Architecture of an Event Processing Application for Monitoring Cardiac Patient Wait Times

1. Introduction

Minimizing wait times and achieving more effective and efficient management of patient flow are important goals for any hospital. Managed process applications are used to improve operational efficiency in organizations, such as hospitals, by monitoring critical events within business processes and reporting on performance (Smith & Peter, 2003).

In this paper, we present an architecture for event processing applications that manage business processes and we use a case study of monitoring patient flow to evaluate our architecture and illustrate our approach. Event processing applications can collect streams of events from sensors for processing to infer critical medical events in real time (Yao, Chu, & Li, 2011). However, to manage business processes, it is critical to understand not only where in the hospital those events occur, but also where in a business process those events are occurring. Metrics, such as wait times, can be computed in real-time by using complex event processing to integrate and aggregate events in support of fine grained monitoring of business processes.

Our work seeks to improve cardiac patient wait times by integrating events from a wide spectrum of sources within a hospital using complex event processing (CEP) to support fine grained monitoring of patient flows and detailed analysis of bottlenecks. Potential sources of events within the hospital include medical equipment, physiological sensors, RFID tag readers, health information systems (HIS), and business process management (BPM) systems.

As an example, a cardiac patient could be equipped with a bracelet that contains a RFID tag to track where the patient is physically located, but location information alone is not enough to understand where the bottlenecks are in the cardiac patient process flow. For instance, is the cardiac patient in the emergency department (ED) waiting for an ECG or is he waiting to see an ED physician to review test results?

To solve this problem, an event processing application needs to correlate business process events with RFID and sensor events. Then complex events, like a patient waited one hour for an ED physician to do an initial consultation, can be used to calculate and display in a real-time dashboard the current average waiting times at key points in the cardiac process flow. This enables hospital staff to better identify and respond to bottlenecks.

A real time dashboard is an important tool in improving integration of care (Mouttham, Peyton, & Kuziemsky, 2011). However, most hospitals today, have a traditional HIS data architecture that does not include integration of BPM and CEP. In traditional HIS data architecture, data is extracted from disconnected data sources (sensor logs, departmental databases) transformed and loaded into a data warehouse view over a period of weeks to support historical reporting (Kimball & Ross, 2002). Incomplete and disconnected data often means that fine grained monitoring of patient flows is not possible. Only coarse-grained metrics to support high-level administration decisions are available which are insufficient for monitoring patient flows.

Integrating business process and location events with CEP supports real-time monitoring of patient flows that includes:

• •

  Monitor patient wait times such as wait time for bed, wait time for surgical procedure, and wait time for physician consult.

• •

  Monitor operations wait time for housekeeping and transport.

• •

  Monitor service times (e.g., how long it takes to get a medical imaging report).

• •

  Emit alerts in real time such as when wait times are above a threshold.

In this paper, we use a case study from a large community hospital in the suburbs of Toronto, Canada to illustrate the problem and describe an event processing application for monitoring cardiac patient wait times which includes:

• 1.

  A business process model of cardiac patient flow in order to understand where the patient could be waiting.

• 2.

  An event-driven data architecture for collecting and processing basic events from multiple sources in order to create complex events delivered to a real-time dashboard.

• 3.

  A detailed analysis of the complex event processing required to integrate and aggregate both RFID and business process events for fine-grained metrics related to wait times in cardiac patient flow.