Complex Event Processing Engine based Study Enrollment Alerts

Research IT offers the Stanford Medicine research community the ability to configure real-time alerts based on specific clinical events at both SHC and SCH. Each alert is custom-designed to notify research(s) when certain criteria are met, for example when a study eligibility event occurs. In one instance, an algorithm monitoring blood pressure and lactate levels in real time is successfully being used to enroll patients in a sepsis study. When pre-defined levels or trends are detected, study coordinators are paged and can immediately verify if a patient is eligible for recruitment. If notifications are not time-sensitive they can be batched for periodic delivery.

How it works

A crucially important aspect of the ongoing collaboration between the hospitals and School of Medicine is the live HL7 message feed sent to the Research IT. This HL7 message feed originates from various core clinical ancillary systems such as pharmacy, admitting, lab, billing and dictation services. Lab and pharmacy orders, lab results, patient admits transfers and discharges and many other essential informational messages are sent in real-time to the clinical data warehouse and can be used by the our clinical alerting engine

The alerting engine can be configured to watch for single sentinel events or occurrences within a specified time frame of one or more related events. Notifications are sent via e-mail, which can be either regular e-mail with anonymized content, or secure e-mail with fully identified content. E-mail to pagers is also an option, though these must be anonymized due to the use of radio transmissions in triggering pagers.

HL7 messages are also stored nightly in the clinical data warehouse, along with supporting data from both Cerner and Epic, in support of daily, weekly or monthly alerts. These periodic alerts can be based on a broader informational base than what is available to the real-time alerting machinery and are well suited to such purposes as periodic review of clinical data as part of an ongoing retrospective study.