This post represents the first in a series of THREE posts that I have written at the request of readers. After my slightly sarcastic review of iSonea where I sort of lumped iSonea and ResApp together in to the “turd” category, but without any real elaboration, I had several requests for clarification. Incidentally, I appreciate it when people do that and I think it’s fair. Also, if I ever make a statement that you think is unsubstantiated, let me know and I will write a follow-up. Sometimes I do make general statements where I have done some background reading but haven’t made a specific citation, and I am happy to clarify.
So… ResApp (ASX : RAP). To be honest, part of the reason why I glossed over this company was because I don’t think there is anything much of excitement to say about it. I like the ASX ticker code, and that’s about it. Sure it’s a neat idea, I am not knocking the basic concept – indeed that’s part of the appeal of this baby company to its shareholders. It’s not very hard to understand what ResApp does. There is also no doubt that the smart phone / “phablet” is going to become a concentration point for all kinds of healthcare service provision, diagnostics, wearables, etc. A quick stroll though your app store of choice shows just how quickly the health application market is taking off.
I guess there are four main reasons why I don’t think this company is a viable prospect :
- The technology may or may not actually work. I don’t think the company has moved the technology along very far from its academic origins at the University of Queensland and any publication or validation I have read referred to incredibly small data sets and very limited clinical repertoires. I have had some considerable experience in taking machine learning technologies to the market (including through FDA channels) and I am skeptical that ResApp has done enough to even ascertain whether the technology has the potential for clinical validation. For example, if you were going to validate that the application could detect pediatric pneumonia, you’d need to do a very significant study and compare with chest X-ray infiltrates and microbiology. Not only is this a tough study to design and execute, but it is also relatively expensive and there is huge variability in morphologic/infiltrate radiographic characteristics, which has to intrinsically impact respiratory acoustics = large patient numbers. I honestly view it as more of an “idea” than a technologically robust implementation. Moreover, automated analysis of respiratory acoustics is something that has been done for close to two decades and there are a hell of a lot of ways to tackle the problem. I see nothing in ResApp’s technology strategy that would block a potential competitor.
- I don’t actually think ResApp is doing much development. If you look at the last 4C filing, this company basically pays their board more money than they spend on R&D/product development. It may be an anomalous “snapshot” given the proximity to some modest financing, but if you look at the use of proceeds statement from their placement, why would you invest in this company? It’s going to spend more money on corporate and administration costs than product development. It’s proposing to sell a consumer health product, but it is going to spend $100,000 on marketing. Certainly with the cash the company is proposing to spend on “clinical trials”, we are probably not talking about X-rays, microbiology, patient follow-up or anything run through a CRO. Basically, this company raised some cash that is hardly going to move the needle on a product, and is – on the whole – wastefully deployed.
- There is no doubt that we have difficulty with early-stage funding in Australia in healthcare/life sciences. But the same cannot be said for IT and applications. If the ResApp technology really had some potential, it would have been far more cost-effective to have developed it as a private eHealth company. I know plenty of investors that would have taken a look at this and if you deduct the offering expenses and corporate/admin costs (mostly associated with being a public company) all you are really talking about is a $2-2.5m raise. Peanuts. You may view this comment as a bit petty, but if there was really something “there”, it didn’t need to be backed into Narhex for this level of capitalisation. I don’t have fundamentally anything against re-purposing a shell, that’s all fine and dandy if the new asset is good, but when I look at the use of proceeds, the real extent of development (noting that there is no realistic cost provision for quality/regulatory, FDA advisory, etc.), etc. and combine this with the “usual suspects” behind the deal, I don’t think it’s more than just a story.
- From a product development and regulatory vantage, the only way that there is going to be any real value capture in this technology is if it becomes an “aid to diagnosis”. That’s a pretty high regulatory bar for an automated diagnostic product in a consumer technology platform. Not only is ResApp not even remotely financed to deliver on that level of product development and validation, but its not even clear that if it did that it would change patient care anyhow. So it either stays as a “novelty” app that someone pays a few bucks for (noting that millions of people have respiratory conditions, so this could turn into rather a large number) or it … ah… stays as a novelty app that someone pays a few bucks for that has an FDA marketing authorisation. That’s the problem with “apps”. There is also plenty of competition out there from other apps that claim to do many of the same things, or at least provide “benchmarks” for diagnosing different conditions – many of them for free. It could be that ResApp plans to make money out of managing the physician referral process but that business model hasn’t really yielded much so far by way of actual successes either.
I personally don’t believe that a machine learning algorithm can do extremely detailed “blanket” analysis of respiratory sounds. All the studies that have been done to date are for a specific patient class against “normal”. That’s a pretty low bar test of efficacy. But let’s say they can – it doesn’t change the fact that diagnosis of respiratory conditions is far more than just advanced acoustic analysis of coughs. There are plenty of other clinical symptoms that are required to make an accurate diagnosis, precisely because sounds don’t tell the whole story. Many asthmatics wheeze, but not all wheezers are asthmatics. But what I really don’t like is the articulated “vision” that ResApp will move the needle in the “developing world”. That’s mostly rubbish – and nothing more than a vague hand-waving justification for throwing some big numbers out there. When diagnosing respiratory conditions in the developing world, the most important clinical information is microbiology, and that isn’t going to get delivered by this particular app company.
Tomorrow I will be posting on “Paradigm Pharmaceuticals” (hint : no paradigm shift) and on Thursday, NanoSonics.