No problem and thanks for the welcome, TG
Regarding your counterpoints, which are also welcome and I hope you don't mind me batting the ball back sometimes:
1. "Don't believe a different socioeconomic part of the world will change the accuracy, first test in Indonesia and also the recent backing by Doctors Without Borders who have been using algorithm for approx 7 months."
Are the Indonesian results in the Australian study, or those from MSF? I thought the study was from WA...Joondalup (a busy ED) and Princess Margaret? I realise there was some early preliminary testing done in Indonesia but I'm not sure the study design was the same. I could stand corrected here.
That said, kids who present to the best hospitals in the USA might be different from those in those in 3rd world settings. The latter might be sicker, relatively late presentations, malnourished, dry as a chip. Then again they may not, but in study methodology, the potential confounders should be considered and controlled for if possible. Matching was mentioned by another poster here.
2. "far more accurate than the ear"...is it? I thought the comparison was clinical diagnosis vs cough test. Clinical diagnosis takes into account a range of parameters of which auscultation is but one. I think you might be right, but it is only speculation. The study does not specifically compare auscultation findings with cough findings, unless I missed it. Also some stethoscopes are better than others. A good Littman knocks the socks off a disposable ward set of tubes acoustically. You can guess what might be available in 3rd world settings more often than not. Also some ears are better than others, ie operator dependent. A grunting, hyperventilating child in a typical ED full of ambient noise and alarms makes it all the more challenging I expect.
3. Old school GPs are going to lap it up (words to that effect).
Old school GPs have a wealth of experience to draw on and the basic issue is 'sick or not sick?' If a GP can take the temperature, observe respiratory effort and take oxygen saturation (granted most probably don't have a paediatric probe), then the question is broadly one of method of eliciting chest noise. I'm not as sure as you seem to be this group of GPs would be as eager as the investor for uptake. They've seen it all before, again and again. Tests, drugs, you name it. Remember, there is a potential big difference between the accuracy of this test in hospital and community settings, because the prevalence of pneumonia is higher as a proportion of all respiratory presentations to the ED than to a GP. The FDA website makes this point very clear, namely the prevalence of the condition in the target population can affect the PPA and NPA. They may wait for a community based study before going further, if at all. As for nurse practitioners and young graduates, as well as telehealth, I would think them more likely the early adopters. But of course, I could be wildly wrong.
4. "This is why the algorithm is so powerful, it can detect and differentiate between different illnesses far far better than the current gold standard of care, which is doc/stethoscope/X-ray/blood test. (We all know the bits in between doc & test)".
The primary endpoint is **agreement** with clinical diagnosis and radiological diagnosis of pneumonia in SMARTCOUGH-C. This is the big enchilada. What happens if the cough test is right and the clinical and radiological diagnosis wrong? You get less agreement. This complicates the issue.
As a standalone test, if expressed as likelihood ratio (sans confidence intervals) it sound favourable, but maybe not dramatic by itself, at least I can't see it anyway. However, they might be significantly better when applied to the clinically more uncertain age group and with the right symptoms and observable signs also inputted.
Which is still good news and may be better than good.
As for superbugs, this is an obvious issue to anyone who follows the news, let alone people who have to make treatment decisions. But to counterweight the issue, given the uncertainty in diagnosis, there will by necessity be a number of people who get antibiotics who would have managed without them, in order to treat the one that needs it. I don't know what that number is but I'd bet there is a study on it.
In the end, I know if I was a parent in the 3rd world, with only a few bob, I'd be more inclined to spend it on antibiotics bought over the counter without a script and put food on the table than a fancy test. Having said that, if charitable or government healthcare organisations foot the bill, then I can't see anyone really saying no to it as it is non-invasive and results would be close to immediate, provided it works.
Expand
(20min delay)