[Contemporary Clinical Trials Communications, 10 November 2023]

This short communication concerns a biomarker adaptive Phase 2/3 design for new oncology drugs with an uncertain biomarker effect. Depending on the outcome of an interim analysis for adaptive decision, a Phase 2 study that starts in a biomarker enriched subpopulation may continue to the end without expansion to Phase 3, expand to Phase 3 in the same population or expand to Phase 3 in a broader population. Each path can enjoy full alpha for hypothesis testing without inflating the overall Type I error.

[Journal of Biomedical Informatics, 28 September 2023]

Background: We often must conduct diagnostic tests on a massive volume of samples within a limited time during outbreaks of infectious diseases (e.g., COVID-19,screening) or repeat many times routinely (e.g., regular and massive screening for plant virus infections in farms). These tests aim to obtain the diagnostic result of all samples within a limited time. In such scenarios, the limitation of testing resources and human labor drives the need to pool individual samples and test them together to improve testing efficiency. When a pool is positive, further testing is required to identify the affected individuals; whereas when a pool is negative, we conclude all individuals in the pool are negative. How one splits the samples into pools is a critical factor affecting testing efficiency.

Objective: We aim to find the optimal strategy that adaptively guides users on optimally splitting the sample cohort into test-pools.

Methods: We developed an algorithm that minimizes the expected number of tests needed to obtain the diagnostic results of all samples. Our algorithm dynamically updates the critical information according to the result of the most recent test and calculates the optimal pool size for the next test. We implemented our novel adaptive sample pooling strategy into a web-based application, ADSP (https://ADSP.uvic.ca). ADSP interactively guides users on how many samples to be pooled for the current test, asks users to report the test result back and uses it to update the best strategy on how many samples to be pooled for the next test.

Results: We compared ADSP with other popular pooling methods in simulation studies, and found that ADSP requires fewer tests to diagnose a cohort and is more robust to the inaccurate initial estimate of the test cohort’s disease prevalence.

Conclusion: Our web-based application can help researchers decide how to pool their samples for grouped diagnostic tests. It improves test efficiency when grouped tests are conducted.

[JoGH, 26 May 2023]

The coronavirus 2019 (COVID-19) pandemic has significantly impacted the global economy and society. One of the key challenges in combating it was predicting its spread to take appropriate measures, such as lockdowns and social distancing. These measures have now been lifted, and many countries are entering the final stages of the COVID-19 pandemic.

It is essential to continue studying the data collected during the COVID-19 pandemic, even as the focus shifts to recovery and rebuilding, to improve our ability to respond to future pandemics and protect public health. The COVID-19 pandemic has provided a wealth of data that can be used to enhance our understanding of the virus and how it spreads. We used data from 3112 counties in the USA obtained from multiple sources, including the daily infection rates from the COVID-19 Data Repository of the Center for Systems Science and Engineering (CSSE) at the John Hopkins University [1], interventions used to control the spread of the virus [2], and demographics from the US Census [3], to train monitoring systems that detect and track future outbreaks or pandemics, allowing us to better prepare or even mitigate them in advance.

Artificial intelligence (AI) models have been used to forecast the cumulative daily number of COVID-19 cases. These models can analyse large amounts of data and make predictions quickly, which is critical in fast-moving pandemics. We built a forecasting model based on the temporal convolutional network (TCN) [4] and implemented a web application [5] that displays 28-day forecasts for every county in the United States. In our evaluation study, we found that our TCN-based model outperformed its extension (an ensemble model) and other state-of-art forecasting models…