No sooner had I given my keyboard the final tap to publish the conclusions of my work yesterday on medical apps than the first item hit my inbox that suggest that certification is a flawed proposition.
The suggestion of this iMedicalApps article is that the Happtique saga has shown certification to be impossible. Instead it is suggested that people make up their own minds based on peer review on sites (you’ve guessed it) such as theirs, and a greater understanding of apps. The key paragraph for me is :
For fitness and healthcare apps in general I have no problem with that conclusion. For medical apps where there is a potential to do significant harm, then I disagree totally. Even if the clinician finds “references for each equation used” (very few apps currently give this information) and has the necessary subscriptions to access all the appropriate learned journals (few do), how can the clinician be sure that the equations have been coded correctly? Does the clinician also need to be a coding expert as well? Of course in reality the code is rarely if ever accessible. And then can the clinician be sure this is the correct code version to use for that version of the operating system?
To me there is no regulatory difference between the software in say a sphygmomanometer and that that might, as standalone software, perform a similar task in a smartphone. No-one would dream of expecting all professional users to understand the complex algorithms used in oscillometric sphygs in order to produce the same results as the original auscultatory technique, so all sphygs sold in the EU for professional use are CE marked; similarly risky medical apps should be certified too.
Throughout recent history, regulation has had to be introduced when customers could not immediately detect whether a product was as advertised, and when the result on customers of it not being as advertised was serious, such as the practice of using alum to whiten bread (as most notably exposed in the Great Bread Scandal of 1757). There should be no exception for apps that can seriously damage health.
Sure there are many to certify, however the time taken will be far less than the time taken for every clinician to look up every equation used in every app and confirm its provenance, even before they have learned iOS and Android coding! Perhaps we need to educate people that safe and effective medical apps are not the same as the download-and-delete 99p wonders on app stores; such medical apps should be expensive because they are the result of extensive research and testing in preparation for certification, and will need to be managed by their suppliers throughout their life. With money available to do proper evidence gathering too, it should be possible to prove their efficacy beyond doubt.