Sage Presentation

Below are a few resources that cover this topic. Traditional 'current-future state', 'benchmarking', and 'survey/structured interview' based approaches are generally not used in medical affairs, but some of the thinking that goes into these, if incorporated helps push teams towards more systematic thinking and towards considering cutting edge technologies. Accordingly, some of the links below for these gap analysis methods are not specific to medical affairs.

Van Norman 2016 https://doi.org/10.1016/j.jacbts.2016.03.002

Clinical trial designations into different phases is not entirely consistent across the industry. Basic guidance is available in the sites linked below. "Lawinsider.com" is particularly useful because of the extensive examples of actual use

"Patient Recruitment in Clinical Trials" by Spilker and Cramer

FDA's Digital Health Center for Excellence
EMA: Qualification of digital technology-based methodologies to support approval of medicinal products
Ethical, Legal and Social Issues of Digital Phenotyping as a Future Solution for Present-Day Challenges, Ana Tomičić et al - 2021
An Ethics Checklist for Digital Health Research in Psychiatry, Shen et al, 2022
R2D3 Machine Learning Tutorial: One of the best dynamic infographics available online to get a sense of how machine learning works, using a representative example
Future of Life open letter: Some years ago, a number of cross-discipline experts got together to work out what 'safety' meant in a world moving rapidly towards reliance on AI. I wonder what they think now?
Black box problem: There is a 'black box' problem when we think about the meaning of 'answers' that AI is providing. While many industries have now accepted this, and increasingly so has the public, it isn't clear whether in clinical research this should be acceptable
An article on multimodal data fusion: Multimodal data fusion (accessing and analytic combining data from divergent sources) may allow advances in signal detection in clinical research
Slides on a new paradigm in clinical trials, which I presented in a workshop a few years ago. I added some more recent slides on digital eendpoints and their use

Figure from Tomičić et al [linked above]. The authors state: "While we noted 207 different terms for devices used in the field of digital phenotyping, the most common terms relating to Digital Phenotyping referred to in our corpus were: Quantified self; Digital Phenotype; Self-tracking; Wearables; Biometric data; Digital Biomarkers; Internet of Things; Lifelogging; Activity tracking; and Self-surveillance; a series of terms such as Ubimedicine; Reality mining; Psychoinformatics; Innernet were used seldom. The lexical field of Digital Phenotyping (Fig) shows to be inconsistent in its use in the literature"

In putting together a peri- and post-launch evidence generation plan, it is important to understand the characteristics of the placebo response seen across indications, various endpoints and studies
Placebo responses are complex phenomena influenced by subject characteristics, conditions under study, amount and quality of site interactions, how and what scales are administered, frequency of scale administration, environment that subjects experience within and outside studies, physical characteristics of investigational agents (e.g. size and color), and many other factors
Slides on interpretation of clinical trial results in indications with historically high placebo response rates

Evidence Generation Presentation - February 27, 2023