RIVM Crown Dashboard (2023)

As a data partner, we often work with dashboards. We work with clients to determine which numbers matter, create connections to unlock data from disparate systems, work on visualizations, and use dashboards as input to our analytics efforts.

How to create a dashboard?

Creating a dashboard is easy. Creating dashboards that will actually be used is much more difficult. We've seen many cockpits that we built unfortunately didn't get reviewed so soon after they were implemented. There are many companies that stopped updating/sending the dashboard for a while as an experiment, just to make sure no one missed it.

You can read how to create a dashboard here.In this article, we use our current knowledge of the upcoming coronavirus dashboard to analyze a real-world example.

Strategies, KPIs and Dashboards

Helpful dashboards contain policy-related information. Graphics and illustrations are chosen for a reason. they areKey Performance Indicators (KPIs), or at least Performance Indicators (PIs), which alone or in combination with other numbers show actions to be taken. This means that before you start creating a dashboard, you need to consider what data is important to your situation, and what the actual purpose of the dashboard itself is.

Basic Guide to Crown Control Panel

Here's the strategy behind the crown board:

"The outbreak of the virus must be contained as much as possible. The most vulnerable people in society must be protected as much as possible. Healthcare in the Netherlands should not be overloaded. At the same time, we want society to do this as much as possible. For the economy and for all The physical and mental health of the Dutch. This means that the government wants to relax measures as much as possible, but only if it is really possible. If it seems that we as a country are taking (or have taken) too many risks, any relaxation measures may need to be reversed .”

The purpose of the dashboard is to show when further policy easing can be applied, when all measures should remain in place, and when a reversal may be required.

The dashboard is only being updated now because of the coronavirus

The most important step in making a dashboard usable is selecting the KPIs and PIs that really matter. The accuracy and precision of these figures must be considered. In this case (a form that requires action) you need to carefully consider the delay of the shape to be able to react quickly.

good foundation

Dashboards are visualizations of data.If this data is unreliable, and neither is the information you get from the control panel.

To some of you who regularly deal with numbers, the coronavirus panel announcement might feel a bit late. Wouldn't that provide valuable information faster? However, we must not forget that there are many challenges in combining and automatically obtaining the correct data. The crown panels are probably only available now because it took so long to get the bases right.

correct number

Evidence relevant to assessing whether the virus remains under control has been carefully analyzed. This ensures that the dashboard actually guides you to the correct action.

Synchronize

Another factor besides time is that the use of dashboards is only now becoming really relevant. In the past, defining measures was relatively easy. The virus is spreading fast and health services are under pressure. Therefore, we must do everything we can to limit transmission. The severity of the epidemic is such that economic and other interests are clearly subordinated to public health interests. Therefore, the action is clear, and any number on the board will not affect them.

other interests

Now that the government has indicated that the virus outbreak is under control, some measures have been relaxed. Other interests come into play. However, it is important to know as much as possible whether the spread of the virus is under control. This way, action can be taken should the situation deteriorate again.

Regional Differences

Regional differences also become significant. If the virus starts spreading again in one area, it may not be necessary to immediately pivot to the entire country.

Notable personalities from Crown Communications

What will actually appear on the crown board? The government has given some examples and we will present the known figures. We assess the data quality of each project. For the current dashboard, it boils down to answering these 3 questions:

• How important numbers are to strategy
• How much delay is planned?
• How precise and/or accurate is this number?

death toll

The ultimate goal of the coronavirus strategy is to prevent people from dying. Therefore, he is a very important figure. It also doesn't mean it's the best number for the board. Because preventing people from dying is achieved by keeping the spread of the virus to a minimum.

When someone dies, it usually means they had the virus weeks before. So there's a big lag in that number. Restoring relaxation according to this number means that the movement is actually performed slowly.

Also, we might doubt the accuracy and precision of this number. RIVM only counts people who test positive, so it's not accurate because not everyone gets tested. It's also inaccurate because the number of tests is increasing every day, so it's not fair to just look at the trend of this number. This figure will be more accurate if all the complainants are checked in June.

If you look at the excess death rate calculated by Statistics Netherlands, the figure is probably a bit more accurate, maybe even more accurate (depending on how stable the death rate trend has been in recent years). However, the delay will be longer since it is calculated on a weekly basis.

Conclusion: This Council should lead to relaxation of measures or reversal of previous relaxations. While this number is the most important for the crown strategy, it is not the most relevant for the dashboard. It will fall on the board, but it will not be used for reverse relaxation activities. If you need it for that, it's actually too late, and there's a good chance we'll be back to "smart locks" again. However, this can be taken into account when deciding to relax further.

