As an internal medicine physician (MD in USA) myself, I have been following the development of this App. with both great anticipation, and with some concern. After downloading the App on my I-phone (to trial it), I inputted the typical symptoms of a left leg blood clot (DVT), providing (by answering the posed questions) one symptom at a time at the appropriate prompting of “Cara”, the App’s assistant.
And I was pleasantly surprised that the App. correctly told me that I most likely had a DVT (deep venous thrombosis), giving me the numeric % probability of this diagnosis. The App also told me that I had a lesser probability of having venous thrombophlebitis (again, a correct lower probability diagnosis), and once again gave me the percent probability.
The App then explained how it arrived at those conclusions, and provided the usual treatment options when I chose those selection tabs. It then recommended that I contact my doctor for confirmation of diagnosis and for treatment. It also gave me the options of checking out predisposing factors and preventive measures
I then tried the App again using a second, cleverly feigned (as only a physician can do) condition, … this time a migraine headache. And once again, the App did a great job asking me the right questions in order to arrive at the most probable diagnoses.
And once again, the App appropriately offered “migraine headache” as the most probable diagnosis (again with probability percentage), as well as other possible diagnoses with lower probabilities. And once again, options were offered to read about predisposing factors, treatment options and preventive measures for migraine headaches.
And once again, I was advised to see my doctor for diagnosis confirmation and for pharmacologic treatment if over the counter treatments were not effective. The App. also discussed non-pharmacologic interventions and over the counter pain meds (Tylenol).
I was also pleased to see that the App acknowledged that, (even though it was designed by physicians), it could NOT guarantee that my true malady was among the probable diagnoses offered. It also cautioned repeatedly
that the app did NOT take the place of a physician.
My only recommendation to the developers is to consider replacing the obtuse medical terminology in the questions with simpler language. Or, alternatively, to define the medical terms used, when it is part of a question. As an example, when questioning me about my symptoms, the App asked if I had ever been diagnosed with venous thromboembolism. It would have been better to ask if I had ever been diagnosed with a blood clot.
So, in conclusion, as a physician (myself), I can honestly say that (at least for the conditions I chose), this new, long-awaited medical App is living up to my high expectations; and, equally importantly, does not have the concerns (including inaccuracies and trying to replace physicians)that I had feared.
The process of medical data collection (through intelligent questioning with option choices provided) is sound and thorough. The App arrives at the same differential diagnoses (most probable diagnoses) as would a physician. And finally, the App does not market itself as, nor purport to be, a substitute for a live physician, nor to have 100% infallibility.