ctlab - Brave New World

Brave New World: Medical Devices, Clinical Information Systems, Networks, and Patient Safety

Colloquium at the Human Factors and Ergonomics Society National Conference
Wedesday, 28 September 2005, Orlando, FL

Richard Cook, MD (co-chair) University of Chicago
Jay Crowley, US FDA
Christopher Nemeth, PhD (co-chair) University of Chicago
Mark Nunnally, MD University of Chicago
Matt Weinger, MD Vanderbilt University

In contrast with other sectors such as the military and aviation, little attention has been paid to the systemic aspects of healthcare.The research that is necessary to obtain that understanding is performed infrequently. This is because research into healthcare cognition is difficult, for a number of reasons. Information is dense at the sharp end. Practitioners devote more attention to their own science rather than on support issues. Clinical practice changes often and there are many points of view as to how it should be performed. Views on what constitutes the preferred approach to clinical care are subjects of debate. (Nemeth, et. al, in press) As a result, sharp end research takes significant sustained effort.

The desire to create systems that are human-centered (Billings, 1997) is worthwhile, yet is far more difficult than many appreciate. This is what David Woods (1994) meant by noting that “the road to technology centered systems is paved with user-centered intentions.” IT development that is intended to be user-centered that does not succeed leaves us with results that are molded solely by technology considerations. As complexity of clinical care grows, information technology (IT) support spans multiple applications that include medical devices, clinical information systems, and work management, accounting, and documentation. Healthcare provides a useful ground to explore the development and implications of IT at multiple levels.

Equipment—Most infusions of powerful, fast-acting medications are now administered by microprocessor-controlled pumps that must be programmed by clinicians. (Hunt-Smith, et.al., 1999) The current generation of infusion devices incorporates multiple modes of operation, involves substantial operator programming, and contains layered, nested menus with complex branching. Interface designs provide little useful feedback about the state of program entry, the history of operation, or the past or present states of infusion devices. Research in our Lab (Nunnally, et.al., 2004) shows that practitioners have to perform additional work in order to coordinate and program these devices. This presents unforeseen complications with significant implications for patient safety. These traits frequently cause experienced device operators become lost while programming, have difficulty tracking device states, and misinterpret device function. For example, infusion pumps are set-up to administer drugs using parameters that do not match the way the clinicians think about infusions. We have found that operators first figure infusions in terms of flow rate (ml/min), then need to convert to other parameters (mg/kg/min, mcg/kg/hr) in order to program the infusion. As a further example, representations of current pump state and programming paths that are available to reach goal states are often ambiguous. This forces the practitioner to develop coping strategies that are effective when used in specific circumstances, yet are vulnerable to failure in actual use.

Systems— The smooth operation of the OR suite requires a particular kind of expertise, which makes coordination among all team members essential. Nemeth, et. al. (2004) describes how both anesthesia coordinators and acute care team members in an OR suite look at the current state of procedures, what has happened, and what they anticipate will happen in order to manage resources throughout each day of procedures. This occurs constantly through the day. Information displays currently portray procedures according to operating rooms and procedures that are assigned to each room, by time of day. Computer-supported versions of the master schedule have attempted to mimic the hard copy version. In the process, unforeseen complications have developed. Case status information is frequently posted late by a half hour or more, forcing team members to do extra cognitive work as they make in-person trips and phone calls to verify case status. The roster on the electronic display shifts each time a case is deleted, causing the reader to search for cases. The computer interface requires practitioners to drill down as a far as four menu levels to find information that was previously available in one glance using the paper version of the master schedule. Computer-supported displays go “down,” preventing team members from seeing or entering information. This causes great difficulty in making the kind of moment-by-moment decisions that are necessary for resource allocation. The displays offer only a static or a “keyhole” view of the day prevent practitioners from making connections that are a necessary part of their cognitive work, allow only truncated descriptions that create misleading procedure descriptions that have caused departments to plan for the wrong procedure, and do not allow for informal and subtle adjustments to the ways that team members communicate.

