Abernethy: FDA does not directly regulate non-diagnostic healthcare workflow software
FDA Chief Information Officer · Food, Drugs, and Tech: 100 years of Public Health · Jan 6, 2020 · at 25:20
Dr. Amy Abernethy, FDA Principal Deputy Commissioner and CIO, outlines the scope of FDA authority regarding healthcare and life sciences software applications.
0:00 / 0:33exact quote · 33.5s
▶ Watch the full episode on YouTube → 720p mp4 · rendered on demand · StarZero watermark
“So it may not necessarily be a software product that's acting As a diagnostic or treatment activity, but it's a software product that's intended to support life sciences more globally, whether that's to make clinical trials more efficient, to match patients to clinical trials, to curate data, to help do workflow in the hospital, and all of those kinds of software products we don't directly regulate, but importantly, those products also need some good signals of here's what good looks like, and here's how you should think about good software controls in those settings as well.”
quote is from the automated transcript, cleaned for reading: filler sounds and stutters are removed, nothing is rephrased. names can be misheard (the analysis reads context, assessments check outside sources). how →
More from FDA Chief Information Officer
Abernethy: HIPAA is outdated and struggles to protect individual privacy today
“Our laws of the past, HIPAA, really contemplated a different world than we live in right now, whereby essentially in 2019 and going forward, it's very hard to maintain privacy of any individual.”
Abernethy: HIPAA-de-identified health data remains likely re-identifiable through modern analytics
“Even information that is de-identified from a HIPAA perspective actually still is probably re-identifiable even in our context.”
Abernethy: The FDA has approved pharmaceuticals based on Phase 1 data
“We've seen drugs approved based on phase one data.”
Abernethy: FDA is testing machine learning to predict border truck inspections
“And so right now we've got an experiment going on where we're looking at machine learning based prediction of which trucks we should inspect on the border.”
Abernethy: FDA jurisdiction over DIY gene-editing requires commercial intent and labeling
“Practically speaking, if the kit was going to be approved for commercial purposes with claims and labeling, et cetera, that's when it starts to get into the FDA perspective. It gets really murky when we live in this landscape of the internet without claims.”
Abernethy: FDA pilots preconditions to allow medical software updates without full review
“Can we come up with, essentially, preconditions for software updates? So that if there are strong quality controls in the way software is developed, well understood product performance as an expect, in terms of the expectation updates, can you now have algorit…”