(without proper algorithms and visualizations!)

Everyone is saying how valuable data is these days. And they are correct.  AND they are wrong. In fact, data by itself, is almost worthless.  It is just a collection of zeros and ones. It means very little.  It is not until you start to do things with it, apply algorithms, visualize it, etc. that is starts to provide meaning and value. At this point it starts to tell you its stories…

Data by itself is not useful. It is just Zeros and Ones,

But wrap that data up in a ‘story’ and it becomes information!

Good information leads to understanding,

Understanding is the foundation of knowledge,

Knowledge is a driver for Change!

 And it all begins with the stories your data can tell you!

Stories are actually what we use to derive meaning from our data. And it is this meaning that provides value. The stories are pulled from the data using various algorithms and visualizations. They allow us to make sense of all the Zeros and Ones – and present the data in a form that brings meaning to us. In most cases they are graphs and charts that show us visually some pattern in the data.

‘Raw’ Data is actually quite boring for the most part. Think of an excel spreadsheet – just a bunch of rows and columns. This might be something like what your data looks like.

While this has use and meaning to someone like myself who deals with raw data day in and day out – it does not have real meaning to others and does not provide much in the way of information without a lot of study. So – as a generalization – this data in its current form does not have a lot of value to the average person or organization by itself.  You can look at the above grid and see the list of allergies for certain patients. But there are over 11,000 records in this dataset (and it contains all the medications, allergies, problems and immunizations for the patient population – but you cannot tell that without a lot more investigation). You cannot easily see any of the patterns or the stories it contains using just the raw data (i.e. the grid). As well, trying to find all the patients who might have a condition or allergy means sifting through all the records. Simply trying to find all the patients with diabetes by searching for all the patients with an A1C above 7.0 would take forever. Now imagine trying to get a better understanding of a number of different conditions and how they have affected your patient population. It would be almost impossible using something like the above ‘raw’ data.

This is where visualization and algorithms come into play. They allow you to better understand your data – they provide meaningful interpretations – and allow you to make decisions based on the stories your data can tell you. Algorithms can go beyond just giving you a list of the patients with say diabetes, to allowing you to find patients who are in danger of developing diabetes and thus taking action to prevent them from developing the condition. Finding patients with a particular condition can be easy if you have good data – but you need an algorithm to tease more subtle meaning from your data…

Now imagine the above ‘raw’ data using visualizations. You might have something like the following:

These simple example visualizations show various breakdowns of the data (by decade, sex, condition type, etc.). They provide you some quick summaries of your data – and provide some basic understandings that are not possible with the ‘raw’ data. The above examples show you very quickly how your patient conditions break down by sex, patient decade, etc.  They also show relative importance/magnitude visually. Take the pie in the upper left hand corner. From this you can very quickly tell that patients in their 60’s and 70’s make up over 40% of all conditions. How might this sort of information change how you treat your patient population? You could not do this using the ‘raw’ data in the grid.

The algorithms and visualizations allow you to start making use of your data, your most valuable asset. It is one thing to have data to care for an individual patient – but what if you could use that data to help improve the care and the health of your entire patient population? You can do this in your own practice, or participate in projects where anonymized and summarized data is used in research, etc. Perhaps you may want to compare your care against your entire Family Health Team or Primary Care Network, or even from entire health regions, or specialty! And this is just the beginning of being able to put you data to work for you!  Visualizations, such as dashboards, allow you to see your data in different ways, interact with it, and see patterns that are not obvious from dealing with it on a patient by patient basis. They allow you to explore your data and discover new meaning that is hidden in all the details. This is the beginning of the real value of your data!!

The funny thing is that while data by itself is not worth much to the average user – but start looking at it, using visualization tools and special algorithms, and you can produce better understandings and therefore real value. The trick is finding the right visualizations and algorithms that make sense to you. It is these algorithms and visualizations that extract the value out of your data and turn it from being worthless to almost priceless!

There is a lot we can learn from the data currently locked in our medical systems. I have heard the phrase ‘Algorithms are the new Doctors, and Data is the new Drug’. While I understand what they are trying to say – I am not sure I totally agree with this philosophy – but I do know that data and algorithms will play a huge part in improving patient care and disease outcomes over the next few years. The question is how can you use your data to do this?