Clinical Work Management—Efforts to reduce what has been termed human error include initiatives to manage the human role in care provision through the use of IT. Such systems may be beneficial, but they can also suffer from difficulties such as being unable to handle marginal conditions that are a regular part of patient care. For example, computerized physician order entry (CPOE) relies on a centralized computer system to track and manage the provision of medication. CPOE is intended to create a continuous connection from physician, to pharmacist, to nurse. The approach is intended to reduce causes of medication error by improving the reliability and accuracy of health care system performance. While IT can improve on some difficulties, it can also introduce others. In 2003, 2.9% of reported errors were attributed to a long-suspected cause of medication adverse events: illegible handwriting. However, in that same year, computer entry accounted for 13% of reported errors. Nearly 20% of hospital and health system medication errors that were reported in 2003 involved computerization and automation. (USP, 2004) Efforts are underway to develop facility-wide centralized systems to manage all medications including infusions. The integration of multiple departments and systems that would be involved in such an arrangement poses similar potential difficulties with unforeseen consequences.

New Capabilities and New Problems
As the use of IT grows, the distinctions blur between products and systems, and systems and networks. For example, depending on the way that a problem with an infusion device is described it can be seen as an issue for pharmacy, bio-medical engineering, purchasing, IT, education, or unit manager. This lack of distinction fragments and decentralizes responsibility for a single medical device, as well as the reporting and resolution of problems that are associated with it. As the trend continues, it can be expected to blur the responsibility for the safe introduction and use of a single device.

Equipment and systems that are intended for use by clinicians must necessarily reflect actual clinical practice in order to be suited for use at the (operator) sharp end. The efforts that are required to accomplish this are not simple, as this is the most complex and varied work setting that IT has tried to support. (Rasmussen, 2000) Support for sharp end cognitive work requires attention to the subtleties and complexities of the real world of clinical care that are unforgiving in their consequences. The challenge for human factors professionals is to anticipate the changing landscape these issues present and to foresee opportunities to shape these systems before they are put in place.

References

Billings, C. Aviation Automation: The Search for a Human Centered Approach. Mahwah, NJ: Lawrence Erlbaum Associates.1997.

Christoffersen, K and Woods, D. How to Make Automated Systems Team Players. Advances in Human Performance and Cognitive Engineering Research. Vol.2. 2002. 1-12.

Cook, RI and Woods, DD(1996). Implications of Automation Surprises in Aviation for the Future of Total Intravenous Anesthesia (TIVA). Journal of Clinical Anesthesia, 8:29S-37S.

Hunt-Smith, J, et al. (June, 1999). “Safety and Efficacy of Target Controlled Infusion (Diprifusor) vs Manually Controlled Infusion of Propofol for Anesthesia.” Anaesthesia Intensive Care. 27, 260-4. June, 1999.

Nemeth, C, Cook, R, O’Connor, M, and Klock, PA (2004). Using Cognitive Artifacts to Understand Distributed Cognition. In Nemeth, C., Cook, R. and Woods, D. (Eds.). Special Issue on Studies in Healthcare Technical Work. IEEE Transactions on Systems, Man and Cybernetics-Part A. 34:6. 726-735.

Nemeth, C, Nunnally, M, O’Connor, M, Klock, PA and Cook, RI (in press). Making Information Technology a Team Player in Safety: The Case of Infusion Devices. Advances in Patient Safety: From Research to Implementation. Agency for Healthcare Research and Quality. Washington,DC.

Nemeth, C. Nunnally, M, O’Connor, M, Klock, PA, and Cook, RI (in press). Getting to the Point: Developing IT for the Sharp End of Healthcare. Journal of Biomedical Informatics.

Nunnally, M, Brunetti, V, Woods, D and Cook, R “Infusion Device Characteristics Related to User Error during Programming and Operation Determined by Finite State Modeling.” Anesthesiology, 96: A520; 2002, October.

Nunnally, M, Nemeth, C, Brunetti, V, and Cook, R (2004). Lost in Menuspace: User Interactions with Complex Medical Devices. In Nemeth, C., Cook, R. and Woods, D. (Eds.). Special Issue on Studies in Healthcare Technical Work.IEEE Transactions on Systems, Man and Cybernetics-Part A. 34:6. 736-42.

Rasmussen, J (2000). The Concept of Human Error: Is It Useful for the Design of Safe Systems in Healthcare? In Vincent, C. and DeMol, B. Safety in Medicine. Kiddington, Oxford, UK: Elsevier Science.

USP (2004). MEDMARX 5th Anniversary Data Report: A Chartbook of 2003 Findings and Trends 1999–2003. Rockville, MD: The United States Pharmacopeial Convention.

Woods, DD. Cognitive Demands and Activities in Dynamic Fault Management: Abduction and Disturbance Management. In Stanton, N. (Ed.). Human Factors of Alarm Design. London: Taylor and Francis. 1994.