This modelling study aimed to estimate the burden for allergic diseases in children during a period of 30 years.

Population-based observational study.

The data on the incidence, mortality and disability-adjusted life years (DALYs) for childhood allergic diseases, such as atopic dermatitis (AD) and asthma, were retrieved from the Global Burden of Disease study 2019 online database. This data set spans various groups, including different regions, ages, genders and Socio-Demographic Indices (SDI), covering the period from 1990 to 2019.

In 2019, there were approximately 81 million children with asthma and 5.6 million children with AD worldwide. The global incidence of asthma in children was 20 million. Age-standardised incidence rates showed a decrease of 4.17% for asthma, from 1075.14 (95% uncertainty intervals (UI), 724.63 to 1504.93) per 100 000 population in 1990 to 1030.33 (95% UI, 683.66 to 1449.53) in 2019. Similarly, the rates for AD decreased by 5.46%, from 594.05 (95% UI, 547.98 to 642.88) per 100 000 population in 1990 to 561.61 (95% UI, 519.03 to 608.29) in 2019. The incidence of both asthma and AD was highest in children under 5 years of age, gradually decreasing with age. Interestingly, an increase in SDI was associated with a rise in the incidence of both conditions. However, the mortality rate and DALYs for asthma showed a contrasting trend.

Over the past three decades, there has been a worldwide increase in new asthma and AD cases, even though mortality rates have significantly declined. However, the prevalence of these allergic diseases among children varies considerably across regions, countries and age groups. This variation highlights the need for precise prevalence assessments. These assessments are vital in formulating effective strategies for prevention and treatment.

Considering the increasing incidence of Alzheimer’s disease (AD) and mild cognitive impairment (MCI) worldwide, there is an urgent need to identify efficacious, safe and convenient treatments. Numerous investigations have been conducted on the use of supplements in this domain, with oral supplementation emerging as a viable therapeutic approach for AD or MCI. Nevertheless, given the multitude of available supplements, it becomes imperative to identify the optimal treatment regimen.

Eight academic databases and three clinical trial registries will be searched from their inception to 1 June 2023. To identify randomised controlled trials investigating the effects of supplements on patients with AD or MCI, two independent reviewers (X-YZ and Y-QL) will extract relevant information from eligible articles, while the risk of bias in the included studies will be assessed using the Rob 2.0 tool developed by the Cochrane Collaboration. The primary outcome of interest is the overall cognitive function. Pair-wise meta-analysis will be conducted using RevMan V.5.3, while network meta-analysis will be carried out using Stata 17.0 and ADDIS 1.16.8. Heterogeneity test, data synthesis and subgroup analysis will be performed if necessary. The GRADE system will be employed to assess the quality of evidence. This study is scheduled to commence on 1 June 2023 and conclude on 1 October 2023.

Ethics approval is not required for systematic review and network meta-analysis. The results will be submitted to a peer-reviewed journal or at a conference.

PROSPERO (CRD42023414700).

To develop an interpretable deep learning model of lupus nephritis (LN) relapse prediction based on dynamic multivariable time-series data.

A single-centre, retrospective cohort study in China.

A Chinese central tertiary hospital.

The cohort study consisted of 1694 LN patients who had been registered in the Nanjing Glomerulonephritis Registry at the National Clinical Research Center of Kidney Diseases, Jinling Hospital from January 1985 to December 2010.

We developed a deep learning algorithm to predict LN relapse that consists of 59 features, including demographic, clinical, immunological, pathological and therapeutic characteristics that were collected for baseline analysis. A total of 32 227 data points were collected by the sliding window method and randomly divided into training (80%), validation (10%) and testing sets (10%). We developed a deep learning algorithm-based interpretable multivariable long short-term memory model for LN relapse risk prediction considering censored time-series data based on a cohort of 1694 LN patients. A mixture attention mechanism was deployed to capture variable interactions at different time points for estimating the temporal importance of the variables. Model performance was assessed according to C-index (concordance index).

The median follow-up time since remission was 4.1 (IQR, 1.7–6.7) years. The interpretable deep learning model based on dynamic multivariable time-series data achieved the best performance, with a C-index of 0.897, among models using only variables at the point of remission or time-variant variables. The importance of urinary protein, serum albumin and serum C3 showed time dependency in the model, that is, their contributions to the risk prediction increased over time.

Deep learning algorithms can effectively learn through time-series data to develop a predictive model for LN relapse. The model provides accurate predictions of LN relapse for different renal disease stages, which could be used in clinical practice to guide physicians on the management of LN patients.

This study aims to investigate physicians’ familiarity and awareness of four diabetes guidelines and their practice of the recommendations outlined in these guidelines.

A cross-sectional study.

An online questionnaire survey was conducted among physicians affiliated with the Specialist Committee for Primary Diabetes Care of China Association of Chinese Medicine, using the snowball sampling method to ensure a broader representation of physicians.

1150 physicians from 192 cities across 30 provinces in China provided complete data.

Tertiary care hospital physicians (TCPs) exhibited the highest familiarity with the Guideline for the Prevention and Treatment of Type 2 Diabetes Mellitus in China (91.3%), followed by the National Guidelines for the Prevention and Control of Diabetes in Primary Care (76.8%), the Standards of Medical Care in Diabetes (72.2%) and the Guidelines for Prevention and Treatment of Diabetes in Chinese Medicine (63.8%). Primary care practitioners (PCPs) exhibited familiarity with these four guidelines at about 50% or less. Self-reported reference to modern diabetes guidelines by physicians is more frequent than traditional Chinese medicine (TCM) diabetes guidelines, with rates at 73.2% and 33.8%, respectively. Approximately 90% of physicians provided instructions on self-monitoring of blood glucose to their patients with diabetes. Less than one-third of physicians referred patients to a specialised nutritionist. In terms of health education management, TCPs reported having a diabetes health management team at the rate of 75.7%, followed by secondary care hospital physicians at 57.0% and PCPs at 27.5%. Furthermore, approximately 40% of physicians did not fully grasp hypoglycaemia characteristics.

Familiarity and awareness of the screening guidelines varied among physicians in different hospital settings. Importantly, significant discrepancies were observed between physicians’ awareness and their self-reported reference to modern medicine guidelines and TCM guidelines. It is essential to consistently provide education and training on diabetes management for all physicians, particularly PCPs.