End-to-End Monitoring of Critical Applications

By Pedro Pacheco

Thanks to an early season skiing accident, I found myself sitting in the Emergency Department of a small hospital in a Colorado ski town. The hospital had recently gone through a major renovation and was proud of its investment in HIT (Health care IT). Most patient workflows are now automated.

Check-in and registration were a breeze. Within a few minutes, my patient record was updated and my case was well documented. With an electronic signature of consent and a scan of my insurance card, all I needed to do was wait for my turn.  A large monitor displayed my name as well as the names of others waiting to be seen, as well as the likely waiting time.

Being a software person with a passion for HIT, I was beside myself. A small hospital in a mostly rural area embraced technology to facilitate patient care and expedite service. This is awesome!

I was eventually taken into an ED bay, where I became even more excited when I noticed the attending physician was carrying a tablet (as opposed to a clunky computer-on-wheels or a clipboard). She scans my wristband, and nothing comes up.

After a few more attempts and some not so pleasant comments about technology, the physician gives up on using the tablet as a well-intentioned IT person is running around telling everyone the “system” is down.

My shoulder, assumed dislocated, is now really painful, and the drugs I was given to manage the pain are slowly taking effect. I am hard to understand on a good day; add the shoulder pain and the effects of medicine; the physician is now struggling to get her head around my case. Especially since all documentation about my accident and me as a patient are now lost.

More and more patients are coming in (mostly from Texas ☺), some with severe and life-threatening conditions. Health care professionals and resources are backing up.  Quality of care and patient safety are now been affected. Lives are now on the line. And, in the middle of it all, there is a very-sophisticated-yet–non-functional integrated EHR solution.

“Things have gone downhill quickly,” I think to myself. Chances are the problem is simple to fix. If only the poor IT person had the right tool, this whole mess would be quickly solved.

Eventually, with a great deal of professionalism and great bedside manners, my case is processed. As I am discharged from the emergency department with a separated shoulder, rather than pay my copay and receive discharge instructions, I am given a hand-written prescription and a note saying the billing department will be in touch to collect the service fees.

Did I mention that the ER was now full of waiting patients?

In a highly integrated patient care environment, one simple failure can lead to an end-to-end collapse; hence the need for an end-to-end monitoring tool.

In other words, when patient safety is involved and critical applications work in unity to support patient care, a robust, end-to-end monitoring solution is paramount.

It's likely that the otherwise well-done mobile patient registration application had some level of monitoring. But this is not nearly enough.

Mobile applications are about end-user engagement. As such, only a portion of the application lives on the mobile device. While mobile applications are difficult to develop and maintain, they represent only the tip of the iceberg.

Behind the scenes, in the dark corners of a datacenter somewhere, live tens, if not hundreds, of back-end system responsible for handling requests coming from the mobile devices as well as other crucial business processes.

With that in mind, a more robust monitoring pattern for a highly sensitive and complex application environment, such as an Emergency Department, is to monitor not only the app as it executes on a customer’s (or patient’s) mobile device, but also the back-end application infrastructure; arguably, more complex tasks.

Needless to say, it is also fundamental to correlate the end-user experience from the mobile device, across the network or internet and into such back-end services.

This is the only way to understand the true end-to-end application performance and ensure a good end-user experience. It also gives the administrators a full end-to-end picture of the application, allowing them to quickly and precisely identify and address faulty components. The end result is quick restoration of service.

There are many monitoring solutions in the market that only provide visibility on the mobile application (aka the tip of the iceberg). The 80-20 rule is applicable here: by monitoring only the mobile application only 20% (if that) of the entire application ecosystem is visible. Conversely, monitoring only the back-end services excludes an al-too-important component of any business process – “the user.” Last, monitoring both sides separately with independent solutions misses the correlation and continuity between user, device, network and back-end services.

AppDynamics is a solution that can monitor these business transactions beginning in the mobile device, into the services tier to the application, through message queues, into legacy systems, databases and eventually, into the termination point while maintaining the continuity of the business context of each transaction.

Anything else is purely superficial.

See how AppDynamics can help your HIT department, download a FREE trial today!

The post End-to-End Monitoring of Critical Applications: A Health Care Example appeared first on Application Performance Monitoring Blog from AppDynamics.