Combining probabilistic forecasts of COVID-19 mortality in the United States.

The COVID-19 pandemic has placed forecasting models at the forefront of health policy making. Predictions of mortality, cases and hospitalisations help governments meet planning and resource allocation challenges. In this paper, we consider the weekly forecasting of the cumulative mortality due to COVID-19 at the national and state level in the U.S. Optimal decision-making requires a forecast of a probability distribution, rather than just a single point forecast. Interval forecasts are also important, as they can support decision making and provide situational awareness. We consider the case where probabilistic forecasts have been provided by multiple forecasting teams, and we combine the forecasts to extract the wisdom of the crowd. We use a dataset that has been made publicly available from the COVID-19 Forecast Hub. A notable feature of the dataset is that the availability of forecasts from participating teams varies greatly across the 40 weeks in our study. We evaluate the accuracy of combining methods that have been previously proposed for interval forecasts and predictions of probability distributions. These include the use of the simple average, the median, and trimming methods. In addition, we propose several new weighted combining methods. Our results show that, although the median was very useful for the early weeks of the pandemic, the simple average was preferable thereafter, and that, as a history of forecast accuracy accumulates, the best results can be produced by a weighted combining method that uses weights that are inversely proportional to the historical accuracy of the individual forecasting teams.

Changes in Tracheal Tube Cuff Pressure and Recurrent Laryngeal Nerve Conductivity During Thyroid Surgery.

BACKGROUND: The aetiology of recurrent laryngeal nerve (RLN) neurapraxia is unclear in most RLN palsies post-thyroidectomy. We hypothesised that high intralaryngeal pressures impede RLN conductivity, in turn contributing to RLN palsy. Therefore, we measured tracheal tube (TT) cuff pressure (as a surrogate for intralaryngeal pressure) and RLN conduction during ten standard manoeuvres in thyroidectomy, to assess for correlation between cuff pressure and RLN conductivity. METHODS: A prospective cohort study of thyroidectomy during 2018. For each thyroid lobe, TT cuff pressure was continuously measured via an air pressure transducer. RLN conduction (amplitude and latency) was measured using continuous neuromonitoring. Changes in mean TT cuff pressure and median nerve conduction from baseline measurements were analysed using Student's t test and Wilcoxon signed-rank test. RESULTS: In a total of 50 RLNs, the mean baseline TT cuff pressure of 19.5 ± 8.9 mmHg increased significantly to 22.0 mmHg during anteromedial rotation of the thyroid (p < 0.05). RLN conduction changed during manipulation of the superior thyroid pole with shortening of latency (-0.49% from baseline, p = 0.05) and reduction in amplitude (-12.0% from baseline, p = 0.02). The timing of these deviations did not correlate with the increased TT cuff pressure. In three cases of temporary RLN palsy, the mean cumulative case TT cuff pressure was significantly higher (24.8 mmHg, p = 0.02). CONCLUSIONS: This study demonstrates that TT cuff pressure and RLN conductivity can change significantly with manipulation of the thyroid and that high TT cuff pressures may be associated with RLN injuries.

Long-Term Survival of Total Hip Arthroplasty Using Implants From Different Manufacturers.

BACKGROUND: Total hip arthroplasty (THA) can be performed using either femoral and acetabular components provided by the same manufacturer (matched components) or components from different manufacturers (unmatched components). We hypothesized that there would be no difference in outcomes following the use of unmatched compared to matched components. METHODS: Data from a nationwide joint registry, the New Zealand Joint Registry (NZJR), were analyzed to assess long-term outcomes of using unmatched implants in THA. RESULTS: The NZJR has recorded a total of 108,613 primary THAs. We excluded combinations with less than 50 implantations, leaving 99,732 arthroplasties (90.5%). The unmatched group consisted of 24,537 (24.6%) THAs. Revision procedures were required in 3434 (4.6%) of the matched group, at a rate of 0.72/100 component years and 1078 (4.4%) of the unmatched group, a rate of 0.69/100 component years (P = .049). THAs with metal-on-metal or ceramic-on-metal bearings were overrepresented in the matched group. When analysis was repeated with these implants excluded, there was no longer a difference in revision rate between groups (4.0% revisions, 0.65/100 component years and 4.3% revisions, 0.67/100 component years [P = .742]). Survival analysis showed 17-year survival for matched components and unmatched to be within 95% confidence intervals at all time points. There was a small, statistically significant improvement in Oxford Hip Scores for the unmatched group compared with the matched group. CONCLUSION: Data from the NZJR confirm that the use of unmatched components in THA has no adverse effect on outcomes.

