The Race to Analytics: Is Medical Device Data Finally in the Race

By Jeanne Phillips, editor 

Today, as digital data has become pervasive throughout the healthcare system, it has streamlined the ability to perform sophisticated information analyses to achieve improvements in numerous healthcare areas, from clinical to operational. At the HIMSS conference, the term has been on everyone’s lips for some time. Yet because too often medical device data is not standardized and often not even stored, the value of much of this important information is lost. From early warning scoring systems that can predict serious patient decline to better management of alarms, this device analytics can have a powerful impact. Finally, a look at the technology and topics presented at HIMSS this year suggests this situation is about to change.

Analytics: An Overview

Healthcare analytics typically involves statistical and qualitative analysis of data to examine or track an area in depth as well as explanatory and predictive modeling to suggest ways of solving problems and achieving goals. Analytics typically involve running historical and current data through sophisticated computer software algorithms developed to meet a particular goal. The resulting analysis may provide clinicians with greater clinical confidence through treatment decision support based on data from similar patients, or it may identify entire populations at risk for certain medical conditions that may be addressed through preventative tactics and education. Analytics applications can also help boost a hospital’s administrative, operational and staffing efficiency, as well as management and maintenance of equipment. They can help with ongoing hospital financial planning, justifying key purchases and insurance claim processing, while driving down costs.

With sufficient data, benchmarks can be set for optimal performance based on a medical facility’s specific profile. In fact, the Gartner Group, a leading IT research and advisory company, cited analytics as a crucial step today in elevating the overall efficiency of hospitals.

What will analytics mean in the future? As analytic applications and medical technology, especially medical devices, grow more robust, they are expected to play an increasingly important role in patient management, in particular. For example, a clinical data analytics system may flag a patient at risk for a a cardiac event or sepsis based on a pattern of medical device readings. These analytics interface with various alarm programs that will notify appropriate medical personnel to about the need for intervention. They also will tie into decision support systems that suggest appropriate treatments formulated from a vast pool of patients with similar condition who have been similarly treated in the past.

Devices used in the home, nursing home and other settings beyond the hospital also will interface with with systems that send out similar alerts–for example, summoning an ambulance when a patient meets certain high-risk clinical criteria that have statistically show to correlate with rapid decline in condition.

Analytic Data Types

Healthcare analytics typically take advantage of four types of data:
• Descriptive, which chronicles an event. For example, a patient had a seizure.
• Diagnostic, which focuses on why the event happened. For example, the seizure was due to a drop in blood pressure.
• Predictive, which looks at what typically happens next for these patients. For example, patients who have one seizure due to low blood pressure often continue to experience seizures if no intervention occurs.
• Prescriptive, which examines what has worked to address the problem in the past. For example, most patients who experience seizures of this type respond will to NAME drug dosing.

The Unique Nature of Medical Device Data

Medical device data is unique in the healthcare system because it provides ongoing, up-to-the-minute information, delivering crucial insights into the patient’s immediate condition. Information from bedside monitoring devices often drives decision-making that has an immediate and important impact on patient care. Therefore, ensuring that devices are always fully operational and support accurate and complete data capture is crucial.

For therapeutic bedside devices, such as ventilators and infusion pumps, appropriate settings and proper functioning are also important. An incorrect setting or a malfunctioning device can mean the difference between a patient life and death.

Hospital administrators and nursing staff also rely on device data to place patients in the proper care area and to help balance nursing workloads.
In addition, hospitals also own and manage large numbers of bedside devices compared, for example, to MRI or CT equipment. Making the most of medical device inventory is important in today’s cost-driven healthcare age. This can present hospitals with significant operational and budgetary challenges.

The Evolution of Device Data Analytics

Device data analytics can play a crucial role in enhancing patient care and addressing many medical device issues. As medical device integration systems have begun to evolve into more robust medical device information systems, only recently has data been appropriately captured in a way that supports analysis, helping hospitals to make the most of this valuable information.

This is because traditional device integration systems function more as a pipeline simply transmitting data to an electronic medical record system (EHR) or other downstream program, for storage and use by that system only. By contrast, a medical device information system is a true system with many moving parts serving many purposes. Before transmission to an endpoint system, it aggregates and standardizes device data in a separate archive where it can be accessed by a wide range of analytics and other programs in near real time. With analysis of up-to-the-minute information, these systems offer the benefit of addressing potential problems before they actually occur. However, analytics applications cannot run on any medical device data unless it is standardized and stored in its own system as is the case with a device information system.

The Power of Medical Device Analytics

Today, emerging device analytic applications are beginning to play a key role in boosting the value of medical device data in many areas and providing a wide range of benefits:

• Early Warning Scoring (EWS)
Medical Device Information systems can easily analyze key patient vital signs and compute and display patient EWS warning scores showing the severity of the patient’s condition. They can also display all individual vital sign scores for immediate nurse review.
Benefits: With EWS data readily available, seriously ill patients requiring greater care can be assigned to specific units to balance nursing workloads. As patient conditions changes, nurses can also easily adjust the frequency of vital sign collection for consistency with patient acuity levels.

• Accurate and Complete Vitals Reporting
Analytics applications can use their own system data to verify on an ongoing basis that the system is functioning properly as well as that all required devices are connected and required parameters are being capture and transmitted downstream in a timely fashion.
Benefits: Enhances patient care by assuring tmiliness, accuracy and completeness of information.

• Medical Device Location and Utilization
Using special analytics software, hospitals can track the specific location of every device and determine utilization rates by location such as units and floors.
Benefits: Equipment can be reallocated to maximize utilization rates if some areas are chronically short on devices, while others have overages. If utilization rates are high across all locations, hospitals can use this information to help justify purchase of additional devices. Biomed staff also are provided with a high level view of where each piece of equipment is situated should they need it for troubleshooting or maintenance.

• Alarm Settings
With an analytics application, clinical staff can see on a single dashboard the current thresholds for all device alarms and adjust them patient by patient as required. Using historical alarm data, they can predict the increase or reduction in alarms with changes in thresholds.
Benefits: Reduces alarm fatigue without compromising patient safety. Often overwhelmed by the din of non-critical alarms, clinicians can suffer audio fatigue, preventing them from discerning alarms signaling a true emergency. Alarm fatigue has consistently been cited as a Top 10 Health Technology Hazard by well-known healthcare analysts ECRI.

• Monitoring Medical Device Information System Health
Data from the medical device information system itself can be analyzed to ensure the all components are maintaining peak performance. This includes monitoring the function of all in- and out-bound interfaces, servers, and battery life and device connectivity. Data can be viewed historically and trended to pinpoint any prior system problems that may have affected system output as well as to see trends that might signal system decline, the need for preventative maintenance or help diagnose the specifics of an existing problem.
Benefits: Helping to remedy potential problems before they happen and impact patient care

As the quantity and quality of data stored in medical device information systems grows, it can be pooled across institutions to and analyzed to create meaningful benchmarks for performance on many parameters discussed above, such as patient outcomes for various disease states, optimal nursing load balancing, cost-effective device utilization, device failures, battery life and alarm management. Statistics from individual hospitals can then be compiled and compared against these benchmarks. This provides medical facilities with the ability to work towards achieving realistic goals that will improve patient care and hospital efficiency, while cutting costs.

With vendors working towards shared standards and longterm archiving, the future for medical device analytics looks bright!