The board and members of the Royal Institute of Engineers (KIVI) analyse current political developments during the corona crisis. We provide facts and interpretation from the technological knowledge and experience of engineers.

The Parliamentary letter regarding the national introduction of the CoronaMelder App (App) of 16 July 2020 prompts the questions and comments below.

After criticism during the start of development, our expert team is now impressed with the development of the App. The market consultation (ref. page 2) in April played apart the expertise in the Netherlands and caused more than a month's delay. After that, the knowledge and expertise in the Netherlands was much better pooled and deployed. Where necessary, use was made of the right experts and facilities, which the Netherlands has in abundance. The focus was strongly on the ICT open source community, which may have resulted in a number of other technological fields being underexposed. However, the letter gives the impression that very many aspects were taken into account, including technology and privacy, among others. The main conditions for successful launch and deployment now seem to have been met.

KIVI is definitely in favour of the introduction of this App. It can give citizens a serious notification that they are at risk of being infected with the coronavirus several days earlier than is currently possible. Especially at the stage when someone does not yet have symptoms but may be infectious. This can limit the spread of the virus. Moreover, it is a much more modern and quicker form of source and contact investigation. A good addition to or even replacement of the manual and telephone methodology used by GGDs. Especially now that the source and contact investigation organisations of several GGDs are overloaded.

In our experience, digital means appeal to young people much more than a phone call to the GGD. They register a report on the phone, but they do not answer a call from an anonymous number (GGD). Young people only become cautious when the risk becomes very close. The app can show the danger clearly.

Our expert team has only a few questions and comments in response to the parliamentary letter:

  1. According to the letter (page 5), measurements have shown that in 27% of the cases where there was contact, the App does not record this(false negative) and also in 27% that there was no sufficiently prolonged contact still records a contact(false positive).These error rates are very high in our opinion and should improve. After all, at the desired acceptance rate of 60%, a maximum of 42% of high-risk contacts are then reported
  2. According to the letter (page 14), the App registers a contact if two phones have been in close proximity for 15 minutes or more. This contrasts with the rules for the 1.5-metre society, which, after all, assumes that even short passages on the street already pose a risk. Can this difference be explained from a virological or epidemiological point of view?
  3. The retention period of contact recordings on the phones and the "backend" is 14 days states page 14 of the letter. However, it can take up to 12 days for someone to develop symptoms(source RIVM). After that, one has to decide to get tested, make an appointment for a test and then it takes another few days before the results are announced. Finally, the sick person then has to upload the notification to the server. This can take a total of (many) more than 14 days. Crucial contact data is then already deleted. It therefore seems sensible to increase the retention period to 3 weeks. Especially if a change at a later stage requires first amending the DPIA and other documents and obtaining the necessary approvals
  4. It makes sense if the App indicates when someone may become infectious and develop symptoms. This should, at a minimum, indicate when a test and self-isolation is useful. Does the action advice mentioned in the letter on page 13 include these elements
  5. No effectiveness analysis is mentioned in the tests and checks, which, according to page 12 of the letter, are still being carried out. However, it is important to know how many persons who received a contact notification actually get COVID-19 and how many persons get the disease but have not been notified. From these and other effectiveness analyses, possible improvements to the App can be derived. Can the minister conduct a meaningful effectiveness analysis in the near future? This should include a comparison with the results of traditional source and contact surveys and with foreign apps
  6. One of the technology and privacy experts on our team analysed the minister's Data Privacy Impact Assessment (DPIA), referred to in the letter on page 3. You can find his report here. Among other things, it lacks a clearly articulated objective, a legal basis and understanding for users of the privacy risks, their interests, rights and obligations. It is important to adjust the App and DPIA based on these findings
  7. The European initiative to connect the apps of different countries is very important to increase support for travellers and for residents of border regions. It is not clear from the letter on page 15 in which countries the App works or is expected to work.
  8. Foreigners who are in the Netherlands can also install the App and get a notification. But do foreigners in the Netherlands know the way? Tourists and foreign students with symptoms would also have to go to the test lane in case of complaints. The App should come with clear instructions on how to request a test? Even if someone does not have a Dutch BSN?
  9. Making the App mandatory is not an option in almost any case is stated on page 17 of the letter. However, a high acceptance rate is desirable to achieve the intended effect. A bonus might help increase acceptance. A positive incentive works better than coercion and punishment. Consider a special queue or discount at events. An exemption to leave contact details when visiting the hospitality industry could also be considered.

In making our analyses, getting factual information did not prove easy. Although the letter on pages 2 and 3 indicates that maximum transparency is the starting point, those involved in the development by the ministry have to sign a confidentiality agreement. This hinders information sharing.

_____________________________
The Hague, 8 August 2020

Do you have any further questions? If so, please contact us at E: jan.wind@kivi.nl
Photo: Pixabay free to use

web stats