Interval forecasts of weekly incident and cumulative COVID-19 mortality in the United States: A comparison of combining methods.

BACKGROUND: A combined forecast from multiple models is typically more accurate than an individual forecast, but there are few examples of studies of combining in infectious disease forecasting. We investigated the accuracy of different ways of combining interval forecasts of weekly incident and cumulative coronavirus disease-2019 (COVID-19) mortality. METHODS: We considered weekly interval forecasts, for 1- to 4-week prediction horizons, with out-of-sample periods of approximately 18 months ending on 8 January 2022, for multiple locations in the United States, using data from the COVID-19 Forecast Hub. Our comparison involved simple and more complex combining methods, including methods that involve trimming outliers or performance-based weights. Prediction accuracy was evaluated using interval scores, weighted interval scores, skill scores, ranks, and reliability diagrams. RESULTS: The weighted inverse score and median combining methods performed best for forecasts of incident deaths. Overall, the leading inverse score method was 12% better than the mean benchmark method in forecasting the 95% interval and, considering all interval forecasts, the median was 7% better than the mean. Overall, the median was the most accurate method for forecasts of cumulative deaths. Compared to the mean, the median's accuracy was 65% better in forecasting the 95% interval, and 43% better considering all interval forecasts. For all combining methods except the median, combining forecasts from only compartmental models produced better forecasts than combining forecasts from all models. CONCLUSIONS: Combining forecasts can improve the contribution of probabilistic forecasting to health policy decision making during epidemics. The relative performance of combining methods depends on the extent of outliers and the type of models in the combination. The median combination has the advantage of being robust to outlying forecasts. Our results support the Hub's use of the median and we recommend further investigation into the use of weighted methods.

Smell and taste loss in COVID-19 patients: assessment outcomes in a Victorian population.

Background: It has been noted that olfactory and gustatory disturbances may precede or accompany the typical features of COVID-19, such as fever and cough. Hence, a high index of suspicion is required when patients report sudden loss of smell or taste, in order to facilitate timely diagnosis and isolation.Aims/objectives: The aim of this study was to assess the frequency of olfactory and gustatory disturbances in COVID-19 positive patients from a cohort representative of Melbourne, Australia.Methods: A retrospective descriptive study was conducted on patients who tested positive for COVID-19. Standardised phone consultations and online follow-up questionnaires were performed to assess clinical features of COVID-19, with a focus on smell and taste disorders.Results: The most frequent symptoms experienced were taste and smell disturbances with 74% experiencing either smell or taste disturbance or both. Post-recovery, 34% of patients continued to experience ongoing hyposmia and 2% anosmia, whereas 28% continued to suffer from hypogeusia or ageusia.Conclusion and significance: This study presents the high rates of improvement of both olfactory and gustatory disturbance in a short-lived period. It also highlights the importance of these symptoms in prompting appropriate testing, quarantine precautions, initiate early olfactory retraining and the potential for continued sensory disturbance.

Structural insights into the mechanism and inhibition of the ß-hydroxydecanoyl-acyl carrier protein dehydratase from Pseudomonas aeruginosa.

Fatty acid biosynthesis is an essential component of metabolism in both eukaryotes and prokaryotes. The fatty acid biosynthetic pathway of Gram-negative bacteria is an established therapeutic target. Two homologous enzymes FabA and FabZ catalyze a key step in fatty acid biosynthesis; both dehydrate hydroxyacyl fatty acids that are coupled via a phosphopantetheine to an acyl carrier protein (ACP). The resulting trans-2-enoyl-ACP is further polymerized in a processive manner. FabA, however, carries out a second reaction involving isomerization of trans-2-enoyl fatty acid to cis-3-enoyl fatty acid. We have solved the structure of Pseudomonas aeruginosa FabA with a substrate allowing detailed molecular insight into the interactions of the active site. This has allowed a detailed examination of the factors governing the second catalytic step. We have also determined the structure of FabA in complex with small molecules (so-called fragments). These small molecules occupy distinct regions of the active site and form the basis for a rational inhibitor design program.

Metarhizium anisopliae pathogenesis of mosquito larvae: a verdict of accidental death.

