VIS-SIG Blog
Wonderful Wednesdays December 2024
Use Visualizations to answer the following questions: Inter-goniometer: do the two goniometers agree? Intra-goniometer: do the goniometers measure consistently the knee joint angle across the three repeated measurements?
Wonderful Wednesdays November 2024
The challenge was about improving a bad chart.
Wonderful Wednesdays October 2024
Why personalized dosing? „Again and again, it has been shown that certain patients benefit more from individual drugs, while others suffer more heavily from side effects… …and still they are usually treated with a standard dose." Free translation of a quote from Prof. Dr. Dr. Gerd Geisslinger, Executive Director, Fraunhofer Institute for Translational Medicine and Pharmacology, Frankfurt
Wonderful Wednesdays September 2024
Visualise the pattern of events (disease progression, death, etc) on a summary data level or on a patient level data. Highlight differences between treatments or subgroups.
Wonderful Wednesdays August 2024
When a new patient-reported outcome (PRO) is developed, there is interest in deriving the following, to support interpretation of data collected using the new tool: Minimal Clinically Important Difference (MCID) / Responder Definition (RD), Meaningful Between-Group Difference, Meaningful Score Categories. These are usually derived by using an existing, validated measure as the ‘gold standard’ to quantify e.g. disease status.
Wonderful Wednesdays July 2024
The data set is based on a platform trial design. How can we investigate the effectiveness of early stopping for futility? What is the relationship between the quality of the surrogate endpoint and the average number of patients enrolled? Is there an interaction with futility stopping and treatment effect scenario? How do FWER and FWER_BA differ from each other with respect to treatment effect scenario and maximum number of cohorts? Can you find other interesting patterns in the data?
Wonderful Wednesdays June 2024
The data set is based on a platform trial design. How can we best show the effect of sample size on the power/type 1 error? Does sample size affect PTP and Disj_Power in the same way? How about in interaction with the type of data sharing? What is the relationship between number of cohorts, cohort inclusion rate and power? How can we investigate the effectiveness of early stopping for futility? What is the relationship between the quality of the surrogate endpoint and the average number of patients enrolled? Is there an interaction with futility stopping and treatment effect scenario? How do FWER and FWER_BA differ from each other with respect to treatment effect scenario and maximum number of cohorts? Can you find other interesting patterns in the data?
Wonderful Wednesdays May 2024
In almost every study there is a flow chart to present the patient's path through the trial. What could the next level of study flow charts look like? Sandard study flow charts are basically organised lookup tables. A true visualisation can do much more. For example colour coding or indenting boxes can help organising. Also additional information can be embedded. This can easily done in R. For patient disposition Sankey diagrams can be extremely useful. This is even more powerful if used interactively. People charts even go without explanation.
Wonderful Wednesdays April 2024
As statisticians, we often have to deal with power or sample size calculations. It can be very beneficial to visualise the impact that different factors have on the sample size or power. Produce data visualisation(s) or share tools that help your audience understand how power and sample size calculations work and what impacts the actual results of these calculations.
Wonderful Wednesdays March 2024
Produce data visualisation(s) providing insights into how microbiome profiles change over time and what the differences are in microbiome profiles between Pre-Term Birth and Term Birth.
Wonderful Wednesdays February 2024
Find a plot in a healthcare related publication and improve the visual presentation with only the data available from the publication.
Wonderful Wednesdays January 2024
Best of 2023
Wonderful Wednesdays December 2023
Demographic data display: The challenge is based on CDISC data. It is a standard ADSL data set for subject level data (demographics and baseline).
Wonderful Wednesdays November 2023
Gatekeeping Procedure: Results are provided from a simulation of a phase III study in schizophrenia patients, investigating efficacy at 3 dose levels (Low, Medium, High) versus placebo. The study includes a primary endpoint and two key secondary endpoints. The study includes a total of 9 hypotheses, and multiplicity adjustments are required to control for type I error.
Wonderful Wednesdays October 2023
Leverage visual analytics to identify data issues.
Wonderful Wednesdays September 2023
The PRIMUM study aimed to improve medication appropriateness in elderly patients with multimorbidity. A complex intervention involving general practitioners (GPs) and their healthcare assistants (HCA) was developed and compared to usual care. Check for a treatment effect on the MAI score after the intervention. Visualize the effect of adjusting the treatment effect for the MAI score at baseline.
Wonderful Wednesdays August 2023
Benefit-Risk Assessment: For a drug to be approved for marketing, regulatory authorities must determine that the drug is effective and that its expected benefits outweigh its potential risks to patients. An Alzheimer’s Disease drug might have a positive benefit-risk profile for lowering blood pressure. To generate a hypothesis, an exploratory benefit-risk assessment will be conducted.
Wonderful Wednesdays July 2023
US Disease Incidence
Visualisation SIG Showcase
This is an entry accompanying the poster at the PSI conference 2023.
Wonderful Wednesdays May 2023
Data visualisations to present Clinical Global Impression Data and to provide comparisons between the different groups (e.g. based on response differences or odds ratios for the different response categories).
Wonderful Wednesdays April 2023
This month’s challenge will be to visualise diary data, whereby patients complete a diary (grading a number of symptoms), each day. For this challenge, we consider a hypothetical diary, consisting of 6 symptoms (Exhaustion, Aching, Tenderness, Depression, Loneliness, Anxiety). 300 subjects (200 active, 100 placebo), grade each of these symptoms on a 4-point scale (0 = None, 1 = Minor, 2 = Moderate, 3 = Extreme) each day, for 100 days.
Wonderful Wednesdays March 2023
This month’s challenge was to improve a published visualisation. The visualisation displayed the mean (and SD) Dermatology Life Quality Index (DLQI) score for 12 studies on alopecia areata. The publication had visualised these using ‘dynamite plots’, a type of visualisation which has received criticism in the past.
Wonderful Wednesdays December 2022
The data is an excerpt of the official VAERS data created by US Food and Drug Administration (FDA) and Centers for Disease Control and Prevention (CDC) that may be associated with vaccines.
Wonderful Wednesdays November 2022
The EORTC QLQ-C30 is a 30-item questionnaire that has been designed for use in a wide range of cancer patient populations and is a reliable and valid measure of the quality of life in cancer patients. It includes a number of different scales, but this challenge is focussed on the global health and quality of life scale (QL).
