comparable to the other methods. Our results also suggested that the signal variations calculated by SPFA could help identify abnormal functional attributes and parts of pathways. The SPFA R code and functions can be accessed at https://github.com/ZhenshenBao/SPFA.

Recommending research articles to consumers of online vaccination information

Online health communications often provide biased interpretations of evidence and have unreliable links to the source research. We tested the feasibility of a tool for matching web pages to their source evidence. From 207,538 eligible vaccination-related PubMed articles, we evaluated several approaches using 3,573 unique links to web pages from Altmetric. We evaluated methods for ranking the source articles for vaccine-related research described on web pages, comparing simple baseline feature representation and dimensionality reduction approaches to those augmented with canonical correlation analysis (CCA). Performance measures included the median rank of the correct source article; the percentage of web pages for which the source article was correctly ranked first (recall@1); and the percentage ranked within the top 50 candidate articles (recall@50). While augmenting baseline methods using CCA generally improved results, no CCA-based approach outperformed a baseline method, which ranked the correct source article first for over one quarter of web pages and in the top 50 for more than half. Tools to help people identify evidence-based sources for the content they access on vaccination-related web pages are potentially feasible and may support the prevention of bias and misrepresentation of research in news and social media.

Outcome measures in a combined exercise rehabilitation programme for adults with COPD and chronic heart failure: A preliminary stakeholder consensus event

Combined exercise rehabilitation for chronic obstructive pulmonary disease (COPD) and chronic heart failure (CHF) is potentially attractive. Uncertainty remains as to the baseline profiling assessments and outcome measures that should be collected within a programme. Current evidence surrounding outcome measures in cardiac and pulmonary rehabilitation were presented by experts at a stakeholder consensus event and all stakeholders ( n = 18) were asked to (1) rank in order of importance a list of categories, (2) prioritise outcome measures and (3) prioritise baseline patient evaluation measures that should be assessed in a combined COPD and CHF rehabilitation programme. The tasks were completed anonymously and related to clinical rehabilitation programmes and associated research. Health-related quality of life, exercise capacity and symptom evaluation were voted as the most important categories to assess for clinical purposes (median rank: 1, 2 and 3 accordingly) and research purposes (median rank; 1, 3 and 4.5 accordingly) within combined exercise rehabilitation. All stakeholders agreed that profiling symptoms at baseline were ‘moderately’, ‘very’ or ‘extremely’ important to assess for clinical and research purposes in combined rehabilitation. Profiling of frailty was ranked of the same importance for clinical purposes in combined rehabilitation. Stakeholders identified a suite of multidisciplinary measures that may be important to assess in a combined COPD and CHF exercise rehabilitation programme.

Trace elements in scalp hair of leukaemia patients

The aim of this study was to determine the concentration of Fe, Ni, Cu, Zn and Pb in scalp hair of leukaemia patients and healthy volunteers, using the optimised XRF method. Leukaemia hair samples were classifi ed corresponding to type, growth and age of the participants. The results showed that the studied trace elements (TEs) in both of leukaemia and control groups were positively skewed. In comparison with the control group, lower Fe, Cu, Zn and Pb and higher of Ni medians were found in all studied leukaemia patients. The median rank obtained by Mann-Whitney U-test revealed insignifi cant differences between the leukaemia patients subgroups and the controls. An exact probability (α < 0.05) associated with the U-test showed signifi cant differences between medians in leukaemia patients and controls groups for Pb (lymphatic/control, acute/control), Cu (lymphatic/control, chronic/control), Ni (lymphatic/control, chronic/control) and Fe (chronic/control). Very strong positive and negative correlations (r > 0.70) in the scalp hair of control group were observed between Ni/Fe-Ni, Cu/Fe-Cu, Zn/Fe-Zn, Pb/Fe-Pb, Cu/Ni-Zn/Ni, Cu/Ni-Pb/Ni, Zn/Ni-Pb/Ni, Zn/Fe-Zn/Cu, Pb/Ni-Ni and Ni/Fe-Pb/Ni, whereas only very strong positive ratios in the scalp hair of leukaemia patients group were observed between Ni/Fe-Ni, Cu/Fe-Cu, Zn/Fe-Zn and Pb/Fe-Pb, all correlations were signifi cant at p < 0.05. Other strong and signifi cant correlations were also observed in scalp hair of both groups. Signifi cant differences between grouping of studied TEs in all classifi ed leukaemia groups and controls were found using principal component analysis (PCA). The results of PCA confi rmed that the type and the growth of leukaemia factors were more important in element loading than the age factor.