Metarhizium anisopliae, a fungal pathogen of terrestrial arthropods, kills the aquatic larvae of Aedes aegypti, the vector of dengue and yellow fever. The fungus kills without adhering to the host cuticle. Ingested conidia also fail to germinate and are expelled in fecal pellets. This study investigates the mechanism by which this fungus adapted to terrestrial hosts kills aquatic mosquito larvae. Genes associated with the M. anisopliae early pathogenic response (proteinases Pr1 and Pr2, and adhesins, Mad1 and Mad2) are upregulated in the presence of larvae, but the established infection process observed in terrestrial hosts does not progress and insecticidal destruxins were not detected. Protease inhibitors reduce larval mortality indicating the importance of proteases in the host interaction. The Ae. aegypti immune response to M. anisopliae appears limited, whilst the oxidative stress response gene encoding for thiol peroxidase is upregulated. Cecropin and Hsp70 genes are downregulated as larval death occurs, and insect mortality appears to be linked to autolysis through caspase activity regulated by Hsp70 and inhibited, in infected larvae, by protease inhibitors. Evidence is presented that a traditional host-pathogen response does not occur as the species have not evolved to interact. M. anisopliae retains pre-formed pathogenic determinants which mediate host mortality, but unlike true aquatic fungal pathogens, does not recognise and colonise the larval host.

An evaluation of Bayesian techniques for controlling model complexity and selecting inputs in a neural network for short-term load forecasting.

Artificial neural networks have frequently been proposed for electricity load forecasting because of their capabilities for the nonlinear modelling of large multivariate data sets. Modelling with neural networks is not an easy task though; two of the main challenges are defining the appropriate level of model complexity, and choosing the input variables. This paper evaluates techniques for automatic neural network modelling within a Bayesian framework, as applied to six samples containing daily load and weather data for four different countries. We analyse input selection as carried out by the Bayesian 'automatic relevance determination', and the usefulness of the Bayesian 'evidence' for the selection of the best structure (in terms of number of neurones), as compared to methods based on cross-validation.

Bismuth(III) complexes derived from non-steroidal anti-inflammatory drugs and their activity against Helicobacter pylori.

The formation of bismuth(III) complexes of carboxylates and benzoates derived from the 1 : 3 reaction of BiPh(3) with the common non-steroidal anti-inflammatory drugs (NSAIDs) ketoprofen, naproxen, ibuprofen, mefenamic acid, diflunisal, 5-chlorosalicylic acid, fenbufen, sulindac, tolfenamic acid and flufenamic acid, has been achieved using both solvent-free and solvent-mediated methods. The thermochemical profiles of the solvent-free reactions were studied using DSC-TGA. All reactions produced the tris-substituted complexes of general formula [BiL(3)](n), with the complexes derived from ketoprofen and sulindac having an additional single bismuth bound H(2)O molecule in the inner coordination sphere. The complexes are stable in air over a period of six months, do not undergo significant decomposition when suspended overnight in water, but decompose in 1 M HCl solution to release the free acid form of the NSAID. All ten complexes show excellent in vitro activity against Helicobacter pylori with MIC values of > or = 6.25 microg mL(-1).

Global dynamic optimization for parameter estimation in chemical kinetics.

We present the first method guaranteed to find the best possible least-squares (chi2) fit of experimental data by a nonlinear kinetic model. Several important advantages of knowing with certainty the best possible fit rather than a locally optimum fit are discussed and demonstrated using data from the recent literature. This is particularly important when the model and the data appear to be inconsistent. With the new method, one can rigorously demonstrate that a nonlinear kinetic model with several adjustable rate parameters is inconsistent with measured experimental data. The numerical method presented is a valuable tool in evaluating the validity of a complex kinetics model.

The human fatty acid synthase gene and de novo lipogenesis are coordinately regulated in human adipose tissue.

Despite its potential importance in obesity and related disorders, little is known about regulation of lipogenesis in human adipose tissue. To investigate this area at the molecular and mechanistic levels, we studied lipogenesis and the regulation of 1 of its core enzymes, fatty acid synthase (FAS), in human adipose tissue in response to hormonal and nutritional manipulation. As a paradigm for lipogenic genes, we cloned the upstream region of the human FAS gene, compared its sequence to that of FAS orthologs from other species, and identified important regulatory elements that lie upstream of the FAS coding region. Lipogenesis, as assessed by glucose incorporation into lipids, was increased by insulin and more so by the combination of insulin and dexamethasone (Dex, a potent glucocorticoid analogue). In parallel, FAS expression, activity, and gene transcription rate were also significantly increased by these treatments. We also showed that linoleic acid, a representative PUFA, attenuated the actions of insulin and Dex on fatty acid and lipid synthesis as well as FAS activity and expression. Using reporter assays, we determined that the regions responsible for hormonal regulation of the FAS gene lie in the proximal portion of the gene's 5'-flanking region, within which we identified an insulin response element similar to the E-box sequence we identified previously in the rat FAS gene. In summary, we demonstrated that lipogenesis occurs in human adipose tissue and can be induced by insulin, further enhanced by glucocorticoids, and suppressed by PUFA in a hormone-dependent manner.