EHIC2QR - Automatische Patientendatenerfassung

By Harder, Jan

EHIC2QR - Automatische Patientendatenerfassung

The time needed to report test results to patients potentially being infected with COVID-19 can be decreased from 7 days to 24 h with the help of our smartphone app.

E2QR_Logo

Inspiration

Microbiological diagnostics are essential for the outpatient care of patients with suspected COVID-19. The examination for SARS-CoV-2 can currently only be carried out safely in a laboratory for medical microbiology using PCR. For this purpose, a data transfer to the laboratory is necessary to be able to assign the test results to the respective patients and to inform them promptly. Also, a possible test that can be performed on the patient (POCT, e.g. lateral flow assay for antigen detection) will not have the same significance as a PCR (see rapid influenza test). The above-mentioned problem therefore remains. An antigen test with sufficient sensitivity and specificity for mass testing is not yet available in sufficient quantities. Currently, the patient data are entered manually on a sample form (“referral slip”). Above all, poor legibility of telephone numbers, names, and addresses is a major problem that occurs in about 10% of all cases. Furthermore, this leads to the fact that test results can only be communicated with a long delay and a very high workload. Likewise, many working hours are required to transfer the patient data to the Laboratory Information System (LIS) (approx. 2.5 person minutes per submission).

What it does

To achieve the goal of minimizing the time needed to label the test samples with a patient’s information, one needs to simplify the process from recording the needed data to printing out the etiquette. This is being implemented by automatically extracting a person’s information on one’s health insurance card (EHIC; European Health Insurance Card). Also, the doctor’s master data is saved within the app once entered.
First of all, a photo of the patient’s EHIC is taken on a smartphone device within our app to extract the needed information for the auto-completion of the patient’s data. In the end, the set of data will be printed out in form of four self-adhesive labels: One of them is provided as a QR code (different forms of syntaxes are supported) while the remaining three consist of Latin letters.
In the following description of all particular steps within the process will be given:

After a doctor has saved his data consisting of his LANr (“Lebenslange Arztnummer”) and BSTNr (“Betriebsstättennummer”), a user can start the process of scanning the EHIC card by taking a photo. The EHIC holds information concerning the first and last name, the birth date, the personal insurance ID as well as the insurance company’s ID and name. Furthermore, one needs to record the patient’s data i.e. the related phone number and address. Last but not least, the date of the examination will be recorded automatically. To verify a patient’s phone number which is manually typed in, a doctor can conduct a test call within the app. In case there is any information that is missing (especially the phone number), the app will display a warning message. After verifying the collected information, the doctor can proceed by printing out the labels. Suitable printer hardware can be connected to the smartphone via Bluetooth connection.

How we built it

Concept The following information can be read from the EHIC card using text recognition: Last name, first name, date of birth, insurance number, identification, and name of the insurance company. Furthermore, the following information is necessary (for correct reporting to the health authority or communication of findings): Telephone number and address of the patient, identification of the patient (in Germany in the form of the lifelong physician number (LANr)), as well as the date the sample was taken. The telephone number must be entered as free text and should currently be checked manually. Missing data can be avoided utilizing a warning notice (so far in >1% of cases). The date is known to the device via the network time. The doctor’s master data only has to be stored once in the app. For the German (region Germany) version, this includes Lifelong physician number (LANr), business premises number (BSTNr), name, and contact details (preferably telephone number). So far the data is translated according to the LIS syntax (established at UKE Hamburg for Swisslab). Alternatively, we will offer other syntaxes as hl7 or ELGA message into a QR code. A lot of different modes should be supported to be open for all LIS. A separate editor is planned for this. Existing modes are also to be passed on to other devices via QR code. The QR code has sufficient data density to display the whole data record on two by two centimeters, even with a moderate resolution of the printer. With the help of the data scanned and typed into the prior forms, we can create a QR code as well as additional labels. The syntax of the data set can be preset in the app. Because of that, only a minimal amount of communication between the laboratory and the sender is necessary. Finally, the patient’s record in the form of a QR code and three self-adhesive labels can be printed out (with a label for sample labeling and one for the patient, four labels in total). Suitable printers are available, for example, from Honeywell or Zebra for mobile use. A connection is made via Bluetooth and is possible under Android X and iOS 10 and above. The test findings can be communicated via another app (or mail or SMS).

Technical implementation The application is using react-native as the foundation. React-Native is a javascript framework for creating native apps without having multiple codebases. Hence we can provide the service to iOS and Android devices with only one piece of code. Along with react-native we are using the following libraries

  • react-native-community/camera roll
  • react-native-firebase/app
  • react-native-firebase/ml-vision
  • react-navigation
  • moment
  • native-base

FAQ

- Is the doctor’s data save?: The doctor’s data is only stored within the app of the user. We export no data via the internet or similar networks.

- Is the patient’s data save?: All data (even the text recognition) is processed locally on our phone. We transmit no data via the internet or similar networks; only Bluetooth is used for printing the labels. All patients data will be deleted after the printing process.

- My laboratory is not mentioned in the list.: If the lab you like to send your samples is not included within the list, we probably do not support a suitable data format. If you are in Germany or Austria you can choose the standard form (“KVB Standard Deutschland“ or „ELGA Standard Österreich“). For further questions please refer to your developer team via mail.

- We are a medical laboratory and like to read your code.: We can support many kinds of QR-Codes. If one wants to, one can create own syntax via the editor or speak to our developer team via e-mail for further support.

Challenges we ran into

  • Getting the app to run on Android
  • Find adequate printer to test the app
  • Find syntax for data transmission to more than one lab

Accomplishments that we’re proud of

  • Participating in an EU-wide hackathon helping to fight COVID-19
  • Help doctors save great amounts of time and making more tests possible in less time
  • Contributing to slowing down the spreading of the COVID-19 by accelerating the tests of patients
  • Successfully implement offline OCR
  • Successfully get the app to work within just two days
  • Working together in a large team in which communication is working very well

What we learned

  • Learning methods of software engineering and project management
  • Hackathons need a lot more organizational work than expected prior
  • The expertise within a particular domain in which the solution is implemented for is essential for the success of one’s business model/app idea (especially in terms of UX / UI design)
  • Expert advice as to what project would help fight COVID-19 is essential
  • Micromanagement and project management
  • Developing a suitable plan for what is next for the project after the hackathon
  • a clear description of the use-case is needed in advance of the hackathon to start with the development

Video

Team

Tom Albrecht, Frederik Maenner, Johnvir Singh Khattar, Martin Tran, Toan Nguyen, Jan Harder