Number of people hospitalized

Hospitalizations are a number that speaks volumes about the current stress on healthcare, so of course it matters. Since almost everyone admitted to the hospital with suspected coronavirus has been tested, we can assume this number is accurate. It is therefore understandable that the RIVM often uses this number to inform the state about dissemination.

The problem with the dashboard is that even people who are hospitalized may have been infected at least two weeks ago. Therefore, this figure is too late to reverse the easing measures in time.

Conclusions: Admissions are accurate and precise, however with a shorter delay than deaths. So that number applies to the dashboard, which, again, is important for the further easing of coronavirus measures.

number of positive tests

The number of people who have tested positive is less than the number of deaths and hospital admissions. And it's not accurate at all since only a few with complaints were checked. It's also not accurate, as the number of tests per day has often fluctuated in recent times. That number will become more accurate as everyone with symptoms is tested.

Bottom line: The number of positive tests is an important number on the board, especially if everyone tested has symptoms.

Positive detection rate

The percentage of positive tests performed is delayed by the same amount as the number of positive tests. Accuracy also depends on the number of people tested, but this number is more accurate. It is less dependent on the total number of tests (but partly on low test rates).

Therefore, this number shows a slightly better trend in the spread of the virus than the total number of positive tests. If this rate is low, we can also conclude that we see positive tests reflecting a higher true infection rate.

Conclusion: The percentage of positive tests is also important to the board. This number shows the reliability of the number of positive tests.

Number of coronavirus patients in intensive care units

The number of coronavirus patients in the ICU is an important indicator of medical stress. It is accurate and precise. The lag time is relatively long, but it at least shows whether there is potential growth in this period. Therefore, it can be used to decide whether you can again take more risks without imposing an immediate burden on your healthcare. Latency must be considered. If that number starts to rise, it will be too late to use easing reversals.

Bottom line: The number of coronavirus patients in the ICU will definitely be on the board, and hopefully used primarily as input for further relaxation efforts, not for reversing it.

R0

The transmission rate (R0) is the average number of people an infected person infects. This is probably the most important number because it indicates whether the virus is spreading further or slowly disappearing.

If R0 is greater than 1, the total number of infections increases. If it is lower than 1, the total number of infections will decrease. It is based on positive testing and possibly an estimate based on other numbers listed above. That doesn't make it the most accurate or precise design. There is also a delay here.

Bottom line: As long as we are reasonably certain the number is below 1, there is room to relax coronavirus measures. When we suspect it is greater than 1, we need to reverse the relaxation. As such, this figure occupies a prominent place on the chessboard.

congestion rate

The numbers above are the ones we encounter most often in communications about the coronavirus. Those are the most significant numbers, but they also include some delays in getting information about the spread of the virus.

So, in addition to these numbers, the RIVM will look for indicators that can provide faster information. Especially at a time when the risk of virus spread appears to be increasing, as is the likelihood that mitigation measures will be overturned.

Some things have been discussed in the news and we can also speculate. Which elements provide relevant information more quickly than those already listed? Since transmission occurs primarily when people are in close proximity, we should pay special attention to signs that indicate crowding. These numbers are generally less accurate and precise, but they also have much lower latency, so are still useful for this dashboard.

How to use data analysis to create a crowd map

It is technically possible to link data from telcos (we won't discuss privacy issues in this article) to see where multiple phones are located near each other. Areas that are overly congested can take steps to avoid such congestion in the future.

We can also requestsocial listeningFor example, look up crowd feeds on Twitter. If these messages refer to cities, or even specific streets or squares, that may be grounds for action in specific safe areas.

Now that the food service industry has reopened, you can even search for news about specific restaurants and/or cafes. That way, companies that took too much risk can fail again, rather than the entire industry potentially falling victim.

Interestingly, the government implemented social listening anyway. For example, there is a bigger picture of the public's response to the coronavirus pandemic.Read what it looks like in practice here.

Conclusion: Dashboards alone don't get us where we need to be. Keep score!

In addition to the use of telecom data and the application of social listening, countless examples can still be found. It is important to properly evaluate why you chose a particular data source and a particular number:

• Is it important to the core strategy?
• is it accurate
• is it accurate
• How much delay is planned?

But most importantly: how would changing this number (possibly in combination with other numbers) lead to action?

So RIVM seems to have chosen a good approach for this board. The strategy is clear, the dashboards have clear goals, the KPIs and other PIs are defined, and the actions to take when the numbers change are (hopefully) clear.

Perhaps goals have been set to determine exactly when action is required. This dashboard has a solid foundation, it's time to start visualizing it properly.

This makes it a great example of how to approach this type of process. This way you can create a dashboard that really works for you.

Are you inspired and want to identify the added value a dashboard can bring to your organization?Please contact us.

Get data insights, use cases, and context once a month?

Join our email list and stay updated: