Circulation 60,475 • Volume 15, No. 4 • Winter 2000

The Capture of Anesthesia Machine & Ventilator Data: Problems & Solutions

Ruben G. Derderian

A key element in improving patient safety is the ability to analyze outcomes and understand the type and frequency of events that contribute to or cause undesired outcomes. Today, the retrospective analysis of outcomes provides very limited information because, historically, the only data available is from the patient anesthesia record, recollection of the attending staff and the post-surgical condition of the patient.

Clinicians and the medical industry have long envied the aircraft industry for its forward thinking in utilizing flight data and cockpit voice recorders to collect essential information that could help understand the events and/or circumstances that lead to an undesired outcome.

Many of us today believe we need additional information to help us identify and understand the causes or contributing factors to bad outcomes. With respect to state-of-the-art medical equipment, we have made significant progress in the development of the "flight data recorder" for anesthesia, but still have much work to accomplish before a true anesthesia data recorder is a reality.

While my comments will focus on anesthesia machines and ventilators, the information applies to the majority of all state-of-the-art microprocessor-based equipment used in the OR today.

Microprocessor technology has been around for years but only recently has it been employed in anesthesia machines and ventilators. Historically these devices had been based upon mechanical and pneumatic technology, which placed significant limitations on the ability to collect and store data. Microprocessor-based technology, on the other hand, affords the ability to collect data on a time-related basis for all data that is electronically controlled, monitored or measured.

As microprocessor-based products evolved, the ability to collect, store and exchange data became a reality. Most manufacturers today utilize software protocols to exchange data and control functions via their RS 232 interfaces between two or more medical devices.

The MIB (Medical Information Bus) protocols define the information available for exchange, how that information is structured and how and when it is available for communication or exchange.

Types of Information Available: Machine function, including self-monitoring and error code generation and storage. Commands. Measured data. High and low alarm settings & violations. Alarm messages. Realtime data, excluding waveforms. Device setting. Text Messages.

The resolution and frequency of the data collected and available for exchange is generally device dependent, but in most cases exceeds the requirements to reconstruct a data stream for retrospective analysis.

While communication between separate medical devices has become possible, the process is complicated by the lack of standard communications protocols between manufacturers. Currently, most manufacturers have developed customized protocols that are device to device specific. In an effort to achieve "plug and play" interface capability between independent devices, manufacturers and the IEEE have formed a Standards Committee that has recently proposed standards (IEEE 1073 MEDICAL INFORMATION BUS) that are under review, but have yet to be formally adopted.

Typically, with the exception of hemodynamic monitoring where continuous waveform data is required, most devices sample at a predetermined rate and store the information for a period of time. In the case of microprocessor-based anesthesia machines, service or machine function data requires minimal memory and may be stored indefinitely. Patient and setting data require substantial memory and are usually lost when the machine is turned off, unless provisions are made to transfer the data to permanent memory media.

Having data available to archive is only the first step in the realization of an "Anesthesia data recorder." We still need to define the patient and machine data that should be retained for retrospective analysis or quality improvement studies. Next, we need to define the archival requirement, how long will the data be retained? And last, we need to select and employ a standardized data storage and retrieval system.

The actual implementation of the "anesthesia machine and ventilator data recorder" will be a slow process. The capability to capture the required data will be limited to state-of-the-art microprocessor-based anesthesia machines. Anesthesia machines older than three years are not capable of collecting the required data. The expense of new technology, coupled with the comfort of working with time-proven, reliable technology of older equipment, will be a strong factor against upgrading to new technology.

Further slowing the adoption of the data recorder is the availability of capital dollars and the staff whose responsibility will be the maintenance of the data archiving system.

As anesthesia data recorders become available, retrospective analysis will require the integration of data from separate devices, an investigation team, and, ideally, a simulation laboratory.

The ability to retrospectively analyze data will provide documentation of information captured and recorded, but it will not provide information about the awareness, thought process, human interaction or other non-recorded activity associated with a bad outcome. A voice recorder, preferably a video with voice documentation, will be required in addition to the data recorder.

State-of-the-art microprocessor-based anesthesia machines provide the capability to capture and retain data for retrospective analysis of bad outcomes; however, the expense of upgrading technology and the comfort of working with time-proven, reliable equipment will be factors that limit implementation of anesthesia data recorders on all anesthesia machines.