Kirstie Jane Whitaker

The DECOVID database contains harmonized pseudonymized electronic health record (EHR) data on all adult (≥18 years old) patients presenting to two large, digitally mature centers in the United Kingdom between 1 January 2020 and 28 February 2021, with follow-up until at least 28 March 2021. The database was originally developed to support the COVID-19 response but is now available via the PIONEER data hub for researchers to explore a wide range of research questions, including exploratory analyses, risk factor assessment, prediction modeling, and comparative effectiveness studies. Raw data were extracted from local EHRs and transformed into a standardized form (Observational Health Data Sciences and Informatics-Common Data Model version 5.3.1). The database includes 165,420 patients across 256,804 hospital presentations. For these patients, highly granular data are available, including patient demographics, longitudinal vital signs, physiology, treatments, laboratory findings, clinical diagnoses, and outcomes. There are 10,030 patients with COVID-19, of whom 1472 died in hospital.

Jupyter Book allows researchers and educators to create books and knowledge bases that are reusable, reproducible, and interactive. This new foundation introduces a scalable way to publish interactive computational content, support structured metadata, and enable content reuse across contexts.

This is a podcast version of the closing panel discussion at The 18th Munin Conference on Scholarly Publishing (Tromsø, Norway, 8–10 November 2023). The main theme of the conference was scholarly communication in its myriad of forms. The push for a reform of research assessment argues for precisely the need to recognize and reward a wider set of research activities and competencies, not least in order to ensure that science – and scientics – remain in close contact with society. The demands for change will surely impact how and where research is disseminated and communicated. A key question then is how to evaluate research activities and outputs beyond academic publications. The panelists reflected on questions such as: how do we measure the impact of citizen science, or document quality in innovative research practices and scholarly outputs? And how to ensure consensus on assessment methods for a wider and more inclusive range of research activities within a diverse academic landscape? The panel consisted of champions of research assessment reform (Yensi Flores Bueso, University of Washington / University College Cork; Kirstie Whitaker, The Alan Turing Institute) and university leaders (Hervé Dole, vice-president for arts, culture and society at Université Paris-Saclay; Jan-Gunnar Winther, pro-rector for research and development, UiT). Questions were prepared by Tanja Larssen (open science adviser, UiT) in collaboration with Per Pippin Aspaas (head of library research and publishing support, UiT), who chaired the panel on the day. First published as a podcast: December 19, 2023. A video recording of the same session – with metadata providing more detailed information on each panelist – can be found at https://doi.org/10.7557/5.7276.

Watch VIDEO. The main theme of the Munin conference is scholarly communication in its myriad of forms. The push for a reform of research assessment argues for precisely the need to recognize and reward a wider set of research activities and competencies. The demand – and initiative – for change will surely impact how and where research is disseminated and communicated. A key question then is how to evaluate research activities and outputs beyond academic publications. With this panel we welcome a discussion on research assessment that brings together perspectives from all the topics of this year’s Munin conference. We invite you to add your insights and diverse expertise to this panel that will reflect on questions such as: how do we measure the impact of citizen science, or document quality in innovative research practices and scholarly outputs? And how to ensure consensus on assessment methods for a wider and more inclusive range of research activities within a diverse academic landscape?

Watch VIDEO. It has been 40 years since the launch of the GNU project and the Free Software Movement in September 1983. Open Source followed in 1998, Open Access in 2002, Citizen Science in 1995, the Bermuda Principles for open genomic data in 1996, and Open Science was first used as a phrase in 1985. Elinor Ostrom published her Nobel Prize winning "Governing the Commons" in 1990. There has been exceptional adoption and promotion of open research practices in the last 20 years, and there has also been significant drift from the original vision of democratising access to knowledge. In the age of artificial intelligence, are we meeting our ethical responsibilities to use data responsibly? Are we using scholarly communications to dismantle oppressive and exclusive power structures? What more work must we undertake? In this keynote presentation, Dr Kirstie Whitaker will identify core tenets of open research in 2023 and propose an integration with the SAFE-D principles for responsible research and innovation. Developed by David Leslie at The Alan Turing Institute in 2019, these principles facilitate reflection and self-assessment of the safety and sustainability, accountability, fairness and non-discrimination, and explainability and transparency of a research and innovation outputs, including a consideration of data quality, integrity, protection and privacy. Kirstie will explore with the audience whether the scholarly publishing community is ready to address the biggest societal challenges of our time, how we assess and incentivise responsible and ethical development of socio-technical solutions, and how our infrastructure can facilitate interdisciplinary teams to create outputs that are greater than the sum of their parts. All attendees will be invited to join the 800+ members of the open source, open collaboration, and community-driven Turing Way community. Developed on GitHub under open source licences, our shared goal is to provide all the information that researchers and data scientists in academia, industry and the public sector need to ensure that the projects they work on are easy to reproduce and reuse.

