BJC Healthcare leverages meshIQ to monitor important patient data transmissions

The Problem

For BJC, the monitoring system for the IBM WebSphere MQ environment was a home-grown application requiring components deployed on multiple servers. On top of that, the solution required a lot of maintenance and fine-tuning and was thus expensive to maintain, ultimately making replacement a necessity.

The business-critical applications at BJC capture data at both doctor’s office and hospital settings. This data includes patient visit details, dates the patient was admitted, transferred, or discharged, lab results, vital signs, radiology reports and medication. Having this data available 24/7 and keeping the systems running smoothly to handle the large volumes of timely data are crucial to daily work. For its part, WebSphere MQ functions as the communication middleware that moves these data from the source to their processes, using individual queues along the way.

For instance, a doctor might dictate a letter through a specific transcription system. The letter is then transmitted using HL7, a standard for interoperability of health information technology through the electronic interchange of clinical, financial and administrative information among healthcare-oriented computer systems. WebSphere MQ is responsible for “store and forward” services, that is, highly reliable delivery of the data put in the queue by an application. An application converts the data from HL7 to XML and moves it to a report queue where the essential formatting and required tag elements are validated. The application will then look up the patient in the system based on the patient’s unique identification number. The next queue brings the message into the database, an outbound queue or an error queue (if it fails).

The same process is applied to lab results and pharmacy orders, both of which are extremely time-sensitive and, of course, patient-critical. Other time-sensitive information that isn’t transmitted can immediately elevate risk factors for patients.

Recognizing the need for a more robust monitoring platform happened to coincide with the system’s migration to clustering. A stronger and more reliable monitoring solution would be especially useful for error queues and queues that had an unusually large number of messages on them. But where would BJC be able to find an application performance management solution strong enough for such a complex data system?