Data First!
Data, Big Data, Business Intelligence (BI) and Business Analytics (BA) seem to be the most spoken terms lately in the information technology (IT) world. To most businesses and organizations the analysis of data can drive customers to your brand name, helps you reduce errors or achieve some fiscal efficiency in your operations. In a forensic mental health organization data potentially means safety and life.
Over three years ago my data management office chief and I saw a demo about an organization and what they had accomplished with data. They had started by combining data from a workman's compensation system and their learning management system. The problem they were investigating was why were they having so many accidents with pallet loader drivers at specific locations? They found that the locations that had the lowest compliance with safety training also happened to have the highest number of accidents. This came from layering two disparate yet related sources of data. Shortly after they pushed out a directive for all employees to become 100% compliant with the safety training requirements. Within a short amount of time the data started to show all sites with much lower numbers of accidents, including the sites with high accident counts prior. This led them on the path of creating a data repository to be used as an enterprise risk management system. This was the proverbial light bulb for us. We needed and wanted the same exact thing to help our organization solve our problems.
My IT organization is constantly tasked with providing old and new sources of data that will help answer questions. The most important question we are asked is how do we make our hospitals safer? Every single technology that we implement gets us closer to finding the causes of violence and aggression. What I mean by getting us closer is each technology system deployed and used provides us another layer of data to analyze. The new data hopefully provides more opportunities for my analytics people to see if we can find the causes to our problems. It is a very simple approach. If you can find the causes then you can craft a corrective strategy. The ultimate goal is to apply a data informed corrective strategy and then have downstream data show that the strategy is effective and produces the desired outcomes.
Gartner Data Analytics Maturity Model:
- Descriptive Analytics: What happened?
- Diagnostic Analytics: Why did it happen?
- Predictive Analytics: What will happen?
- Prescriptive Analytics: How can we make it happen?
In Gartner's data analytics maturity model my organization's abilities are starting to move into number two: Diagnostic Analytics. We are starting to be able to see the "Why did it happen?" question answered. But we are only just at the start. We need much more operational data to be firmly capable in this area. For now it is only a dream to reach the number three slot: Predictive analytics. I can only imagine having the ability to be able to predict something happening before it actually does.
We are now for the first time able to see a problem showing up in our data and start to look for root causes. We have projects in the works now that will provide pharmacy prescribing data, employee scheduling data and our Personal Duress Alarm System data. Now imagine we see a sudden spike in assaults in a specific hospital unit. We will be able to look at what is being prescribed. Are there specific psychotropic drugs being prescribed that have aggressive behavior as a side effect or maybe we are running a shift with many vacancies and a lot of overtime so as a result our staff are tired and not as security aware as they would be under normal conditions. All of these data examples may be contributing factors to the problem. Remember, you can't manage what you don't measure. If you don’t have the data you don’t have anything to measure!
We take a data first approach in all of our projects. We embed staff from our data management office in each of our IT projects. For us it is not all about taking a paper based business process and moving the process into the electronic realm. It is all about what do we get out of the technology system in the end. Does the technology actually provide us with actionable information? In some cases we even have our data management office staff as project directors.
The key to data management is that there is no quick win or magic technology that helps you leap through the Gartner analytics maturity model to the number four slot. To build an effective data strategy and operation it takes:
- Smart and determined data people;
- Good data technologies;
- Hard work;
- Time;
- Good source data;
- Well thought out strategies;
- Incremental progress and success.
Lastly you have to have good sponsorship that is committed to having the ability to make informed decisions instead of half hearted guesses.