Childhood adversity is one of the strongest predictors of adolescent mental illness. Therefore, it is critical that the mechanisms that aid resilient functioning in individuals exposed to childhood adversity are better understood. Here, we examined whether resilient functioning was related to structural brain network topology. We quantified resilient functioning at the individual level as psychosocial functioning adjusted for the severity of childhood adversity in a large sample of adolescents (N = 2406, aged 14–24). Next, we examined nodal degree (the number of connections that brain regions have in a network) using brain-wide cortical thickness measures in a representative subset (N = 275) using a sliding window approach. We found that higher resilient functioning was associated with lower nodal degree of multiple regions including the dorsolateral prefrontal cortex, the medial prefrontal cortex, and the posterior superior temporal sulcus (z > 1.645). During adolescence, decreases in nodal degree are thought to reflect a normative developmental process that is part of the extensive remodeling of structural brain network topology. Prior findings in this sample showed that decreased nodal degree was associated with age, as such our findings of negative associations between nodal degree and resilient functioning may therefore potentially resemble a more mature structural network configuration in individuals with higher resilient functioning.

The moderation of user-generated content on online platforms remains a key solution to protecting people online, but also remains a perpetual challenge as the appropriateness of content moderation guidelines depends on the online community that they aim to govern. This challenge affects marginalised groups in particular, as they more frequently experience online abuse but also end up falsely being the target of content-moderation guidelines. While there have been calls for democratic, community-moderation, there has so far been little research into how to implement such approaches. Here, we present the co-creation of content moderation strategies with the users of an online platform to address some of these challenges. Within the context of AutSPACEs – an online citizen science platform that aims to allow autistic people to share their own sensory processing experiences publicly – we used a community-based and participatory approach to co-design a content moderation solution that would fit the preferences, priorities and needs of its autistic user community. We outline how this approach helped us discover context-specific moderation dilemmas around participant safety and wellbeing and how we addressed those. These trade-offs have resulted in a moderation design that differs from more general social networks in aspects such as how to contribute, when to moderate, and what to moderate. While these dilemmas, processes and solutions are specific to the context of AutSPACEs, we highlight how the co-design approach itself could be applied and useful for other communities to uncover challenges and help other online spaces to embed safety and empowerment.

Objective: The accurate prediction of seizure freedom after epilepsy surgery remains challenging. We investigated if (1) training more complex models, (2) recruiting larger sample sizes, or (3) using data‐driven selection of clinical predictors would improve our ability to predict postoperative seizure outcome using clinical features. We also conducted the first substantial external validation of a machine learning model trained to predict postoperative seizure outcome.Methods: We performed a retrospective cohort study of 797 children who had undergone resective or disconnective epilepsy surgery at a tertiary center. We extracted patient information from medical records and trained three models—a logistic regression, a multilayer perceptron, and an XGBoost model—to predict 1‐year postoperative seizure outcome on our data set. We evaluated the performance of a recently published XGBoost model on the same patients. We further investigated the impact of sample size on model performance, using learning curve analysis to estimate performance at samples up to N = 2000. Finally, we examined the impact of predictor selection on model performance.Results: Our logistic regression achieved an accuracy of 72% (95% confidence interval [CI] = 68%–75%, area under the curve [AUC] = .72), whereas our multilayer perceptron and XGBoost both achieved accuracies of 71% (95% CI_MLP = 67%–74%, AUC_MLP = .70; 95% CI_{XGBoost own} = 68%–75%, AUC_{XGBoost own} = .70). There was no significant difference in performance between our three models (all p > .4) and they all performed better than the external XGBoost, which achieved an accuracy of 63% (95% CI = 59%–67%, AUC = .62; p_LR = .005, p_MLP = .01, p_{XGBoost own} = .01) on our data. All models showed improved performance with increasing sample size, but limited improvements beyond our current sample. The best model performance was achieved with data‐driven feature selection.Significance: We show that neither the deployment of complex machine learning models nor the assembly of thousands of patients alone is likely to generate significant improvements in our ability to predict postoperative seizure freedom. We instead propose that improved feature selection alongside collaboration, data standardization, and model sharing is required to advance the field.

