MuleSoft and Tableau: Unlock and act on patient data
Connect disparate data sources and visualize data to make better decisions faster with Tableau using MuleSoft APIs.
In this demo, learn how one hospital combines public data sets and connections to their EHR systems to highlight and action important information. See how data is collected from theCovidProject, CDC, OpenDataDC, and the Provider’s own EMR, and then visualized in Tableau.
Pankaj Ahire (00:08):
I will demo this today with two personas, one as a CIO of Healthy Hospitals, and two, as an analyst who built all of this information. I'm the CIO of Healthy Hospitals. And this dashboard is the cornerstone of our preparedness, not only for the current pandemic, but also as a next step towards the future. We combined public data sets and connection to our electronic healthcare record systems to highlight and action information that's pertinent to Healthy Hospitals. MuleSoft allows this critical real-time healthcare integration and Tableau exposes the metrics through intuitive visualizations. Let me show you how, and let's walk through what we are seeing here. The metrics have been designed by Healthy Hospitals and implemented in a MuleSoft API, and then pulled into Tableau. On the left-hand side, we can see the number of positive cases diagnosed, divided by available hospital beds. This is a leading indicator because it anticipates the need for hospital beds.
Pankaj Ahire (01:20):
A large increase in positive cases likely indicates that there may be need for resources in the state, and here on top-right, we have a lagging indicator and a lagging indicator is those who are already hospitalized or where the number of available hospital beds. The color here is the metric, and this also shows who in the state were hospitalized and what's the increase in hospitalization over yesterday. Then on the lower left-hand side, we have the top 10 states with hospital beds and also the ICU beds. This is so I can anticipate the macro needs from my indicators and then move patients to those states that actually have resources. Finally, we have there's the comorbidities that data sets that's published from CDC. Interestingly, this comorbidities dataset also has ICD-10 codes, and I can use those ICD-10 codes to match Healthy Hospital patients that are at risk with those conditions.
Pankaj Ahire (02:27):
In one dashboard, Tableau allows me to deduce the macro trend for resource allocation as well as translating it to my actual patients. And then MuleSoft connects these disparate systems, calculates my business indicators, and exposes endpoints for actionable orchestration. Now, let's look at this from an analyst's perspective. I have to pull the data as it's currently available, I have to integrate public data sets, I have to calculate business metrics and predetermined actions all speedy with maximum efficiency by reuse and all that should be compliant with the enterprise standards. My first stop is Anypoint Exchange to see all the assets that I can reuse. I know that Healthy Hospitals uses OpenEMR. So I'm going to search for OpenEMR. It turns out, this is exactly the system API that I need to reuse. I also noticed MuleSoft has a MuleSoft Accelerator for Healthcare, in which there are plenty of assets that I can actually use as I build my integration and this new innovative, advanced analytics dashboard.
Pankaj Ahire (03:43):
So adopting design-first approach of MuleSoft, I come up with the following design. You can see over here, I'm reusing the OpenEMR system API. I'm also connecting to the various public data sets that I have from the COVID tracking project to CDCs comorbidities data and the open data that publishes the number of hospital beds in the US. I process them all in a process API, which not only integrates these data at different granularities, but also implements and keeps ready the actions that need to be taken. The correlation process API is then used by the experience layer at the very top, which publishes data for usage in Tableau. Anypoint Platform speeds me up immensely by promoting reuse at various different levels. I can not only reuse data types such as the message types that need to be sent to the EMR, but I can also reuse during implementation.
Pankaj Ahire (04:51):
Here you can see the implementation for the process API. Anypoint Platform is able to convert the system APIs into connectors themselves, which I can simply compose together for a very speedy and intuitive implementation. Finally, the platform also allows me to automatically adhere to standards compliance according to my organization by securing each one of these APIs automatically, and also a live look at my architecture with all of these APIs now running. I can even troubleshoot using visualizer and Anypoint Platform to automatically look and monitor my APIs. Switching back to Tableau, let's explore how this Tableau dashboard was built. You can see that this Tableau dashboard actually has all of these data sources as sheets. Each of these sheets presents a data set that was coming in from MuleSoft. Switching back to the persona of the CIO. Let's see how now we can finally action this.
Pankaj Ahire (06:03):
The CIO is actually able to select a comorbidity condition. For example, if I click on this and is able to pick an actionable item, that's already ready and waiting that we message the 100 patients that have circulatory diseases. And we can see here that the Tableau dashboard has automatically posted to the electronic health record system off at the hospitals. Mr. Phil Bedford seems to be one of the patient on home... We have already posted a patient reminder message. So we switch to the EMR system. Look at Mr. Phil Bedford click on as medical record. Here you go. We get to see this message in their medical record, asking them to check for their primary care provider for important information about the pandemic in their area.
Pankaj Ahire (06:56):
In this demo, we have only shown connecting to the EMR, but the possibilities with MuleSoft are endless. Tableau allows you to unlock the data and visualize a new innovative way. MuleSoft allows you to extend this same architecture that you have used for this pandemic to future pandemics, to future states such as connecting more systems like maybe Health Cloud or Salesforce, connecting even more systems, even more innovative advanced analytics use cases such as mentoring management, automatically anticipating in mentoring shortages. The possibilities are endless.