InetSoft Webinar: Hospital Human Capital Analytics Project

Below is the continuation of the transcript of a webinar hosted by InetSoft on the topic of Using Analytics to Increase Staffing Productivity and Improve Hospital Operations. The presenter is Abhishek Gupta, Chief Data Scientist at InetSoft.

The first step of this hospital human capital analytics project was to focus on data governance. The information is highly important. If the quality of data is not there, it will impact our project. Our data governance expert helped us to standardize order reports, scorecards and definitions.

We built in-house a data-dictionary tool that helped us to manage all the metrics that we created, and these metrics are ready to show on the reports that we push out. Each metric has a definition, a contact person for questions, the source owner, and also when the data is refreshed.

For data infrastructure, we have built a new data system. We integrate the data into the hospitals project throughout the years and so when the labor project comes, we just mainly reuse information we already built. Remember we use information as the key for our success.

ETL of the data is very complicated. For this particular project we are using Microsoft tools, the SQL Database, the Integrated Data, SSIS and .NET Framework. It is not the code that makes the process fast and slow. It is the implementation.

We have been constantly improving the speed of the process, we use boatload processing, and that is faster than row by row. We borrowed the concept from our Big Data system Vertica. We built it to be scalable so it has the ability to expand, to support increasing workloads and entities.

We improved our goals to optimize load time from a couple of hours, down to 10 minutes. This is something new that we learned from this project. Using technologies is just part of the work. Changing cultures and how you utilize data and how you can share those ideas to the users would be far more important than the technology.

Under product development, we take new ideas and apply years of knowledge and experience to turn it into reusable product. We usually anticipate what the customer wants. For this particular labor project, we received very little detail. What we did was instead was create a prototype and in six weeks. We went live with the prototype and the nurse manager and assistant nurse manager came back and gave us ideas of how to make it better.

We found that it's far quicker and more effective than asking for what they want before the project starts and wait for decisions. Based on the innovations of the team, we quickly build these prototypes. When the users look at it, they thought we finished a product, but in fact it is just the prototype. It helps the user make decisions quicker, because they enjoy seeing it.

View a 2-minute demonstration of InetSoft's easy, agile, and robust BI software.

The labor project already went through seven versions. The first time we went live with this project in September 2018. And today we already have seven versions. That's telling you we adapt to change. In two months we already changed, fixed and add new features. The last feature we would be push. This is for the manager who wants use a mobile device. They are out there on the run, taking care of the nurses and taking care of the patients. So push is what we would do next.

To deliver a wealth of information we use InetSoft and web applications. Let me explain how we do that. The traditional InetSoft software we use to display a beautiful report and analyze the data. The visualization is amazing. But we use an API technology that came with InetSoft and our developer built right on top of InetSoft to receive information feedback from the nurse manager and in-charge nurse six times a day.

The condition and the request from the beginning would be if you are over budget, you must type in a comment and explain how you went over budget for your staffing and how you plan for the next shift. So we use the API from InetSoft to collect this information.

Originally the data is available daily, weekly, biweekly and monthly. For this particular new project, management has asked us to provide data six times a day for six shifts. In addition the in-charge nurse will enter information when they went over staff, send feedback to their supervisor or to the assistant nurse manager and communicate how they can staff the next shift.