Objective: Neurosurgery is a safe and effective form of treatment for select children with drug‐resistant epilepsy. Still, there is concern that it remains underutilized, and that seizure freedom rates have not improved over time. We investigated referral and surgical practices, patient characteristics, and postoperative outcomes over the past two decades.Methods: We performed a retrospective cohort study of children referred for epilepsy surgery at a tertiary center between 2000 and 2018. We extracted information from medical records and analyzed temporal trends using regression analyses.Results: A total of 1443 children were evaluated for surgery. Of these, 859 (402 females) underwent surgical resection or disconnection at a median age of 8.5 years (interquartile range [IQR] = 4.6–13.4). Excluding palliative procedures, 67% of patients were seizure‐free and 15% were on no antiseizure medication (ASM) at 1‐year follow‐up. There was an annual increase in the number of referrals (7%, 95% confidence interval [CI] = 5.3–8.6; p < .001) and surgeries (4% [95% CI = 2.9–5.6], p < .001) over time. Duration of epilepsy and total number of different ASMs trialed from epilepsy onset to surgery were, however, unchanged, and continued to exceed guidelines. Seizure freedom rates were also unchanged overall but showed improvement (odds ratio [OR] 1.09, 95% CI = 1.01–1.18; p = .027) after adjustment for an observed increase in complex cases. Children who underwent surgery more recently were more likely to be off ASMs postoperatively (OR 1.04, 95% CI = 1.01–1.08; p = .013). There was a 17% annual increase (95% CI = 8.4–28.4, p < .001) in children identified to have a genetic cause of epilepsy, which was associated with poor outcome.Significance: Children with drug‐resistant epilepsy continue to be put forward for surgery late, despite national and international guidelines urging prompt referral. Seizure freedom rates have improved over the past decades, but only after adjustment for a concurrent increase in complex cases. Finally, genetic testing in epilepsy surgery patients has expanded considerably over time and shows promise in identifying patients in whom surgery is less likely to be successful.

Objective: The accurate prediction of seizure freedom after epilepsy surgery remains challenging. We investigated if 1) training more complex models, 2) recruiting larger sample sizes, or 3) using data-driven selection of clinical predictors would improve our ability to predict post-operative seizure outcome. We also conducted the first external validation of a machine learning model trained to predict post-operative seizure outcome. Methods: We performed a retrospective cohort study of 797 children who had undergone resective or disconnective epilepsy surgery at a single tertiary center. We extracted patient information from medical records and trained three models – a logistic regression, a multilayer perceptron, and an XGBoost model – to predict one-year post-operative seizure outcome on our dataset. We evaluated the performance of a recently published XGBoost model on the same patients. We further investigated the impact of sample size on model performance, using learning curve analysis to estimate performance at samples up to N =2,000. Finally, we examined the impact of predictor selection on model performance. Results: Our logistic regression achieved an accuracy of 72% (95% CI=68-75%, AUC=0.72), while our multilayer perceptron and XGBoost both achieved accuracies of 71% (95% CI _MLP =67-74%, AUC _MLP =0.70; 95% CI _{XGBoost own} =68-75%, AUC _{XGBoost own} =0.70). There was no significant difference in performance between our three models (all P >0.4) and they all performed better than the external XGBoost, which achieved an accuracy of 63% (95% CI=59-67%, AUC=0.62; P _LR =0.005, P _MLP =0.01, P _{XGBoost own} =0.01) on our data. All models showed improved performance with increasing sample size, with limited improvements above N =400. The best model performance was achieved with data-driven feature selection. Significance: We show that neither the deployment of complex machine learning models nor the assembly of thousands of patients alone is likely to generate significant improvements in our ability to predict post-operative seizure freedom. We instead propose that improved feature selection alongside collaboration, data standardization, and model sharing is required to advance the field.