Aims This study aimed to review the performance of machine learning (ML) methods compared with conventional statistical models (CSMs) for predicting readmission and mortality in patients with heart failure (HF) and to present an approach to formally evaluate the quality of studies using ML algorithms for prediction modelling. Methods and results Following Preferred Reporting Items for Systematic Reviews and Meta‐Analyses guidelines, we performed a systematic literature search using MEDLINE, EPUB, Cochrane CENTRAL, EMBASE, INSPEC, ACM Library, and Web of Science. Eligible studies included primary research articles published between January 2000 and July 2020 comparing ML and CSMs in mortality and readmission prognosis of initially hospitalized HF patients. Data were extracted and analysed by two independent reviewers. A modified CHARMS checklist was developed in consultation with ML and biostatistics experts for quality assessment and was utilized to evaluate studies for risk of bias. Of 4322 articles identified and screened by two independent reviewers, 172 were deemed eligible for a full‐text review. The final set comprised 20 articles and 686 842 patients. ML methods included random forests (n = 11), decision trees (n = 5), regression trees (n = 3), support vector machines (n = 9), neural networks (n = 12), and Bayesian techniques (n = 3). CSMs included logistic regression (n = 16), Cox regression (n = 3), or Poisson regression (n = 3). In 15 studies, readmission was examined at multiple time points ranging from 30 to 180 day readmission, with the majority of studies (n = 12) presenting prediction models for 30 day readmission outcomes. Of a total of 21 time‐point comparisons, ML‐derived c‐indices were higher than CSM‐derived c‐indices in 16 of the 21 comparisons. In seven studies, mortality was examined at 9 time points ranging from in‐hospital mortality to 1 year survival; of these nine, seven reported higher c‐indices using ML. Two of these seven studies reported survival analyses utilizing random survival forests in their ML prediction models. Both reported higher c‐indices when using ML compared with CSMs. A limitation of studies using ML techniques was that the majority were not externally validated, and calibration was rarely assessed. In the only study that was externally validated in a separate dataset, ML was superior to CSMs (c‐indices 0.913 vs. 0.835). Conclusions ML algorithms had better discrimination than CSMs in most studies aiming to predict risk of readmission and mortality in HF patients. Based on our review, there is a need for external validation of ML‐based studies of prediction modelling. We suggest that ML‐based studies should also be evaluated using clinical quality standards for prognosis research. Registration: PROSPERO CRD42020134867

BackgroundMost of the animal models commonly used for preclinical research into Alzheimer’s disease (AD) largely fail to address the pathophysiology, including the impact of known risk factors, of the widely diagnosed sporadic form of the disease. Here, we use a transgenic rat (APP21) that does not develop AD-like pathology spontaneously with age, but does develop pathology following vascular stress. To further the potential of this novel rat model as a much-needed pre-clinical animal model of sporadic AD, we characterize APP21 transgenic rats behaviorally and histologically up to 19 months of age.MethodsThe open field test was used as a measure of activity; and the Morris water maze was used to assess learning, memory, and strategy shift. Neuronal loss and microglia activation were also assessed throughout the brain.ResultsAPP21 transgenic rats showed deficits in working memory from an early age, yet memory recall performance after 24 and 72 h was equal to that of wildtype rats and did not deteriorate with age. A deficit in strategy shift was observed at 19 months of age in APP21 transgenic rats compared to Fischer wildtype rats. Histologically, APP21 transgenic rats demonstrated accelerated white matter inflammation compared to wildtype rats, but interestingly no differences in neuron loss were observed.ConclusionsThe combined presence of white matter pathology and executive function deficits mirrored what is often found in patients with mild cognitive impairment or early dementia, and suggests that this rat model will be useful for translationally meaningful studies into the development and prevention of sporadic AD. The presence of widespread white matter inflammation as the only observed pathological correlate for cognitive deficits raises new questions as to the role of neuroinflammation in cognitive decline.Electronic supplementary materialThe online version of this article (10.1186/s12974-018-1273-7) contains supplementary material, which is available to authorized users.

