Research News

Research & Results

Statins may be effective treatment for patients with ulcerative colitis

September 15, 2021. "People with ulcerative colitis who are also taking statins have about a 50% decreased risk of colectomies and hospitalization, according to a Stanford Medicine study." DBDS faculty researchers Purvesh Khatri is featured and Nigam Shah is mentioned. 

Data-powered consult service shown to help doctors diagnose illness, guide treatments

September 15, 2021. "Stanford Medicine researchers created a new type of medical consult that harnesses millions of electronic health records to bring new insights to patient care." DBDS faculty researchers Nigam Shah is featured and Trevor Hastie is mentioned. 

RNA splicing programs define tissue compartments and cell types at single cell resolution

September 13, 2021. Members of the Salzman Lab, including co-first authors Julia Olivieri (pictured here) and Roozbeh Dehghannasiri, Peter Wang, Julia Salzman, and colleagues explore the extent splicing is regulated at single-cell resolution in this new eLife publication. 

Why radiologists should consider earlier follow-up imaging for many Lung-RADS cases

August 19, 2021. New evidence from the Stanford Plevritis Lab suggests providers may want to consider ordering follow-up CTs for probably benign nodules earlier than currently suggested. Doing so reduced mortality rates, among other health benefits.

New npj Digital Medicine publication by Akshay Chaudhari, demonstrating prospective use of deep learning to improve the quality of 4x low-dose PET imaging studies

August 18, 2021. Through an external multi-institutional, multi-vendor, and multi-reader evaluation, DBDS Courtesy faculty member Akshay Chaudhari et al. showed that deep learning (DL) can help maintain image quality of fourfold low-dose PET scans without compromising qualitative and quantitative outcomes for patients. This study showed the generalizability of the DL system across different PET scanners, acquisition protocols, and patient habitus.

New Genetic Tech Can Fight Inherited Heart Disease – And Families Can, Too

August 18, 2021. DBDS researcher Dr. Euan Ashley is interviewed in this American Heart Association (ASA) news release, saying "In this new world, we're able to actually sequence genes & give people definitive answers. And that's powerful – first of all, just because having an answer is a powerful thing. But even much more important, it's actionable." 

New preprint on the analysis of longitudinal randomized trial data by Adjunct Prof. Alejandro Schuler

August 17, 2021. Adjunct Professor Alejandro Schuler (pictured here) released his preprint “Mixed models for repeated measures should include time-by-covariate interactions to assure power gains and robustness against dropout bias relative to complete-case ANCOVA" on arXiv. The work shows that a popular method of analyzing data from longitudinal randomized trials can actually result in worse confidence and more bias than a naive method unless the effects of baseline covariates are allowed to vary in time.

New Plevritis Lab research on the ENGAGE framework for risk-based lung cancer screening published in Cancer 

August 12, 2021. A paper from DBDS Plevritis Lab researchers, Iakovos (Jacob) Toumazis (pictured here) and Sylvia Plevritis, and colleagues describing the findings of applying the ENGAGE framework – a dynamic risk-based lung cancer screening framework delivering personalized policies – has been published in Cancer. Accompanying this publication, an editorial exploring how the ENGAGE framework will continue to enlighten and shape the ongoing conversation around the delivery of personalized lung cancer screening has also been published.

New BMJ publication investigating the real-world impact of a healthcare ML/AI system by BMI PhD candidate Ben Marafino

August 11, 2021.  A new article in BMJ authored by BMI PhD candidate Ben Marafino, together with Professor Mike Baiocchi (Epidemiology and Population Health) and BMI PhD alum (now adjunct professor of DBDS) Alejandro Schuler, is among the first to characterize the real-world impact of a ML/AI system at scale in healthcare. The study, which was carried out in collaboration with Kaiser Permanente, examined data on nearly 2 million patients before and after the implementation of a predictive-algorithm driven intervention to reduce readmission, and employed a novel hybrid regression discontinuity/difference-in-differences design approach. Co-authors include Gabriel Escobar, Vincent Liu, and Colleen Plimier, all of Kaiser Permanente.

Tackling COVID-19 Among Prison Populations in California and Beyond

August 9, 2021. The Stanford Health Policy prison project team, which includes DBDS/BMI researcher Elizabeth T. Chin, is out with two more studies to help prisons prevent and reduce the spread of the coronavirus. Read the Stanford Health Policy News release


Work on clinical trial design accounting for efficient estimation published in IJB by Adjunct Prof. Alejandro Schuler

August 6, 2021. Adjunct Professor Alejandro Schuler's paper, “Designing efficient randomized trials: power and sample size calculation when using semiparametric efficient estimators," was published in the International Journal of Biostatistics. The work proposes a method that leverages efficient estimators to prospectively design smaller and faster randomized trials while attaining the same power that would otherwise only be possible with a larger sample size.