CertiHealth^TMCertified digitization in the healthcare sector

With the move towards digitization and the change in working methods, patient dossiers, for example, are increasingly being digitized. What applies to the transfer of paper documents and electronic form also applies analogously to the recording of speech. File notes and medical notes can now be recorded, interpreted and digitized using specialized applications. Imaging procedures and video recordings must be cataloged and assigned.

Many of these processes are still in their infancy. However, information about patients and particularly sensitive patient data must be especially protected. Handling them is strictly regulated and requires increased controls. The new Data Protection Act, for example, requires a data protection impact assessment to be carried out BEFORE the planned solution is put into operation.

Many things can go wrong, from classic errors during data capture (duplicates, missing pages, incorrect recognition) to incorrect semantic interpretation of the captured data.

In order to ensure that the digitization processes correspond to the current state of the art and are handled properly, we have expanded our testing services to include the testing and certification of such processes and products.

You can find a short presentation in English here.

Case study 2:
Language transformation

One provider makes it possible to record conversations between doctor and patient using a cell phone. The app is also used to convert and interpret the spoken text. The result is used for documentation and is transferred in text form to the electronic patient record (ePD).

What are the challenges? Speech recognition algorithms normally use artificial intelligence methods. It is therefore important to determine which methods are used and how they are parameterized. In principle, the transformation must be traceable. This also means that the “Human in the Loop” (HITL) must be a mandatory control instance at least at the beginning of such a process. At the same time, it must be checked that only the really necessary data is recorded and stored. Such a procedure places high demands on the implementation as well as on the necessary control system.

Healthcare: Body of knowledge and test catalog

In Switzerland, there is still no digitization standard for the healthcare sector. With CertiHealth^TM and thus offers providers and users a reliable and solid standard for the compliance of digitization in the healthcare sector.

In Switzerland, the Business Records Ordinance is the basic standard for the digitization of data and for ensuring its integrity and security. The basic principles laid down therein can also be applied to health data without further ado. Consequently, the audit catalog is based on these principles. In addition, there are specific requirements that can be derived from national and international laws and standards. Interestingly, Switzerland has no legal requirements for these processes. For this reason, we have supplemented our best practice catalog with controls from these sources:

• ePDG (Electronic Patient Record Act and Ordinances (ePDV); the focus is on security requirements.
• Data Protection Act (DSG), in particular the provisions on privacy by design and privacy by default; data protection impact assessment
• Medical Device Regulation (MepV) and EU standards[1]
• EU AI Act, in particular for the use of procedures with AI support
• TR Resiscan (BSI, DE)
• Semantics (optional): Consideration of ePD standards such as Snowmed CT (optional) or patient-focused procedures such as COBEDIAS®

[1] We do not test the medical devices themselves, but their use in a digitalization process, if applicable