IntroductionInflammation is emerging as an important risk factor for atherosclerotic cardiovascular disease and has been a recent target for many novel therapeutic agents. However, comparative evidence regarding efficacy of these anti-inflammatory treatment options is currently lacking.Methods and analysisThis systematic review will include randomised controlled trials evaluating the effect of anti-inflammatory agents on cardiovascular outcomes in patients with known cardiovascular disease. Studies will be retrieved from Medline, Embase, the Cochrane Central Register of Controlled Trials, as well as clinical trial registry websites, Europe PMC and conference abstract handsearching. No publication date or language restrictions will be imposed. Eligible interventions must have some component of anti-inflammatory agent. These include (but are not limited to): non-steroidal anti-inflammatory drugs (NSAIDs), colchicine, prednisone, methotrexate, canakinumab, pexelizumab, anakinra, succinobucol, losmapimod, inclacumab, atreleuton, LP-PLA2 (darapladib) and sPLA2 (varespladib). The primary outcomes will include major adverse cardiac events (MACE), and each individual component of MACE (myocardial infarction, stroke and cardiovascular death). Key secondary outcomes will include unstable angina, heart failure, all-cause mortality, cardiac arrest and revascularisation. Screening, inclusion, data extraction and quality assessment will be performed independently by two reviewers. Network meta-analysis based on the random effects model will be conducted to compare treatment effects both directly and indirectly. The quality of the evidence will be assessed with appropriate tools including the Grading of Recommendations, Assessment, Development and Evaluation profiler or Confidence in Network Meta-Analysis tool.Ethics and disseminationEthics approval is not required for this systematic review. The findings will be disseminated through a peer-reviewed journal.PROSPERO registration numberCRD42022303289.

Diabetes mellitus is a major cause of morbidity and mortality, accounting for 1.5 million deaths worldwide in 2012 and attributable to 11.9% of deaths in Canada in 2009. [1][2][3] The prevalence of diabetes has also been steadily increasing, with an estimated 3.4 million Canadians or 9.3% of the population affected in 2015. [4][5][6] By 2025, the prevalence is predicted to rise 44% to 5 million Canadians, or 12.1% of the population. 5 With considerable advances in glycemic control measures and management strategiessuch as early lifestyle modifications and novel pharmacol ogic interventions -which have the potential to reduce morbidity and mortality, screening for diabetes is cost-effective. 7-10 However, despite advocacy for early diagnosis and intervention, 11 diabetes often goes unnoticed and appropriate interventions are delayed as a substantial proportion of individuals who ultimately receive diagnoses may be asymptomatic in the initial phases for many years. 12,13 In Canada, the prevalence of undiagnosed diabetes is estimated to be 1.1%-3.1%. 14 Diabetes Canada and the Canadian Cardiovascular Harmonized National Guidelines Endeavour (C-CHANGE) guidelines thus recommend that all adults aged 40 years and older be screened every 3 years, and those at very high risk regardless of age (e.g., with cardiovascular disease or cardiac risk factors, and some ethnic groups) be screened every 6 to 12 months to ensure early diagnosis and initiation of appropriate interventions to reduce morbidity and mortality. 15,16

Background: Most of the animal models commonly used for preclinical research into Alzheimer's disease (AD) largely fail to address the pathophysiology, including the impact of known risk factors, of the widely diagnosed sporadic form of the disease. Here, we use a transgenic rat (APP21) that does not develop AD-like pathology spontaneously with age, but does develop pathology following vascular stress. To further the potential of this novel rat model as a much-needed pre-clinical animal model of sporadic AD, we characterize APP21 transgenic rats behaviorally and histologically up to 19 months of age. Methods:The open field test was used as a measure of activity; and the Morris water maze was used to assess learning, memory, and strategy shift. Neuronal loss and microglia activation were also assessed throughout the brain.Results: APP21 transgenic rats showed deficits in working memory from an early age, yet memory recall performance after 24 and 72 h was equal to that of wildtype rats and did not deteriorate with age. A deficit in strategy shift was observed at 19 months of age in APP21 transgenic rats compared to Fischer wildtype rats. Histologically, APP21 transgenic rats demonstrated accelerated white matter inflammation compared to wildtype rats, but interestingly no differences in neuron loss were observed. Conclusions:The combined presence of white matter pathology and executive function deficits mirrored what is often found in patients with mild cognitive impairment or early dementia, and suggests that this rat model will be useful for translationally meaningful studies into the development and prevention of sporadic AD. The presence of widespread white matter inflammation as the only observed pathological correlate for cognitive deficits raises new questions as to the role of neuroinflammation in cognitive decline.